PDA

View Full Version : Sneak preview on Chinese off the shelf CNC.



John Stevenson
10-29-2007, 07:12 PM
http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/KX1front.jpg

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/KX1side.jpg

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/KX1rear.jpg

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/KX1cover.jpg

Not certain just how much the off the shelf model will be to this one as it's down to the factory to sort production out.

Based on the X1 small mill it has 220 oz/in motors and 4mm pitch ballscrews, angular contact bearings and limit switches on all axis.

The spindle is belt driven from a 3 phase 600w brushless motor with feedback and dynamic breaking. Usable speeds are from 140 to 4,500 rpm.
Max speeds are 100 in / min in rapids.

The whole machine is self contained and only requires a power cable and a parallel cable from the computer.
Running on Mach3 via a breakout board to four independent 3 amp drivers, [ this one is 4th axis capable ] and no I have no idea on cost.

.

lazlo
10-29-2007, 07:29 PM
Pretty neat John -- a miniature Tormach :)

Were you involved in the design? Is it from Sieg?

JRouche
10-29-2007, 08:19 PM
Yeah, thats a darn good lookin machine. I like the most obvious piece on it, right in yer face.. E STOP.. Haaa.. JRouche

Oh, and I like how they drive the X-axis from the rear, good for the big bellied folks...

hitnmiss
10-29-2007, 09:00 PM
Nice, whats the spindle taper?

tattoomike68
10-29-2007, 10:08 PM
http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/KX1front.jpg

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/KX1side.jpg

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/KX1rear.jpg

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/KX1cover.jpg

Not certain just how much the off the shelf model will be to this one as it's down to the factory to sort production out.

Based on the X1 small mill it has 220 oz/in motors and 4mm pitch ballscrews, angular contact bearings and limit switches on all axis.

The spindle is belt driven from a 3 phase 600w brushless motor with feedback and dynamic breaking. Usable speeds are from 140 to 4,500 rpm.
Max speeds are 100 in / min in rapids.

The whole machine is self contained and only requires a power cable and a parallel cable from the computer.
Running on Mach3 via a breakout board to four independent 3 amp drivers, [ this one is 4th axis capable ] and no I have no idea on cost.

.

hey, I saw one at Wal Mart in a blister package, it came with a USB cable and barbie CAM software.

No... it looks like a nice home unit, any of us home guys would love to have that.

Spin Doctor
10-29-2007, 11:49 PM
It was only a matter of time.

A.K. Boomer
10-30-2007, 12:29 AM
For some reason that whole entire thought has escaped me, its not that I never realised they were or werent into the CNC market its more as iv nevr even thought about it --- but this brings to my attention that I guess they havent been? and why not? theyve gotten into everything else, it does kinda blow me away, So will they have machines to compete with the likes of Haas? this is insane, its never gonna end. No let me restate that Ittl end and ittl do it quite abruptly...

John Stevenson
10-30-2007, 04:42 AM
A few answers.

Spindle taper on this one is No 2 morse. They are also doing a larger version based on the SX3, again with 3~ brushless motor that will have R8 taper.
The spindle taper on this one is self extracting, as you undo the drawbar it pushes the No 2 Morse out.

The reasoning behind this is that many of their manual machines have found their way into private hands and been converted.
This is also a sub industry grown up for kits and even complete machines by third party providers.
Not always with the correct support though.

Sieg and the big box players like Grizzly have identified there is a market out there for an off the shelf machine and these two have been produced to fill that gap.

Now this is where it gets hazy as I have no idea on thier marketing but this is what I believe.
There will be three machines in each range but who stocks what is donw to Sieg and their customers.
Base model will be the KX1 or 3 -NU which is a basic machine with ball screws, motors, spindle motor etc but no drive electronics. This is for the roll your own guys.
Next is the same machine with a DOS controller made by Siemens in China with either a Chinese screen for home market or English for export. This will be a turnkey but with a limited conrol.

Lastly is the same machines with full electronics and Mach3 installed, either Chinese or English. This is the flagship model.

The machine in the pic was the NU model converted to Mach3, all the conversion notes and remedial work has been shared with the Chinese to enable their engineers to mirror this.
Now whether the production version comes out like this one I have no way of knowing, that's an internal decision but it should only be cosmetic.
They may even go for a remote driver electronics box as the NU was originally set up this way.

AK,
They have been in the CNC market for a while now, it's just that their home market is large enough to take all they can make.
In a recent sales blurb I received on generators it showed their factory with rows of CNC machining centers, all with China Fadal on them.
One think they don't lack is drive and money.

.

small.planes
10-30-2007, 05:01 AM
Hi John,
Its good to see you finally got the covers back on ;)

Dave

S_J_H
10-30-2007, 08:35 AM
These machines are going to be very hot sellers IMO! As it is now machines like the x3 mill sell about as fast as they ship them over here. I bought my x3 with no other intent but to convert to cnc. I think I heard they may even offer up the 10x22 c6 lathe or maybe a smaller lathe as a cnc package.

CNCzone.com is quite a large site now. Huge interest in smaller cnc home shop machines! I don't understand why this site still does not have a cnc subforum to sort of go with the new digital machinist magazine.

Steve

bob_s
10-30-2007, 09:52 AM
So will this be the price point?

http://cgi.ebay.com/4-Axis-CNC-Mill-Tabletop-Precision-New-Complete_W0QQitemZ220164534596QQihZ012QQcategoryZ1 2584QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

Or the next US business extinction?

lazlo
10-30-2007, 10:12 AM
So will this be the price point?

http://cgi.ebay.com/4-Axis-CNC-Mill-Tabletop-Precision-New-Complete_W0QQitemZ220164534596QQihZ012QQcategoryZ1 2584QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

The next business extinction.

Considering that the 1400 lb Tormach sells for $6800, I think this Sieg X1 CNC will sell for a lot less than $2500...

lwalker
10-30-2007, 10:55 AM
I'd rephrase that as "the next business extinction unless he finds a point to compete on besides price!"

There are plenty of people out there for whom price isn't the be all and end all. I'd even hazard a guess that there's a market in buying these Chinese CNC machines and modifying them for other uses: high speed spindles for milling/drilling PC boards; precision glue or solder paste applicators; etc.

Just like the X1 is a cheap base for products out there so too can the CNC'd X1 be a base for more advanced products.

It's still way too early to give up!

noah katz
10-30-2007, 01:41 PM
"http://cgi.ebay.com/4-Axis-CNC-Mill-...QQcmdZViewItem"

The machine weighs 170 lb, and look at that skinny column.

I don't see how you could reasonably do anything except plastic or light cuts in soft metals.

JimB
10-30-2007, 08:42 PM
That looks very nice. I built a CNC X1 with the big table from Little Machine Shop. I used the screws that came with it. Also used a turned up belt drive spindle, Mach3, SheetCAM and a Xylotex driver. You can sink some money in a homemade job. I gave it to the High School Robotics team I'm a mentor at. Great starting point for the kids who picked up on the programming right away. Once I got them started they made that thing work hard cutting small aluminum parts for the robot. It will cut aluminum and is a great place to start. It will make you want a Tormach tho....really bad. After seeing it work one of the other senior mentors found a Tormach locally for sale and got it, he didn't give it to the school tho. (smiley face here)

Me, I'm back using my lathe built around 1910 that Wierd had stored on the back 40 of his ranch. Jim

ptjw7uk
10-31-2007, 09:44 AM
Anyone any idea how much it will go for as you can get an x3 with 4th axis for 3150 on flea bay.
I may well wait and see what they go for, still undecide about putting ball screws on my x1 as there is not much room for the ballnut on a 12mm ballscrew. So far not found out a reasonable price on a 10mm ballscrew.

Peter

MikeHenry
10-31-2007, 01:23 PM
Considering that the 1400 lb Tormach ...

Actually, the Tormach weighs around 1100 lbs.

Mike

lazlo
10-31-2007, 02:35 PM
Actually, the Tormach weighs around 1100 lbs.

Sorry Mike -- I thought I had seen 1400 lbs on the Tormach web site, but maybe that was shipping weight.

My point though was that the Tormach is a pretty sturdy CNC outfit, and it sells for $6800 (which is a steal, in my opionion), so you would think a little X1 mill would sell for a lot less...

oldtiffie
10-31-2007, 11:59 PM
I am keeping an open mind here and trying not to prejudge or condemn anything until I've seen it - "it" being CNC - on a mill.

I have a HF45 "clone" - if you will - mill which is quite sturdy and does a good job - a real good job. But the only "digital enhancement" it will get is the Easson 3-axis DRO when ever I get around to fitting it.

I have another very good smaller mill on my "3-in-1" which has a very good 250-900 lathe on it as well. I also have a very good "mini-lathe". None of these will be CNC-ed - "DRO-ed"? - maybe.

I am interested in the concept of the "turn-key" job that John Stevenson showed us as I'd like to be able to "put my toe in" and "dabble" with a dedicated CNC set-up which I'd like to use as CNC should be used for lighter jobs.

I trust John's advice - all the time - every time.

My days and need for smokin', smellin' "blue-chip" production are dead and buried and that's how they are staying.

I can see the need for CNC and I'd like to try it and get some-thing useful done that might be better made on non-CNC or even do some stuff on CNC that is a "CNC only" job and some of either. Perhaps do a job several times in different ways, - CNC and non-CNC and learn a lesson or two and if that's it - scrap it. If its usable or useful - use it.

I've got stuff in my shop that I've either never used (never will?) or has not been used for quite a while that I like to "brush up" on.

I like to "make" or "do" things when ever I have the time and a need to do it.

Sometimes my Shop doesn't see me for weeks. That's fine so long as its there when I need it all is well in the world.

Whether there is use or rhyme or reason for what or how I do things is of no consequence.

I just like to try things out and keep old skills up and try new ones.

If I ditch everything that I did a for a year and had nothing of substance to show for it - so what? I've enjoyed the ride. (AutoCAD is a prime example - see a previous post).

If I get bored shi*less in my shop - I just clean it up and leave it until I want to use it again.

I utterly respect how anyone else deals with matters pertaining to their shop.

I enjoy this forum as it is a huge asset in many ways - some tangible, others not. It sure helps me in my shop and new ideas and approaches that I need to think about and perhaps "try".

CNC is very much in this category - if the time, cost, specifications and need are all OK at the time.

oldtiffie
11-01-2007, 07:23 AM
John S.

Any idea of the specs and capacity of that machine and if so can or will you post them please.

Also- any idea of when and where distribution will start or if started what the addresses (international) are?

Any indicative costings?

Thanks.

Swarf&Sparks
11-01-2007, 07:28 AM
I'll second that request Mick.
Cheers John.
Rgds, Lindsay

dp
11-01-2007, 09:44 PM
There is a possible consequence of there being a wide-spread commodity tool like this. Imagine there are tens of thousands of these in shops and garages around the world. The next thing that will spring up is websites offering g-code for sale and as open source, distributed over the Internet, for far flung users to use and tweak to produce widely popular things.

Go with open source for a bit. What is the first thing the OS crowd is going to make? More CNC Machines. 20,000+ volunteers building bits and pieces of machines and giving away the bits is a pretty big free labor pool.

Then will come the commodity 3D scanners that allow pirating existing products. Banks of lawyers will form up and start the lawsuit process. Congress will pass laws forbidding plundering commercial work (RIAA does it with digital music, why not?). The Homeland Security people will see this as everyone being able to machine perfect baseball sized spheres of plutonium or coin stamping machines, and they will declare such machines to be munitions or tools of counterfitters, illegal for domestic use or export.

Interesting to think about a world where nothing is immune to cloning, anyway.

There might be some tongue in cheek humor in this post. Add SFSFs as appropriate.

oldtiffie
11-01-2007, 10:27 PM
There is a possible consequence of there being a wide-spread commodity tool like this. Imagine there are tens of thousands of these in shops and garages around the world. The next thing that will spring up is websites offering g-code for sale and as open source, distributed over the Internet, for far flung users to use and tweak to produce widely popular things.

Go with open source for a bit. What is the first thing the OS crowd is going to make? More CNC Machines. 20,000+ volunteers building bits and pieces of machines and giving away the bits is a pretty big free labor pool.

Then will come the commodity 3D scanners that allow pirating existing products. Banks of lawyers will form up and start the lawsuit process. Congress will pass laws forbidding plundering commercial work (RIAA does it with digital music, why not?). The Homeland Security people will see this as everyone being able to machine perfect baseball sized spheres of plutonium or coin stamping machines, and they will declare such machines to be munitions or tools of counterfitters, illegal for domestic use or export.

Interesting to think about a world where nothing is immune to cloning, anyway.

There might be some tongue in cheek humor in this post. Add SFSFs as appropriate.

Well dp,

if they are "cloning" its OK - but if they are "reproducing", I hope they do it in the dark.

oldtiffie
11-01-2007, 11:21 PM
There is a possible consequence of there being a wide-spread commodity tool like this. Imagine there are tens of thousands of these in shops and garages around the world. The next thing that will spring up is websites offering g-code for sale and as open source, distributed over the Internet, for far flung users to use and tweak to produce widely popular things.

Go with open source for a bit. What is the first thing the OS crowd is going to make? More CNC Machines. 20,000+ volunteers building bits and pieces of machines and giving away the bits is a pretty big free labor pool.

Then will come the commodity 3D scanners that allow pirating existing products. Banks of lawyers will form up and start the lawsuit process. Congress will pass laws forbidding plundering commercial work (RIAA does it with digital music, why not?). The Homeland Security people will see this as everyone being able to machine perfect baseball sized spheres of plutonium or coin stamping machines, and they will declare such machines to be munitions or tools of counterfitters, illegal for domestic use or export.

Interesting to think about a world where nothing is immune to cloning, anyway.

There might be some tongue in cheek humor in this post. Add SFSFs as appropriate.

Thanks dp.

As I read one of your posts today, I realised that you were "into" Turbo Pascal (TP).

Now that being so, perhaps you programmed in "Forth", in which case all this "cloning" may only be someone who took his "Command" from "on high" a little too seriously and far too literally - and did it - and hence the scenario you envisage/d - out of control(led) "cloning".

The Command was "Go Forth and multiply" if I recall correctly.

I'm off - I can hear the Cavalry coming (clippity-clop-clippity-clop). I'm off to the stampede to Calvary (not Canada!! - that's Purgatory!!! Guess who's there!!!).

lazlo
11-01-2007, 11:27 PM
Imagine there are tens of thousands of these in shops and garages around the world. The next thing that will spring up is websites offering g-code for sale.

That's a fascinating idea DP. Take Daryl's Model Engineering scenario -- you visit a guy's table where not only did he "program" a pretty little model engine, but he'll sell you the G-Code so you can build an exact copy for $49.99 :)

I'm curious if G-Code is for sale in the commercial CNC world. Due to realities of the machine, I doubt G-Code would be portable across machines, especially of different sizes and horsepower, but if everyone can run down to Harbor Freight and buy a Sieg CNC machine for $X,XXX dollars...

oldtiffie
11-01-2007, 11:39 PM
I am not very bright - period.

That's a given.

This "G" stuff has me intrigued.

Is there a "G" spot?

Is "G-code" code for or can it actually be a "G-string"?

Now put "G strings" in with "G code", press the "G-spot/s" and with hardly a slap of a "Whale's Tail" they'll be off duplicating and reproducing like crazy - in their millions - just like in China???

Or have I go it wrong again?

Anyway.

I'm off back to see if there is any progress on the "Sneak preview on Chinese off the shelf CNC" discussion.

See ya there.

dp
11-01-2007, 11:54 PM
I am not very bright - period.

That's a given.

This "G" stuff has me intrigued.

Is there a "G" spot?

Is "G-code" code for or can it actually be a "G-string"?

Now put "G strings" in with "G code", press the "G-spot/s" and with hardly a slap of a "Whale's Tail" they'll be off duplicating and reproducing like crazy - in their millions - just like in China???


Hehe -- hopefully we won't have to wait to see what CNC standards the porn industry adopts as happened with VHS vs BetaMax and BlueRay vs HD DVD.

lazlo
11-01-2007, 11:56 PM
This "G" stuff has me intrigued.

Is there a "G" spot?

Fortunately, the two are not related. G-Code is essentially how you program a CNC machine to make a part: it's a set of standardized codes which tell the CNC controller what type of action to perform, such as:


Change a pallet
Set tool information such as offset
Rapid move
Controlled feed move in a straight line or arc
Series of controlled feed moves that would result in a hole being bored, a workpiece cut to a specific dimension...


Here's a couple examples of G-Code. The standard is controlled by the EIA, and the industry tried to standardize around an old Fanuc controller, but most CNC vendors make proprietary changes to the G-Code, so it's not really portable across machines:

G1 X1.0 Y1.0 F20.0 ---- go to X1.0, Y0.0 at a feed rate of 20 inches/minute
G2 X1.0 Y0.0 I0.0 J-1.0 ----go in an arc from X0.0, Y1.0 to X1.0 Y0.0, with the center of the arc at X0.0, Y0.0
G1 X0.0 Y1.0 F20.0 ----go to X1.0, Y0.0 at a feed rate of 20 inches/minute
G2 X1.0 Y0.0 R1.0 ----go in an arc from X0.0, Y1.0 to X1.0 Y0.0, with a radius of R=1.0

The beauty of Mach is that, as a software CNC controller, the Mach G-Codes are a defacto standard for the hobby CNC world.

So, in theory, if one person wrote a G-Code program for a series of parts for a model engine that were run on one Sieg X1 CNC machine, the same G-code should work on any other Sieg X1 CNC machine.

dp
11-02-2007, 12:02 AM
That's a fascinating idea DP. Take Daryl's Model Engineering scenario -- you visit a guy's table where not only did he "program" a pretty little model engine, but he'll sell you the G-Code so you can build an exact copy for $49.99 :)

I'm curious if G-Code is for sale in the commercial CNC world. Due to realities of the machine, I doubt G-Code would be portable across machines, especially of different sizes and horsepower, but if everyone can run down to Harbor Freight and buy a Sieg CNC machine for $X,XXX dollars...

Someone just needs to come up with a CNC machine that is the IBM PC of CNC machines and the rest will happen on its own. Unless the government helps.

For home shop wood workers this comes close and is on the right track:
http://www.sears.com/shc/s/p_10153_12605_00921754000P?vName=Tools&adCell=A3

lazlo
11-02-2007, 12:07 AM
Someone just needs to come up with a CNC machine that is the IBM PC of CNC machines and the rest will happen on its own. Unless the government helps.

Ugh, a government standard would suck! Remember ADA? :) I can just imagine Homeland Security restrictions on what you're allowed to CNC on Government G-Code :mad:

I thought about that Sears 2.5D CNC router last night -- if Sieg is really going to make these CNC'd X1's by the thousands, seems like they would make a whole lot more money by getting a Western company to re-brand them, like Sears did with the CNC router.

I can just picture Bob Villa pitching the "Sears Craftsman Professional Desktop Milling System." :D

dp
11-02-2007, 12:19 AM
Ugh, a government standard would suck! Remember ADA? :) I can just imagine Homeland Security restrictions on what you're allowed to CNC on Government G-Code :mad:

I do remember ADA very well and is exactly what I was thinking of when I wrote that! I took an ADA class at the University of Washington several years ago (gad - 15!) and was astonished at what a committee can do to languages.



I can just picture Bob Villa pitching the "Sears Craftsman Professional Desktop Milling System." :D

That's another thing - there needs to be a Bob Villa/Norm Abrams face on this. I was thinking someone like Angelina Jolie in bib overalls.

oldtiffie
11-02-2007, 12:21 AM
Fortunately, the two are not related. G-Code is essentially how you program a CNC machine to make a part: it's a set of standardized codes which tell the CNC controller what type of action to perform, such as:


Change a pallet
Set tool information such as offset
Rapid move
Controlled feed move in a straight line or arc
Series of controlled feed moves that would result in a hole being bored, a workpiece cut to a specific dimension...


Here's a couple examples of G-Code. The standard is controlled by the EIA, and the industry tried to standardize around an old Fanuc controller, but most CNC vendors make proprietary changes to the G-Code, so it's not really portable across machines:

G1 X1.0 Y1.0 F20.0 ---- go to X1.0, Y0.0 at a feed rate of 20 inches/minute
G2 X1.0 Y0.0 I0.0 J-1.0 ----go in an arc from X0.0, Y1.0 to X1.0 Y0.0, with the center of the arc at X0.0, Y0.0
G1 X0.0 Y1.0 F20.0 ----go to X1.0, Y0.0 at a feed rate of 20 inches/minute
G2 X1.0 Y0.0 R1.0 ----go in an arc from X0.0, Y1.0 to X1.0 Y0.0, with a radius of R=1.0

The beauty of Mach is that, as a software CNC controller, the Mach G-Codes are a defacto standard for the hobby CNC world.

So, in theory, if one person wrote a G-Code program for a series of parts for a model engine that were run on one Sieg X1 CNC machine, the same G-code should work on any other Sieg X1 CNC machine.

Thanks lazlo - understood.

I hope that THIS "Industry standard" works better than some others I can think of: so-called but ever-changing "compatibly" between various CAD suppliers as regards importing, working on, amending/revising/updating say DXF and *.dwg and similar files between various CAD packages that have "adopted" these "industry standards" which are often amended/revised in anything but a regulated industry-wide coherent manner.

The first a user knows of this is when they either import a file in an "Industry standard" format/ting, go to use it and it "crashes" or when another user is most irate that his different (say CAD) package either won't recognise your file, can't open it, or reduces all of the layering to a "flat" package or sets all layers and entities to layer zero. There are more.

I just hope the "Industry standard" Mach G-Codes which as you say "are a defacto standard for the hobby CNC world" are a better indicator of some other "de facto" states and "standards" than some I can think of have been.

I'd guess that "de-bugging" would need to be a good skill to have as well as just "programing".

I like the "Mach" concept as it appears to be.

I can accept proprietory hard-ware and soft-ware as long as I know about it.

I could live with "Mach" in a "non-cohabitable" "de facto" way I suppose - as long as "marriage" and "consumation" are not pre-requisites.

oldtiffie
11-02-2007, 12:54 AM
............................
..............................

That's another thing - there needs to be a Bob Villa/Norm Abrams face on this. I was thinking someone like Angelina Jolie in bib overalls.

You're WHAT?

You gonna put AJ "through the mill" (or the "wringer").

Well!! I never!!! If THAT'S how YOU get your "Jollies" ...................... (half ya luck!!!!).

Give me the bib as AJ can do better without it - I need the bib for my dribbling (drooling??).

That was just another diversion (yeah, I know - mine - again) but I'm off back to the "Sneak preview on Chinese off the shelf CNC".

John Stevenson
11-02-2007, 05:00 AM
Fortunately, the two are not related. G-Code is essentially how you program a CNC machine to make a part: it's a set of standardized codes which tell the CNC controller what type of action to perform, such as:
Change a pallet
Set tool information such as offset
Rapid move
Controlled feed move in a straight line or arc
Series of controlled feed moves that would result in a hole being bored, a workpiece cut to a specific dimension...Here's a couple examples of G-Code. The standard is controlled by the EIA, and the industry tried to standardize around an old Fanuc controller, but most CNC vendors make proprietary changes to the G-Code, so it's not really portable across machines:

G1 X1.0 Y1.0 F20.0 ---- go to X1.0, Y0.0 at a feed rate of 20 inches/minute
G2 X1.0 Y0.0 I0.0 J-1.0 ----go in an arc from X0.0, Y1.0 to X1.0 Y0.0, with the center of the arc at X0.0, Y0.0
G1 X0.0 Y1.0 F20.0 ----go to X1.0, Y0.0 at a feed rate of 20 inches/minute
G2 X1.0 Y0.0 R1.0 ----go in an arc from X0.0, Y1.0 to X1.0 Y0.0, with a radius of R=1.0

The beauty of Mach is that, as a software CNC controller, the Mach G-Codes are a defacto standard for the hobby CNC world.

So, in theory, if one person wrote a G-Code program for a series of parts for a model engine that were run on one Sieg X1 CNC machine, the same G-code should work on any other Sieg X1 CNC machine.

What Lazlo has said is correct, there are certain standards but Mach has taken most on board very broadly.

In Lazlo's example above there are two ways of describing an arc, one using the letters I and J and incremental points and the other using R for radius.

Now some controllers [ Mach3 is a controller ] need to use either but mach can accept both and work with whatever it's given so in Lazlo's code above Mach will read this OK.

The biggest differences occur over tool handling, speeds and feeds.
Fortunately for tool handling that bit of the code is at the beginning and easily editable.
Speeds again are usually at the top but not all small machines can handle automatic speed, they require the speed to be set by belts or pot manually before the start. The Sieg is fully automatic on speeds, everything is controlled by the computer [ mach ].

Feeds are different as the plunge feed for the vertical axis [ Z ] is usually a lot less than X and Y so you get feed changes all thru the program and something like an engraving file with literally 1,000's of small line moves will have these spread over possibly 30,000 lines :D

However all isn't lost as notepad will allow a quick search and replace to handle this.

My large mill can read Mach code with only two lines for tool hight having to be altered, then speeds and feeds.

The same applies in the other direction.
Someone posted the code of a big sprocket he was machining on a late model Fagor machining centre on PM recently.
I saved the code, opened it up in Mach and ran this on the screen to see. It ran fine other than the usual tool, speeds and feeds and the fact that it was about 20" diameter and wouldn't fit on the bed :rolleyes:

However if I had changed the tool, speed and feeds as layed out above, then clicked the scale button at the side of the DRO's in Mach, entered 0.1 for each scale my little piddly machine would then have made a 2" sprocket from the same code.

So no sharing code isn't as far away as it sounds, I have sent code of various badges to people, it's only one step away from sending DXF files from CAD anyway.

I had this exact conversation with David Clark, the current editor of MEW and he's all for it but they won't publish in MEW as models are the domain of ME, MEW is just for workshop and tooling items.

I can see the first things being something like a simple steam engine of the type beginners already do.

On the Sieg X3 mill we convert to CNC we sent a blank plate out and the code for the new user to cut a name plate to cover the hole where the head handwheel used to be as a starter to get used to the machine.

.

oldtiffie
11-02-2007, 06:10 AM
John,

I guess that the Mach software incorporates the CAD/CAM/Modelling as part of the total "Turn-key" integrated package.

Is that so?

Can you please post the specs, capabilities etc. etc. of the machine you posted (the new one)?

TIA

John Stevenson
11-08-2007, 04:17 PM
Tiffie me old mate,
Sorry for the delay but been getting some details.

http://www.siegind.com/cnc_machine.pdf

This is from Sieg themselves and hasn't been proof read for western eyes as yet :D so enjoy the Chinglish.

Technically it's pretty correct and should cover the machines that are liable to be seen in the US and OZ.

As far as I know the KX6 lathe is China market only at the moment.
We have been promised one to convert to Mach but at the moment we have too much on to push it.

.

lazlo
11-08-2007, 05:36 PM
http://www.siegind.com/cnc_machine.pdf

Wow, they do have a pre-packaged X3 CNC system (and not just the little X1 system)!
This is going to rock the hobby CNC community. Anyone who's selling X1 or X3 CNC kits...

So I guess the picture on the last page explains why you were in a hurry to CNC that picture of the Sieg CEO :D

Edit: Smithy, the US importer of the Sieg CNC systems, already has the Sieg X1 and X3 CNC systems advertised for sale:

http://www.smithy.com/product_order.php?cid=11&scid=16&pid=1006

The X1 CNC system ("SmithyCNC 516 Bed Mill") is $3,395.00
The X3 CNC system ("SmithyCNC 622 Bed Mill") is $4,695.00

For comparison, you can buy a complete Super X3 CNC system from Syil for $3,995:

http://www.syilamerica.com/catalog/product_info.php?cPath=1&products_id=28

Let the price wars begin! :D

By the way, here's a video of the Syil X3 CNC. 26 IPM with .1 DOC in aluminum. Bizzare toolpath -- I don't know what CAM he was using:

http://www.youtube.com/watch?v=yv3-DwEHh5o

John Stevenson
11-08-2007, 06:18 PM
http://www.siegind.com/cnc_machine.pdf

Wow, they do have a pre-packaged X3 CNC system (and not just the little X1 system)!
This is going to rock the hobby CNC community. Anyone who's selling X1 or X3 CNC kits...

Damn, never though of that and I make X3 CNC kits :(
Ah well another flat battery in the car park of life.

The whole idea of the factory CNC machines IS the fact they are not happy with some of the kits. OK the guy who buys a machine and converts it himself is no problem in fact he'll be catered for with the suffix NU machine which is a bare machine, NU = nude.

Where they are not happy are the people buying these machines adding a kit of spurious parts that don't necessarily match the machine, taking the money and leaving the punter to sort all the problems out.
There are some right horror stories on CNCZone where the punter has had to sort his own problems out.
Where they want to score is getting a reliable machine out there with a support network in place.





So I guess the picture on the last page explains why you were in a hurry to CNC that picture of the Sieg CEO :D

At last fame, well for my travel bag in any case, that's the one on the chair in the foreground.

.

lazlo
11-08-2007, 06:22 PM
Damn, never though of that and I make X3 CNC kits :(

Can't you arrange to be the UK distributor? :)


At last fame, well for my travel bag in any case, that's the one on the chair in the foreground.

You should have superimposed yourself in the background John ;)

bob ward
11-08-2007, 06:35 PM
Sir John, the plug & play X3 CNC mill should be available in the western world in 6 to 12 months?

lazlo
11-08-2007, 06:36 PM
Bob, look at my post immediately above John's. According to Smithy, you can buy it right now...

http://bbs.homeshopmachinist.net/showpost.php?p=314953&postcount=37

John Stevenson
11-08-2007, 06:50 PM
Sir John, the plug & play X3 CNC mill should be available in the western world in 6 to 12 months?

As Lazlo says it looks as if Smithy has them, whether they do or they order on demand I have no idea.
I may find a lot more out at the end of this month when I have to go back to China to sort some things out with the UK import machines.

I would guess that you are probably correct with that time scale figure though.

.

Spin Doctor
11-08-2007, 06:53 PM
Lets see. The Tormach is about 2,200 bucks more. For that you get a bigger table and I would asume a more robust spindle.

John Stevenson
11-08-2007, 07:09 PM
Lets see. The Tormach is about 2,200 bucks more. For that you get a bigger table and I would asume a more robust spindle.

Yes a lot more robust.
I installed a Tormach for a friend as he couldn't handle it. It weighs a genuine 1/2 a ton so it needs a really solid bench.

Having roughly the same conversation today with a guy I repeated that there is always a market for any size machine.
For some the Tormach is too big to heavy, for others the X3 is too big and too heavy and we have even had some say the X1 was too big for what they wanted so there is a market for all sizes.

Costs are semi relevant as well, I think it was on this group that I read machines are like buildings, all the cost is in the corners :D

Some years ago I bought a VCR for about 250 it could record one program, now they record 10 programs a year in advance and have far more features and cost 100.
So that means a simple one should cost 10 ?

No afraid not as the cost of converting a small mill isn't much less than converting a larger mill, you just have to do your homework and make sure you buy whatever you need first time out,- whatever that is .

.

oldtiffie
11-08-2007, 08:10 PM
Tiffie me old mate,
Sorry for the delay but been getting some details.

http://www.siegind.com/cnc_machine.pdf

This is from Sieg themselves and hasn't been proof read for western eyes as yet :D so enjoy the Chinglish.

Technically it's pretty correct and should cover the machines that are liable to be seen in the US and OZ.

As far as I know the KX6 lathe is China market only at the moment.
We have been promised one to convert to Mach but at the moment we have too much on to push it.

.

Thank you John - sincerely.

The pdf file and your advice are "spot on" - as usual. I have no problem with the "Chinglish" as its a damn side better that a lot that is sourced and developed in "English-speaking" countries. The information and specifications etc. are all there and very well presented.

I don't give a "big rat's ar$e" about where it comes from or who makes it - as long as it does the job - which I can confidently say it will - in "Spades" - for ME, in my HSM shop as I want to use it.

So.

Based on your advice/comment, the pdf file and my requirements - and providing that the "local" (OZ) dealer and other support is professional and there when required then I am quite at ease with paying the "going rate" for those "goods and services" as I may require.

I am interested in the CNC mill package as a total/"Turn-key" set-up an this covers my requirements - beautifully.

My guess is that I'd buy most of the accessories as well - including the tools and the rotab etc. I'd make what I need if its worth the effort but otherwise I'd just buy them in and get on with getting on with the machine.

I also need to have full and competent Dealer support - parts/spares/accessories as well as Technical either "on-line" or able to be called out or called on. I value that sort and level of service and support. I have it here with my suppliers. I am quite happy to pay the "going rates" for such professional services as I may require from time to time.

I just just want a complete system that I can have installed and then get on with having a "play-around ", and a "buggerise-around" with the machine and software that will let me see what I can do with it in my HSM shop. It will work in conjunction with my HF(here)-45 vertical mill, my mill and lathe on my 3-in-1 machine, my surface grinder, tool and cutter grinder, cutting and welding gear etc. as these machines will remain my primary focus for "making stuff".

FWIW, most of the stuff I make is either for "making (other) stuff" or just a "try-out" to see if or how something "works" or not and more often than not just toss it into the trash or "spare-stuff" bins. I am more interested in methods and processes than usable end results. But it is a bonus if it is "usable".

The CNC mill will be another interest that can be explored and used in conjunction with those primary machines. It will be to make and do things that those machines can also do and some things that are best difficult to do on them - if they can be done at all on them.

The small-ish CNC mill will meet all of those needs - perfectly.

The power supply requirement suits us in OZ. My supply is 230v 50~ single phase with 60A per phase. I was considering getting a 3>1 (OZ-made) converter but I may not have to - but I will if I need to.

The 3MT quill suits me as all of my tooling is 3MT to suit my mills (2) - so I am well in front there.

I would get most of the accessories although I could make some of them.

I have not got nor will I ever have the skill-set/s required to make or set up a CNC machine of any manner shape or form. I admire and respect those that have and use those skills. I am more a "Taxi driver" than a "Motor/computer mechanic or diagnostician or Technologist" in that regard.

This set-up lets me - more or less - just buy it in, have it set up and set to work for me, and lets me just "switch her on, start her up and get going". That's more or less me to a "T".

Another big factor is that I have limited time. The end of time and luck are getting closer - perhaps very close. Perhaps should have been here already. So perhaps I am on borrowed time but I want to be able to make the most of what is left of it and what-ever faculties I have from here on in.

I just want a set-up such as in the pdf file that I can use in my HSM small shop as and when I feel like it - and this set-up seems to fill the bill nicely.

Needless to say, I would appreciate "ball-park" advice (say +/- 3.6 months for time and +/- 30% for cost) as soon as you are able to let me know.

I fully understand the needs for business ethics and commercial confidentiality in all their shapes and forms and would neither want to or have you breach or/nor compromise them or yourself in any way on my behalf.

Keep up the excellent work and advice John, as it - and you - are held in very high regard and esteem here - despite what I may say from time to time.

I think I can appreciate your effort on our and my behalf in putting in the time and effort that you do for us on the HSM forum.

Thank you sincerely.

Keith Krome
11-09-2007, 02:36 AM
Regarding G-codes.

The basic movements are standardized, however, the canned cycles (threading for example) use different G-codes depending on the particular control manufacturer's preferences. This is why CAM software needs a post processor for specific machines. The post processor formats a tailored G-code for a particular machine.

I've seen 2 different CNC machines (lathes in this case) with different controllers. Both are native G-code machines (as opposed to "conversational" programming). However, because the upper G-codes are left open to the control builder in the standard, a particular G-code on one control can have a completely different function on another control.

John Stevenson
11-09-2007, 10:32 AM
Here's another link to the SX 3 with some decent pictures.

http://www.machsupport.com/forum/index.php?phpsessid=ijvg3sokenbt0kgvqbk5qpo6d0&topic=3446.0

.

speedsport
11-09-2007, 10:53 AM
I sent a Email to my friend Frank at SIEG telling him it would be really good if he could arrange to have one of the new CNC mills shipped to me as a Christmas gift, haven't heard back from him yet.

Sunkenmetal
11-09-2007, 04:13 PM
what a POS just save your money and get a HAAS TM1.... You dont have to tram the head on those.....

Peter N
11-09-2007, 04:20 PM
There again the Haas TM-1 costs about 10 times as much, and is probably a little bit on the big side for an HSM.

Peter

Sunkenmetal
11-09-2007, 04:25 PM
There again the Haas TM-1 costs about 10 times as much, and is probably a little bit on the big side for an HSM.

Peter

A little big, its no bigger then a crappy knee mill, you get what you pay for.

John Stevenson
11-09-2007, 06:17 PM
what a POS just save your money and get a HAAS TM1.... You dont have to tram the head on those.....

I think you are on the wrong forum Sunshine, anyone recomending Haas TM1's and Mastercam XR2 isn't a home shop guy.

.

KiloBravo
11-09-2007, 07:12 PM
I am surprised nobody mention these guys.

http://www.industrialhobbies.com/

It's a bigger version of the RF45 with custom CNC.

I think they run around $6k.

The CNC kit alone is $4.5K.

oldtiffie
11-09-2007, 07:31 PM
I am surprised nobody mention these guys.

http://www.industrialhobbies.com/

It's a bigger version of the RF45 with custom CNC.

I think they run around $6k.

The CNC kit alone is $4.5K.

Hi kben77.

Looks good - very good.

I've got the HF(here)-45 and its staying as it is as I intend to go for a new Seig etc. CNC mill as a "turn-key" package.

The smaller CNC suits me.

I am in Australia so support for USA product may be problematical.

oldtiffie
11-13-2007, 05:44 AM
Hi John.

I would appreciate your advice re. my previous interst in the Seig x 4 CNC as regards a post I made a post on/at:
http://bbs.homeshopmachinist.net/showthread.php?t=26316&page=6



Evan,
Here is my form tool, probably took a little longer to make than yours though.:D

http://i109.photobucket.com/albums/n48/S_J_H/cnc%20bench%20lathe/cnclathe071.jpg

Hi Steve.

That sure is one marvelous machine. You must be very proud and satisfied - as you deserve to be.

I was very intrigued with the profile (hemisphere) and threading videos - I missed seeing them previously.

Was your tool path determined by modeling the profile of the cutter into/onto the model of the work?

I was very intrigued with the thread cutting as there is no physical connection (ie gear-train) between the spindle and the "X" and "Y" feed-screws.

I guess that there must a monitoring of the positions of the spindle and the lead-screw with respect to each other in the software.

How is that achieved?

I noticed too that the screwing tool was cutting on the left flank/side and that there was no need that I could see to "dress-up" the trailing/right thread groove flank/side either.

Am I correct in assuming that the tool was tracking at 30 degrees left to follow the right thread flank?

I could see the CNC motors on the "X" and "Y" lead-screws but I could not see how the spindle was driven other than the main motor as regarding getting and keeping the spindle and screws "in synch" with each other so very tightly under load at those speeds.

The reason I ask is that as you may be aware from my posts on John Stevenson's thread re. the new "Sig" x 4/4 vertical mill that I am interested in buying one. I was wondering and would require that the vertical mill spindle could be synchronised with the "X", "Y" and "Spare" motors as you have achieved with your lathe spindle and ball-screws.

What I'd like to be able to do on the Seig is to have the spindle synchronised to the rotab so as to generate a true involute gear with a spindle-mounted hob rather than just cut it incrementally and so leave a series of "flats" or "'steps" on the mating/meshing tooth faces of the finished gears.

I have one other question.

Could you have used you CNC lathe and a single-pointed tool to make the profile that Evan made with his profile tool?

I realise that your lathe and probably software if different to the Seig mill and I hope John Stevenson reads this post and responds to it.

I don't mean to hi-jack your excellent post at all - I was enthralled with it.

Again many thanks for a marvelous insight into your CNC lathe that you made.

John Stevenson
11-13-2007, 06:43 AM
Tiffie,
A few replies to your last post without taking anything away from Steve's beautiful job and it is.

Steve uses the same software as the mill, Mach3 but he uses the turn version. When you download the software you get a choice of three applications to run, mill which runs mill and router as they are virtually the same, Lathe and Plasma which runs plasma cutters, Oxy-fuel and laser cutters as they have a different type of non contact head.

They all come bundled in the same package.

Threaded is achieved by having a simple shaft encoder on the spindle, that simple that it is just a slot in a disk that is read every revolution. many people say that you need a multi-line encoder to be accurate but the truth is the simple one works - full stop.

That syncs the spindle to the leadscrew for screwcutting. The screwcutting operation can be programmed to approach at an angle just as you would if you were manually machining it or you can plunge cut if needed, the choice is yours, same applies to spring passes at the end to take the spring out, choose as many as you want.

The spindle sync operation whilst fine for threading won't at the moment handle a shaft encoder of enough accuracy to allow the spindle to be locked to a rotab.
This is a work in progress and it will happen but further down the line.
Quite a few people are asking for this as it's at the core of tool changing when you have driver dogs on the spindle tooling, it needs to know where they are for loading and unloading.

It is possible to do what you want at the moment and even do it on manual machines.

Take a read of this text.

http://homepage.ntlworld.com/stevenson.engineers/lsteve/hidden/Gearhobbing.txt

And the pictures are at

http://homepage.ntlworld.com/stevenson.engineers/lsteve/hidden/hob%20indexer1.jpg

And up to 26, just keep changing the number at the end after indexer.

The electronic bit which isn't too bad, only about 3 chips a few resistors and some thumb wheels for setting the gear count was covered in MEW 108 by Brian Thompson, it's creator.

I will leave Steve to explain how his machine cuts profiles.....

.

S_J_H
11-13-2007, 08:42 AM
Tiffie, John covered most of your questions real well. One thing I quickly learned about cnc threading is that you need plenty of spindle power and as consistent a motor speed as possible. While the spindle speed is fed into MACH with a single pulse per rev proximity sensor and Mach will adjust the feed rate to match, any gross rpm variance while cutting ruins the thread as by the time mach tries to compensate the damage is already done. This is why there is now a 1hp 3ph motor on this little machine and it now cuts beautiful threads.
Also I have an external 10 light/dark strip photo sensor(tachulator) to be sure how well mach is reading the single pulse proximity sensor. Up to 4000rpm they both read within 1rpm of each other. Also keeps me from making a silly mistake like having the wrong pulley ratio punched into mach.
The spindle speed is controlled by Mach using a small addon speed control board . It is from cnc4pc.com and only cost about 30$. It connects the computer(mach3) to the VFD drive. This is a slick setup as the machine can run with CSS(Constant surface speed). You punch in say 200sfm and then do a facing or turning cut. Mach will automatically increase the spindle speed to keep the SFM constant.
In that threading video I used a normal 29.5' infeed angle and 1 final spring pass. What is also not so apparent is the end of the threads was also cut with a 45' chamfer. All of this is super easy to setup in Mach.
The bushing Evan made could be cut very easily and quickly in steel with a single point tool. That is a piece of cake for a cnc lathe.
Here is a mt3 taper and a little radius cut. These early tests were done back when the machine had the little dc motor. It cuts a whole lot better now with the bigger motor and vfd.
http://i109.photobucket.com/albums/n48/S_J_H/cnc%20bench%20lathe/cnclathe054.jpg
Here is a shot of Mach3's simple threading wizard- There are other more complex threading wizards as well and just like Mach mill, many wizards for tapers,radius etc.http://i109.photobucket.com/albums/n48/S_J_H/cnc%20bench%20lathe/mach3threadingwizard.jpg
Steve

oldtiffie
11-14-2007, 04:39 AM
For: John Stevenson and S_J_H (Steve).

Many sincere thanks for the time thought and effort that you both put in your replies to my separate queries.

I will try to respond to both on the same/this and any future posts on this thread.

John, my apologies re. the duplicate posts on concurrent threads but it was all that I could think (of) to do. Just as well as it turned out as the thread got closed for no reason that I could see.

First of all, most of my questions have been answered or addressed in both replies (from John ands Steve).

Thanks for the text file and the pics (26) John as they were great.

(Just as an aside, that "Victoria" mill, if it is the same as I used at one time (50 years ago), is a great machine - light, sturdy, accurate and reliable. There is no mistaking the over-arm hanger bearing support for the arbor journal - and the thumb-operated oil-pump was a dead give-away.

I had sorted out how the hobbing process worked - for spur/straight, helical and worm-wheel - to generate the gears required. I had most of it from memory (50+ years ago!!), a book I have, the previous post you put up in the last couple of months and the text file that you provided the links too.

This is gear-generation in its truest sense as opposed to gear-cutting which in all its forms is at best a compromise - good and not so good.

The cutting I did all those years ago had the horizontal spindle (arbor) linked to the universal dividing-head with gearing so that the correct ratio of revs arbor:dividing head spindle matched the gearing ratio required for the gear to be produced. For spur gears the diving head spindle was parallel to the "X" feed for a plain hob and off-set by the hob helix angle at the pitch circle for spur gears cut with a spiral hob. On a universal mill the off-set angle was set on the mill table (ie the "X" feed).

For spiral/helical/worm-wheel gear production the table was off-set to suit the difference between the helix angles of the hob and the gears.

There were occasions when the differential dividing-head was driven from the "X" lead-screw via rod-and-gear linkages as well.

Needless to say this required intense concentration and any slip anywhere was a disaster.

But when it "worked" the gears produced were superb in terms of size, profile, surface finish and accuracy.

The only thing that could surpass it was the gear shaper/shaver machines (and the gear-grinders - which as I recall - were contracted out).

Steve's lathe and the synchronising of the spindle and "X" and "Y" feeds got exactly the same but better result in a beautifully elegant way under CNC control.

I recalled the set-up on John's mill - as per the text and pics on the links - but could not see how it could be done (as Steve did) without a mechanical linkage.

I was concerned about the speed control and accuracy of the spindle to concurrent "X" and "Y" on Steve's lathe both under light and heavy load in case of "slip" and/or "lag" somewhere.

I had concerns about the Seig mill having enough power to achieve the load and also just what the effective speed range is at which adequate torque is available and accurate consistent angular synchronisation is achievable.

The Seig, like my HF-45 is a vertical mill with an MT3 spindle held in by a draw-bar.

My arbor adaptor is as rigid and as accurate as I can get. But I anticipate having to make a fixture similar in design and principle to the "Victoria" mill arbor support to be fitted to the Seig vertical column vee-way/slide. This is intended to reduce the "flex" and chatter that might be caused by the cantilever situation that the hob on the arbor would otherwise operate under.
http://i200.photobucket.com/albums/aa294/oldtiffie/Stub_arbor1.jpg

http://i200.photobucket.com/albums/aa294/oldtiffie/Stub_arbor2.jpg

I intend to use a spiral hob with the rotab on-end and "angled" to suit on the table.

The operation of course will be the gradual generation of the gear as the hob is fed into the blank.

I was very interested in a comment John made in a previous post/thread that he had problems cutting a 127 tooth gear as it was "thin" and caused problems under cutting load. I took that point as it made a lot of sense to me. I guess that I'd have to support the blank on two "rings" a bit less than the gear full depth diameter.

I was very interested in the pics of Steve's computer screen as well as it answered a lot of questions. I was very intrigued at cutting the thread at 500 RPM and not needing to either disengage (or re-engage) the "lead-screw" or needing to have a "run-off/out" grove either and the way that the tool withdrew under cutting load.

I intend to mount a "tool post" on a fitting mounted on the Seig vertical column "Vee" way so that I can use the mill as a lathe. I intend to have the "lathe" as a mini-lathe head-stock mounted on the mill table not too unlike Steve's lathe set-up. I was concerned about threading and the accuracy of taper turning but Steve's post provided very elegant answers to that as well.

John, you said that there are 3 packages in the Seig "Mach" software. Is is possible to "open and use" both the "mill" and "lathe" packages (I am quite prepared to pay for it) so that I can use the Seig mill as a modified lathe as well?

Have you any indication as to what and when the power/torque and synchronising of the Seig spindle will be.

I note that Steve said he had to use 3 phase power and a VSD where-as the Seig as I understand it is single phase. Is that so?

I apologise to all for this very long post but it was necessary to try to get it all together.

As an aside.

I can get my memory to operate very well over say anything between 10 and 50-60 years ago, not so good but quite passable for say a month to 10 years ago, but anything less than a month can be a problem. I still can't remember what I am crossing a room or going somewhere for. Under the shower I even forget if I've used the soap (so do it twice) or whether I've washed the leg I'm not washing - so "do" that again as well - just in case.

Hmmmmmmmmmmmmm.

John Stevenson
11-14-2007, 08:09 AM
John, you said that there are 3 packages in the Seig "Mach" software. Is is possible to "open and use" both the "mill" and "lathe" packages (I am quite prepared to pay for it) so that I can use the Seig mill as a modified lathe as well?

Have you any indication as to what and when the power/torque and synchronising of the Seig spindle will be.

I note that Steve said he had to use 3 phase power and a VSD where-as the Seig as I understand it is single phase. Is that so?


Tiffie,
The Mach package comes all in at one price of $159
BUT you can go to the web page and download the demo program for free.
http://www.machsupport.com

This allows you to use all the mill functions but will stop running after 500 lines of code.

The turn side is more limited as turn programs tend to be a lot shorter and the screwcutting is disabled in the lathe demo.

This demo will not time out and only requires a licence key to work unlimited.

The limit of syncronising the spindle will need to be addressed by both Mach 3 software and external boards, people are working on both at the moment, as regrads a time scale I can't answer that.

Although Steve is using a 3 phase motor he's supplying it single phase via a VFD.

The Sieg uses a 3 phase brushless DC motor with feed back, much like a Vector drive VFD and again this is fed single phase.

.

S_J_H
11-14-2007, 08:34 AM
I note that Steve said he had to use 3 phase power and a VSD where-as the Seig as I understand it is single phase. Is that so?


I did not have to use 3ph and the vfd drive. I had that setup prior on my 9x20 and decided it would be much better utilized on the cnc lathe.
The c6 speed control board will also work with the x3 Sieg DC controller and it also will control my Minarik brand dc drive.
At the bottom of this page is wiring for the seig x3 and how it works. I don't know about the SX3 drive though. I know Hoss from cnczone is using the c6 board with his Sieg "freak x2" though.
Here is a link to the c6 board.
http://www.cnc4pc.com/Store/osc/product_info.php?cPath=25&products_id=58
I could show how to set it up in Mach3 if needed.
Steve

lazlo
11-14-2007, 08:49 AM
I intend to use a spiral hob with the rotab on-end and "angled" to suit on the table.

If you're using a dividing head to hold the gear blank, then you're gear cutting, not hobbing, even if you're using a hob to do the cutting. John posted a good Youtube video last week showing a guy cutting gears with a hob on his CNC lathe.


I recalled the set-up on John's mill - as per the text and pics on the links - but could not see how it could be done (as Steve did) without a mechanical linkage.

Tiffie, the way you've written this, I can't tell if that's a question? John's setup is an electronic gear hobber: there are two rotating spindles: the spindle of the mill, and the spindle of the powered workhead driving the gear blank. The number of teeth that are cut on the gear is determined by the ratio of the RPM of the cutting spindle (the mill) versus the RPM of the workhead driving the gear blank.

On a commerical gear hobber, or on the Jacobs gear hobbing kit, the two spindles are synchronized mechanically, via a gear train. On John's setup, the spindles are syncrhonized electronically: a shaft encoder on the mill generates a pulse train that is divided to get the correct RPM for the workhead to generate the number of gear teeth desired, and then used to drive a Gecko stepper controller driving the workhead.


I was concerned about the speed control and accuracy of the spindle to concurrent "X" and "Y" on Steve's lathe both under light and heavy load in case of "slip" and/or "lag" somewhere.

John and I have discussed this issue several times by PM, especially with respect to how much horsepower is needed to drive the workhead holding the gear blank. Also, with a feed-forward stepper control, there's no way for John's setup to deal with slip or lag between the spindles, which is directly related to the horsepower requirements: i.e., make sure the workhead motor driving the gear blank is big enough that it can't slip.

I've just about finished collecting all the pieces I need to build a servo version of Sir John's electronic hobber. A servo setup would have the advantage that it has feedback, and can deal with slip, lag, and starting and stopping the spindle between cuts.

Good luck! :)

Robert

John Stevenson
11-14-2007, 04:38 PM
I did not have to use 3ph and the vfd drive. I had that setup prior on my 9x20 and decided it would be much better utilized on the cnc lathe.
The c6 speed control board will also work with the x3 Sieg DC controller and it also will control my Minarik brand dc drive.
At the bottom of this page is wiring for the seig x3 and how it works. I don't know about the SX3 drive though. I know Hoss from cnczone is using the c6 board with his Sieg "freak x2" though.
Here is a link to the c6 board.
http://www.cnc4pc.com/Store/osc/product_info.php?cPath=25&products_id=58
I could show how to set it up in Mach3 if needed.
Steve
Steve,
Here's a shot of the KX3 / SX3 board and motor.

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/SX3%20board.jpg

The board far left is the brushless drive board with a small 0 - 10v analog board plugged in that the computer talks to when fitted to the KX3

The two boards to the right of the red screwdriver replace the analog board when fitted to an SX3, One is the start stop and speed control that goes on the front of the SX3 head and the smaller board is the spindle speed DRO.

The brushless motor is far right with a 6" rule laid on this, these are 240 volts , 1,000W.

Three phase wires going in and 5v + & - plus 3 feedback cables out.

Same layout for the KX1 but the motor on this one is only 400W.

.

sconisbee
11-14-2007, 05:33 PM
you know John if you need an extra tester for these machines you can always send one to me, ill look after it! :D .....ok maybe not lol, but they do look good, and its nice to see some out of the box solutions for the smaller hobbiests and i suppose even the smaller businesses....i for one would love to have a CNC x3 some of the stuff i produce for boats gets kinda complicated and repetative.

JRouche
11-14-2007, 09:26 PM
By the way, here's a video of the Syil X3 CNC. 26 IPM with .1 DOC in aluminum. Bizzare toolpath -- I don't know what CAM he was using:

[url]http://www.youtube.com/watch?v=yv3-DwEHh5o

LOL.. That IS a bizarre toolpath. Looked like a kid on a spirograph wrote it. Oh? I know now!! It was written in Chinese. Haaa. Lot of wasted paths there.. JRouche

S_J_H
11-14-2007, 10:47 PM
Thanks for sharing the sx3 motor and drive setup John. Seems quite a bit different than my x3. It seems Sieg is really improving their product line.
On my list of to-do's is a big overhaul of my x3 cnc setup.

lol, I just watched that tool path video. Very odd and it won't make many want to run out and buy or build a cnc mill. But I guess it was just some random circular cutting action.

Steve

oldtiffie
11-15-2007, 01:40 AM
Deleted - incorrectly posted before completion.

oldtiffie
11-15-2007, 05:35 AM
If you're using a dividing head to hold the gear blank, then you're gear cutting, not hobbing, even if you're using a hob to do the cutting. John posted a good Youtube video last week showing a guy cutting gears with a hob on his CNC lathe.

Tiffie, the way you've written this, I can't tell if that's a question? John's setup is an electronic gear hobber: there are two rotating spindles: the spindle of the mill, and the spindle of the powered workhead driving the gear blank. The number of teeth that are cut on the gear is determined by the ratio of the RPM of the cutting spindle (the mill) versus the RPM of the workhead driving the gear blank.

On a commerical gear hobber, or on the Jacobs gear hobbing kit, the two spindles are synchronized mechanically, via a gear train. On John's setup, the spindles are syncrhonized electronically: a shaft encoder on the mill generates a pulse train that is divided to get the correct RPM for the workhead to generate the number of gear teeth desired, and then used to drive a Gecko stepper controller driving the workhead.

John and I have discussed this issue several times by PM, especially with respect to how much horsepower is needed to drive the workhead holding the gear blank. Also, with a feed-forward stepper control, there's no way for John's setup to deal with slip or lag between the spindles, which is directly related to the horsepower requirements: i.e., make sure the workhead motor driving the gear blank is big enough that it can't slip.

I've just about finished collecting all the pieces I need to build a servo version of Sir John's electronic hobber. A servo setup would have the advantage that it has feedback, and can deal with slip, lag, and starting and stopping the spindle between cuts.

Good luck! :)

Robert

Hi lazlo.

Thanks for the reply.

First of all, I intend not to get into another slanging/quoting/"citing" war.

I will respond as briefly as I can to your post - but the quotes are unavoidable.

I will try (again) to detail what my requirements are and how I see them being met with the "Seig" 4X vertical CNC mill that John S posted that after all, is the subject of this thread.



Originally Posted by lazlo
If you're using a dividing head to hold the gear blank, then you're gear cutting, not hobbing, even if you're using a hob to do the cutting. John posted a good Youtube video last week showing a guy cutting gears with a hob on his CNC lathe.

I knew and know that. There is no argument from me on that. It is not what I intend to do which is gear-hobbing.



Originally Posted by lazlo
Tiffie, the way you've written this, I can't tell if that's a question? John's setup is an electronic gear hobber: there are two rotating spindles: the spindle of the mill, and the spindle of the powered workhead driving the gear blank. The number of teeth that are cut on the gear is determined by the ratio of the RPM of the cutting spindle (the mill) versus the RPM of the workhead driving the gear blank.

Same reply as above except that this is gear-hobbing which I intend to do.


Originally Posted by lazlo
John and I have discussed this issue several times by PM, especially with respect to how much horsepower is needed to drive the workhead holding the gear blank. Also, with a feed-forward stepper control, there's no way for John's setup to deal with slip or lag between the spindles, which is directly related to the horsepower requirements: i.e., make sure the workhead motor driving the gear blank is big enough that it can't slip.

I am neither impressed nor deterred by any private discussions between anybody else as that is - private. I can only consider and comment on the forum only what is posted on the forum.

Steve seems to have achieved a successful outcome in this regard - even if others have not to date.

John seems to indicate that Seig has something in the pipe-line to achieve a similar result to Steve.

That being the case, so long as the outcome is as I require, my interest ends there and then.

I am not interested in how it came about other that to say that Steve's results are self-evident to me and meet my need unless or until they are proven to be otherwise.

I thoroughly congratulate Steve on his achievements as in my view they are outstanding and should be regarded and acknowledged as such.

The work-head motor - in my requirement - will be driving a 90:1 "Vertex" rotab so I guess that the motor requirement will not be too large at all. I shouldn't imagine that it will be significantly larger - if at all - than the motor Steve used to drive his "X" lead-screw.

I am confident that the "Seig" main (vertical) cutter spindle has the power to drive the cutter/hobbing cutter/end-mill etc. that I might reasonably require it to do.

I am quite open to reasoned advice as to why these items may or will not be the case.

Steve seems to have solved the "lead-lag/stepper/servo" conundrum quite elegantly and I guess that level of "tracking accuracy" will suffice for my purposes on the rotab - unless or until I am advised otherwise. His "hemisphere" and "morse taper" were excellent examples as well and served to increase my confidence in his underlying concept and his application and practical demonstration of it.

Evan's advice/reply on this sort of matter was very well put in regard to his CNC mill, "stepping" and similar levels of accuracy on a recent "Gear(ing)" thread.



Originally Posted by lazlo
I've just about finished collecting all the pieces I need to build a servo version of Sir John's electronic hobber. A servo setup would have the advantage that it has feedback, and can deal with slip, lag, and starting and stopping the spindle between cuts.

I have no problems with a servo alternative to a stepper motor control. And until it is proven to work it is a "work in progress" where-as Steve's solution is "out there" for all to see and consider as a working practical solution.

How or why he did it is Steve's business - but do it he did!!.

It is not my business nor am I interested in those "under the hood" developments and applications as they are outside my knowledge. I am only interested to see if and that they "work".

The "technical bits" can be discussed and debated among the "Techies" who live in that higher ethereal plane and occasionally come down on that cloud that lesser mortals like myself are under.

I am a "Taxi driver" not a "mechanic/guru/technologist etc" whose interest is "under the hood".

I just need to drive my Taxi and get competent assistance if and when needed - and pay the "going rate" for it.

So far John's Seig X4 and Steve's synchronising solution are at or well on the way to meeting my driver requirements for my Taxi.

Now onto my requirements.

My requirement is that that the mill spindle and the rotab on which the job is mounted be synchronised to a level of accuracy as achieved by Steve between the lathe spindle and the "X" and "Y" CNC drives. I could see no cumulative or "spot" errors in his screw thread - at all. So he had the synchronisation correct to a high order of accuracy so far as I could tell.

I would also wish that if I chose to mount a lathe head on the mill table that I could synchronise that spindle motor with the mill "X" and "Y" CNC motors (using a tool mounted on a fixture which in turn was fitted to the mill vertical column vee-way/dove-tail). I would be simulating/emulating Steve's lathe.

If I need to buy the "Mill" and "Lathe" software - so be it.

If I need John or Steve to make or supply parts and/or modifications and/or advice in that regard for the "Seig" I will do it and gladly pay what they ask as I will be getting an excellent service and not just buying a "product".

As I understand it, the Sieg mill is a 4X unit - or more particularly a 3X with a "spare" CNC control available for use.

My requirement - as I thought I'd made clear - is for the spindle to be synchronised with the rotab spindle with the 4th. controller to be mounted on the rotab spindle such that the rotab will rotate in any given direction at any given ratio between the mill spindle and (rotation of) the rotab table.

My requirement is that when the mill spindle upon which the hobbing cutter is mounted is started the "X" feed can be engaged/started so that the rotating cutter will progress along the job as it is being rotated by the rotab.

The hobbing cutter and the job will have the same pitch (circular for the job and linear for the hobbing cutter which will have the form of a rack with the required pressure angle. ie the DP (inch/USA) or MOD (metric - in my case).

I do not anticipate cutting any gear larger than 6 inches OD and with a MOD no larger than MOD1.0 or 1.5. The largest gear on my lathe is my 127 gear which has a PA of 20deg and is MOD1.0 (OD = 129mm). Space considerations of the Seig and the rotab etc will be constraints as well.

Any other gear will be cut by a hob in the normal "gear-cutting" (ie "non-hobbing") manner using either hobbing cutter or a "numbered" (ie "traditional") commercially-available gear cutter. I may well "send it out" and have it/them made by a local Specialist Gear-maker who has a superb set-up for generating them.

All of this pre-supposes or presumes that I will buy the "Seig" X4 - but despite that it is looking very good thus far - I am not I am not committed to anything - yet.

I hope this discussion is at or nearing its end as I have other matters which I want to address to John and which others are cordially invited to comment and advise upon - on the forum.

Thank you.

lazlo
11-15-2007, 09:00 AM
First of all, I intend not to get into another slanging/quoting/"citing" war.

Tiffie, you need to calm down. You completely mis-read my post.


Same reply as above except that this is gear-hobbing which I intend to do.

Tiffie, I'm not being pedantic -- if you're using a Dividing Head or a Rotab, you're gear cutting, and not gear hobbing. John's setup is a gear hobber.

So this video that John posted last week, where the guy is using a dividing head and cutting the gear teeth on his CNC lathe with a hob is gear cutting (not hobbing), and therefore uses a very different setup than John built:

http://www.youtube.com/watch?v=MItgd-faHFw


I am neither impressed nor deterred by any private discussions between anybody else as that is - private.

I have no interest in impressing you Tiffie. I was just adding an important piece of information that is not in John's description of the "Stevo Ee-lek-Tronic Gear Hobber" -- if you use a purely feed-forward control system, like a stepper, you need a big enough motor on the workhead that's driving the gear blank to ensure that the blank doesn't slip, or the gear will be ruined. John used a 1/4 HP motor to drive his workhead/gear blank!


The work-head motor - in my requirement - will be driving a 90:1 "Vertex" rotab so I guess that the motor requirement will not be too large at all. I shouldn't imagine that it will be significantly larger - if at all - than the motor Steve used to drive his "X" lead-screw.

That's the major point that you're missing Tiffie. A dividing head or a rotab is not going to work as a gear hobber, but it will work great as a gear cutter, as in the Youtube video I linked above. The workhead in John's gear hobber has precision high-speed spindle bearings, and is running at relatively high RPM, which a dividing head or rotab can't do.

But a rotab or dividing head can obvious index the gear blank between cuts, if you're doing gear cutting.

Make sense?

So John's "Ee-lek-Tronic Hobber" isn't CNC, and doesn't need CNC.

So a Sieg CNC setup isn't going to help you to hob gears like Sir John is doing, but it would make an excellent gear cutting setup, like the YouTube video above.

Cheers,

Robert

John Stevenson
11-15-2007, 10:32 AM
Tiffie, me old mate,

Either due to my description or your interpretation of it some confusion has crept in..

Take this quote.

My requirement is that that the mill spindle and the rotab on which the job is mounted be synchronised to a level of accuracy as achieved by Steve between the lathe spindle and the "X" and "Y" CNC drives. I could see no cumulative or "spot" errors in his screw thread - at all. So he had the synchronisation correct to a high order of accuracy so far as I could tell.

I would also wish that if I chose to mount a lathe head on the mill table that I could synchronise that spindle motor with the mill "X" and "Y" CNC motors (using a tool mounted on a fixture which in turn was fitted to the mill vertical column vee-way/dove-tail). I would be simulating/emulating Steve's lathe.

Now the problem is that the method of synchronisation used by Steve to do threading on the lathe isn't accurate enough to allow a rotab to follow the spindle to do hobbing.
Steve's spindle is only updated every revolution whereas in hobbing it has to be updated all the time.

Until a further version of Mach is released and a card is made to enable a high count encoder to work with Mach then it's a dead duck [ quack ]
My encoder runs at 4,000 counts per spindle rev, Steve's works at 1 and that's the main difference.

For low speed work I can't see why a rotab can't do the driving but the limit will be on low speed and minimum tool forces.

Hopefully I will try this shortly as I'm in the process of converting a small 4" tilting table with a tiny type 17 stepper motor to run with the KX1 CNC.
This is to try an idea out to see if I can cut a perfect bevel gear with full tapered form and no touching up as you have to do at the moment.

Did I mention it will use a 20 pence tool to do it with ? :D
Today I ordered 4 pounds of machinable wax to work with.

If it works it will be the subject of another post, if it doesn't this post will be edited :D

.

lazlo
11-15-2007, 10:50 AM
Thanks John -- I had a feeling Tiffie wasn't to listen to any explanation that I posted ;)


For low speed work I can't see why a rotab can't do the driving but the limit will be on low speed and minimum tool forces.

But to reiterate my previous post, although a rotab doesn't have the bearings to spin fast enough to hob, it works fine to index a gear cutter, and it still a great way to make gears.

You just won't get the theoretically perfect involute tooth form you get with your electronic gear hobber, which is generated by the gear blank continuously rotating around the circular pitch past the hob.

oldtiffie
11-15-2007, 07:31 PM
Tiffie, me old mate,

Either due to my description or your interpretation of it some confusion has crept in..

Take this quote.


Now the problem is that the method of synchronisation used by Steve to do threading on the lathe isn't accurate enough to allow a rotab to follow the spindle to do hobbing.
Steve's spindle is only updated every revolution whereas in hobbing it has to be updated all the time.

Until a further version of Mach is released and a card is made to enable a high count encoder to work with Mach then it's a dead duck [ quack ]
My encoder runs at 4,000 counts per spindle rev, Steve's works at 1 and that's the main difference.

For low speed work I can't see why a rotab can't do the driving but the limit will be on low speed and minimum tool forces.

Hopefully I will try this shortly as I'm in the process of converting a small 4" tilting table with a tiny type 17 stepper motor to run with the KX1 CNC.
This is to try an idea out to see if I can cut a perfect bevel gear with full tapered form and no touching up as you have to do at the moment.

Did I mention it will use a 20 pence tool to do it with ? :D
Today I ordered 4 pounds of machinable wax to work with.

If it works it will be the subject of another post, if it doesn't this post will be edited :D

.

Thanks John.

That is precisely the sort of response I wanted and need - concise, focused, straight to the point, all the info I require, no bull**it and at a level that I can relate to directly.

I am more comfortable and more certain that I will be "going" for the Seig X4 which is the subject of this thread.

Knowing that a solution and answer to my requirement is "pending" or "in the pipeline" from Seig/Mach is just about what I require. In short, that it is "do-able" and will be done "in due course".

If that requires getting you or someone similar (IS there one??) to supply such a solution on a "third party"/"add-on" basis - then that's what it will be.

I am comfortable about the rotab synchronisation bit. I will await your advice on this matter. I don't anticipate any significant loads or speeds (on, of, and at the rotab) - at all. If you determine what reasonable loads limits are, then I will use and apply them.

I would still want to emulate Steve's lathe on the Seig as previously stated.

If I have to stick to cutting threads on my current lathe - then so be it - but with luck, the rest of it should be do-able.

Is it possible to use the 4th. "port??/channel" on the Seig to drive what-ever motor is to drive the rotab and the "lathe head-stock" on the mill?

Is it possible to buy both the "mill" and "lathe" versions from the same package of the Mach software and have them both on the same computer?

Can I have one copy of the "Mill" and "Lathe" files on my computer in the shop and another on the computers in the house and can I transfer files between the computers in the shop and the house or do I need additional licences? I ask this as many/most software developers/distributors allow three copies at the one address. Do I need to net-work the computer in the shop/shed? I have 2 computers in the house (W2K and XP) on an Ethernet net-work and they work fine.

The Seig set-up and your "follow-up" info as well as the excellent jobs on Steve's lathe and Evan's CNC mill and their information and discussion have been hugely helpful.

The info I need is at the "man in the street" level so that I can see what is "about" and get the info I need in a form that is both useful and usable by me at the level I require.

Once I am satisfied that the "Taxi" is what I want/need, I will set about buying it so that I can "drive" it with the necessary "under the hood" back-up available if, when and as required.

I will be interested to see how you get on with the "machinable wax". I saw the recent thread on it and it was very interesting.

Also, as a matter of interest, I intend to use "Tufnol" (Cotton Fabric Based Phenolic) http://www.ceaco.com.au/tufnol.html for gear material unless there are compelling reasons for me to do otherwise. I can see it as a very viable alternative to cast-iron or steel. I used "Tufnol" many years ago and it was marvelous as it machined very well and required little or no coolant and it performed well under reasonable loads. It, or a similar product, was used for years in some car/automobile engines as the "drive" gear on the crank-shaft in the cam-shaft drive (with push-rods - before or not used on OHV set-ups). It went very well although it did wear a a bit and eventually the gears "crowned out" and stripped - after about 80 > 100,000 miles). Did not take kindly to over-revving or "thrashing" the engine.

Again, many thanks to yourself, Steve and Evan for getting me started and keeping me informed and showing what's "do-able" at my level and meeting my requirements in my HSM shop.

It's all looking very good from here.

I look forward to your info and this thread continuing.

S_J_H
11-15-2007, 11:03 PM
oldtiffie, when you buy a license to mach3 you get both mach mill and turn in the package. Plasma too but that's not my bag.
The mach3 following is massive now. There will be tons of people to help you out. Yes you can keep multiple copies on different computers. I keep mach on 3 different pc's. I do most of my cad/cam on a laptop so I can sit in my living room and relax and do that type of work. I have a dedicated pc in my home shop also loaded with mach to run the software for my lathe and mill. I also keep a version on my desktop where I surf the web just for downloads etc.
IMHO the price is well worth it for mach3. What is it now just over 150$? I spent more than that last week for dinner at a restaurant.
My x3 mill was easy to convert. The lathe was a lot more work because so much was scratch built. I will never pretend to be an expert at this stuff. I just read and try and learn from others and gladly try and return the favors in some fashion.
I watched that gear hobbing video series by Shorty on youtube. Tell you what, Ol' Shorty is what this hobby is all about. Is he a member on this site?
Steve

oldtiffie
11-16-2007, 02:54 AM
oldtiffie, when you buy a license to mach3 you get both mach mill and turn in the package. Plasma too but that's not my bag.
The mach3 following is massive now. There will be tons of people to help you out. Yes you can keep multiple copies on different computers. I keep mach on 3 different pc's. I do most of my cad/cam on a laptop so I can sit in my living room and relax and do that type of work. I have a dedicated pc in my home shop also loaded with mach to run the software for my lathe and mill. I also keep a version on my desktop where I surf the web just for downloads etc.
IMHO the price is well worth it for mach3. What is it now just over 150$? I spent more than that last week for dinner at a restaurant.
My x3 mill was easy to convert. The lathe was a lot more work because so much was scratch built. I will never pretend to be an expert at this stuff. I just read and try and learn from others and gladly try and return the favors in some fashion.
I watched that gear hobbing video series by Shorty on youtube. Tell you what, Ol' Shorty is what this hobby is all about. Is he a member on this site?
Steve

Thanks for the info Steve - it was "spot-on". Thanks for the interest and time.

First off, I anticipate that I might have to use a "Seig" (or equivalent) 1000W 3-phase motor - as shown in one of John's recent pics on this thread - on the Seig cutter spindle. But that's OK. Don't know why I think so - but time and continuing good advice all round will sort that out as well.

I had no idea what mach3 cost - never even thought to have a look. But even if it was US$500+ it would be well worth it. The question for me is more a "show cause" situation where instead of asking what it cost, asking for a "show cause" as to why I don't. It is a "piddling" amount for such a vital essential item in the overall scheme of things that I would have to "show cause" as to why I should not buy it. And I can't. So I buy it what-ever the cost. US$2K would not be out of the ball-park in that situation.

What CAD software do you use? What others are there that will do the job? Which is the best of them? Do/can they write the "G" (or what-ever) code to/for mach3 or are there other intermediate steps? Is a CAD/"Modelling" or CAD/CAM package required? If so, any suggestions?

Putting the CNC/electronics aside for the moment, the way John set up his mill in his pics and the excellence of craftsmanship in your CNC lathe and Evan's CNC mill and the explanations the three of you provided were fabulous in every way.

All three and others of like ilk are superb Machinists in every sense of the word.

"Shorty" in the u-Tube items are an excellent case in point - as you say. He "got the message over" in fine style such that the audience that he was addressing - what-ever their "qualification" or "experience" - or lack of it - just had to "get it".

In that respect I have to say that the skills of a Machinist do not require any formal education or "time served" or "experience" etc.

(Formal requirements as a condition of or for "Registration" etc. are the exception. Plumbers, Electricians, Engineers, Doctors, Veterinarians, Accountants etc. are in that category - there are many others).

Skill at these levels is a state of mind, attitude, dedication and a willingness to learn and "try" as these are what counts. Lane - possibly the best machinist on this forum - has made that very point numerous times.

The 3 of you have that in abundance.

I have seen any amount of people who have finished a course, Apprenticeship, Traineeship or Tertiary/Degree/Diploma courses that were not good machinists at all - they merely saw the "distance" out and got a "Pass" and were not "flash" at all in applying any knowledge that they might have acquired. I must qualify those remarks by saying that this is not necessarily the general case, but there are sufficient numbers to be of concern.

I have had the privilege to work with some extra-ordinarily gifted competent and versatile "Qualified" people in my life - even Lawyers, Accountants, Engineers, Plumbers, Electricians, Financial Advisers, Administators and the like - and others listed above - great great people. Even at $50, $75, $100, $300, $500 plus per hour was a great investment (I never saw them as "cost" items) as I invariably got much more back because of their services than I paid for them - some 10+ times over - and "set my wife and I up" where we are today.

There are some excellent Machinists on this forum irrespective of whether they have any formal "qualifications" at all. I am damn sure that if I were hiring I'd grab them based on their ability, versatility and ability to "think out-side the square" - just as you, John, Evan and some others have more than adequately demonstrated on this forum.

Being able to communicate as well as the three of you and others do is a huge bonus to me and I hope to others as well.

Back to the "Seig".

I anticipate it having an effective life of about 5 years (very rough "guess-timate" here) as that will get me to about 76 which is as far as I can reasonably expect to go in using this "stuff". After that? Who knows? Who cares? Not me!!

I thank you all.

oldtiffie
11-16-2007, 06:14 AM
Tiffie, me old mate,

Either due to my description or your interpretation of it some confusion has crept in..

Take this quote.


Now the problem is that the method of synchronisation used by Steve to do threading on the lathe isn't accurate enough to allow a rotab to follow the spindle to do hobbing.
Steve's spindle is only updated every revolution whereas in hobbing it has to be updated all the time.

Until a further version of Mach is released and a card is made to enable a high count encoder to work with Mach then it's a dead duck [ quack ]
My encoder runs at 4,000 counts per spindle rev, Steve's works at 1 and that's the main difference.

For low speed work I can't see why a rotab can't do the driving but the limit will be on low speed and minimum tool forces.

Hopefully I will try this shortly as I'm in the process of converting a small 4" tilting table with a tiny type 17 stepper motor to run with the KX1 CNC.
This is to try an idea out to see if I can cut a perfect bevel gear with full tapered form and no touching up as you have to do at the moment.

Did I mention it will use a 20 pence tool to do it with ? :D
Today I ordered 4 pounds of machinable wax to work with.

If it works it will be the subject of another post, if it doesn't this post will be edited :D

.

Thanks again John.

Great advice.

It just so happens that there is a concurrent thread that addresses the short-comings and limitations of the "1 check per rev" encoder?? (well, registration - for want of a better word for me) at:
http://bbs.homeshopmachinist.net/showthread.php?t=26365

It relates to a single "shot" system on a lathe such as Steve's - the system then used was (a?) "Frog" but it seemed very similar.

It cleared up and explained the need for high spindle RPM and the need for very accurate spindle speed control.

The original thread was at:
http://bbs.homeshopmachinist.net/showthread.php?t=3949&highlight=frog+cnc+thread

I thought the discussion was excellent from my perspective.

I can now better appreciate the need for your suggested 4,000 "samples" per rev of the main cutter spindle.

Am I correct in assuming that with the higher sampling rate that a slower work speed will be possible - and if so - over what range?

This is looking better by the hour.

Many thanks.

John Stevenson
11-16-2007, 07:44 AM
Tiffie,
Hobbing is usually done at between 120 and 300 rpm max so as to extend the life of the expensive cutters.

It's also dependent on numbers.

A low count gear say a 20 tooth will have to move 4 times faster than an 80 tooth so when you are cutting 20 teeth the software has to be able to handle the number of pulses.

It it can't handle these in the time available and we are talking micro seconds per pulse then you have to slow the spindle speed down to also slow the pulse train down.

The 4,000 figure was arrived at by trial and error, any greater than 4,000 and it can get egg bound and loose steps,

Any less and the operation becomes very jerky.
This is a ball park figure as we are not talking a dead 4,000 but in that area.;

.

lazlo
11-16-2007, 08:55 AM
Hobbing is usually done at between 120 and 300 rpm max so as to extend the life of the expensive cutters.
...
A low count gear say a 20 tooth will have to move 4 times faster than an 80 tooth so when you are cutting 20 teeth the software has to be able to handle the number of pulses.
It it can't handle these in the time available and we are talking micro seconds per pulse then you have to slow the spindle speed down to also slow the pulse train down.
The 4,000 figure was arrived at by trial and error, any greater than 4,000 and it can get egg bound and loose steps,
Any less and the operation becomes very jerky.

John, that's exactly what I explained to Tiffie 3 posts up, but he's not understanding the difference between the electronic gear hobber you built, and CNC gear cutting, as in the YouTube video.

Do you still have that video of your electronic hobber running? Maybe if he saw it running, he'd understand the difference between the two methods.

http://bbs.homeshopmachinist.net/showpost.php?p=316072&postcount=68

A dividing head or a rotab is not going to work as a gear hobber,

The workhead in John's gear hobber has precision high-speed spindle bearings, and is running at relatively high RPM, which a dividing head or rotab can't do.

http://bbs.homeshopmachinist.net/showpost.php?p=315884&postcount=61

John's setup is an electronic gear hobber: there are two rotating spindles: the spindle of the mill, and the spindle of the powered workhead driving the gear blank. The number of teeth that are cut on the gear is determined by the ratio of the RPM of the cutting spindle (the mill) versus the RPM of the workhead driving the gear blank.

On a commerical gear hobber, or on the Jacobs gear hobbing kit, the two spindles are synchronized mechanically, via a gear train. On John's setup, the spindles are syncrhonized electronically: a shaft encoder on the mill generates a pulse train that is divided to get the correct RPM for the workhead to generate the number of gear teeth desired, and then used to drive a Gecko stepper controller driving the workhead.

lazlo
11-16-2007, 09:00 AM
By the way John, I think Tiffie is saying, in a weird kind of way, that he doesn't want to know how the sausage is made, he just wants to know if he can make gears if he buys a Sieg CNC System.

So just say "Yes" and he'll be happy, and we can get back to the thread :D

BobWarfield
11-16-2007, 12:03 PM
Stepping away from sausages and hobbing (hmmm), I really like the thought that there will be a market for gcodes.

I fool around quite a lot with cnc, and have made my own site to chronicle some of it (www.cnccookbook.com). What I've found is that the mechanicals of converting a machine are dead easy. If you can build a little steam engine or nice piece of tooling, it's no challenge at all. You can even go first class with ballscrews and all that good stuff, or make a machine from scratch like the cool lathe we've seen pictured.

Next up are the electronics. That's still pretty easy. It's building blockish with breakout cards and Gecko drives or similar. I've used both Geckos and Xylotex and it all has worked great. Perhaps the most complicated thing is building a power supply for the motors, but there are easy formulas available or you can just by one of those too.

By far the hardest part is the g-codes. Writing them from scratch is pretty close to computer programming. I am a software developer, and I don't really care for it much.

Your next step up is the wizards in Mach. They're not bad. Think of them as giving you a really smart "power feed" capability for your machine, although in fairness they go beyond even that notion.

But to really get facile, you need a CAD program and a compatible CAM program. CAD makes the drawings, CAM converts them to gcodes. I haven't played with the low end too much. I know John S. likes Dolphin, and Mach even has a low end called LazyCAM.

Things get all weird when you get into CAM. These programs can get awesomely expensive. I use Rhino 3D and OneCNC, and I paid more for the pair than my IH Mill. They take a lot of getting used to in terms of how to think about it.

My recommendation would be to really look for a community college CNC class or hopefully a buddy that's all over it. I could definitely see a case for a club or small group of machinists to go in together and share the software. You only use it for a short part of the machining process and it costs a lot of money.

You see why I think the g-code idea is a good one? I almost think someone with CAD/CAM package could make a business generating g-code for drawings they're sent.

This whole CAD/CAM piece seems like really the dark side of the hobby so far. I don't think we quite have the cheap yet very powerful equivalent of Mach 3.

Cheers,

BW

PS Someone mentioned governments being afraid of this technology. Here is a site to think about. There used to be a site, cncgunsmithing.com, where the guy had published complete CAD and g-code for making firearms. We're talking semi-auto pistols and AR-15 lower receivers. I notice his site has disappeared. Not sure what happened, but once you have a complete CNC shop, you can make almost anything.

snocat_trf
11-16-2007, 03:05 PM
he changed his domain to: http://www.cncguns.com

oldtiffie
11-16-2007, 08:23 PM
Tiffie,
Hobbing is usually done at between 120 and 300 rpm max so as to extend the life of the expensive cutters.

It's also dependent on numbers.

A low count gear say a 20 tooth will have to move 4 times faster than an 80 tooth so when you are cutting 20 teeth the software has to be able to handle the number of pulses.

It it can't handle these in the time available and we are talking micro seconds per pulse then you have to slow the spindle speed down to also slow the pulse train down.

The 4,000 figure was arrived at by trial and error, any greater than 4,000 and it can get egg bound and loose steps,

Any less and the operation becomes very jerky.
This is a ball park figure as we are not talking a dead 4,000 but in that area.;

.

Thanks John.

I have no problem at all with that concept.

The key, as you say, is that the number of pulses in a given time is the main limiting factor and speeds of the - say - spindle and rotab (or "lathe head-stock") have to be adjusted accordingly to maintain a maximum of say 4,000 pulses per spindle rev. in a given time to maintain the required close synchronisation.

I have no problem with the speed limitations imposed by cutters, material and set-up rigidity.

I can foresee no need to make a hobbing cutter as I'd buy one in from McMasters or similar. It would need a 1" bore and keyed to suit a standard 1" arbor. I only require a MOD1 (metric) at present.

My preferred materials are: aluminium, "machinable wax", "Tufnol", steel and brass/bronze. Cast-iron as well I suppose.

Costs and cost-recovery are not such a consideration here. Everything I buy is "written off" as soon as I get it as I have no need to amortise nor depreciate them.

I am retired so "cost-effectiveness" is not a consideration - as long as I've got the cash and the need and the item required is available - I buy it. I almost invariably go for the "best there is for the price" as this is a hobby.

I have the resources to sharpen and maintain tools that need it.

I have no need to "push" my equipment either. If I take a month for a "commercial one-hour job" - who cares, not me - it just does not matter to me.

So as I see it, as soon as the "4,000 pulses per rev. encode" problem is resolved, my requirements are met. As that is in the "pipe-line" either within mach3 (or 4?) and "Seig" or by or from a "Third Party" I have every reason to be very optimistic and to keep appraising the "Seig" X4 which is looking better by the day.

I would hope that resolving the "4,000" matter that my intended use of the Seig vertical CNC mill as a lathe (and perhaps as a cylindrical grinder) - see previous posts on this thread -will be realised sooner rather than later.

Can you recommend a suitable CAD/CAM package please?

I intend to stick with the "Turn-key" aquisition policy as it suits me as I've seen no good reason for me to do otherwise.

Others have every right to think (I should do?) otherwise.

If I am seen to be using my "Sieg" "off-the-shelf set-up" "Taxi" (see previous posts) as a "Sausage machine" - so be it.

And if it is?

Likewise.

Tough.

Again, many thanks.

oldtiffie
11-16-2007, 10:00 PM
Stepping away from sausages and hobbing (hmmm), I really like the thought that there will be a market for gcodes.

I fool around quite a lot with cnc, and have made my own site to chronicle some of it (www.cnccookbook.com). What I've found is that the mechanicals of converting a machine are dead easy. If you can build a little steam engine or nice piece of tooling, it's no challenge at all. You can even go first class with ballscrews and all that good stuff, or make a machine from scratch like the cool lathe we've seen pictured.

Next up are the electronics. That's still pretty easy. It's building blockish with breakout cards and Gecko drives or similar. I've used both Geckos and Xylotex and it all has worked great. Perhaps the most complicated thing is building a power supply for the motors, but there are easy formulas available or you can just by one of those too.

By far the hardest part is the g-codes. Writing them from scratch is pretty close to computer programming. I am a software developer, and I don't really care for it much.

Your next step up is the wizards in Mach. They're not bad. Think of them as giving you a really smart "power feed" capability for your machine, although in fairness they go beyond even that notion.

But to really get facile, you need a CAD program and a compatible CAM program. CAD makes the drawings, CAM converts them to gcodes. I haven't played with the low end too much. I know John S. likes Dolphin, and Mach even has a low end called LazyCAM.

Things get all weird when you get into CAM. These programs can get awesomely expensive. I use Rhino 3D and OneCNC, and I paid more for the pair than my IH Mill. They take a lot of getting used to in terms of how to think about it.

My recommendation would be to really look for a community college CNC class or hopefully a buddy that's all over it. I could definitely see a case for a club or small group of machinists to go in together and share the software. You only use it for a short part of the machining process and it costs a lot of money.

You see why I think the g-code idea is a good one? I almost think someone with CAD/CAM package could make a business generating g-code for drawings they're sent.

This whole CAD/CAM piece seems like really the dark side of the hobby so far. I don't think we quite have the cheap yet very powerful equivalent of Mach 3.

Cheers,

BW

PS Someone mentioned governments being afraid of this technology. Here is a site to think about. There used to be a site, cncgunsmithing.com, where the guy had published complete CAD and g-code for making firearms. We're talking semi-auto pistols and AR-15 lower receivers. I notice his site has disappeared. Not sure what happened, but once you have a complete CNC shop, you can make almost anything.

Thanks Bob,
for the very informative and insightful post.

I appreciate the effort, time and resources you put into it.

I can imagine that there would be a need for "boiler-plate"? g-code/s "out there" - the same as there was (and is) for "AutoLISP" code for use in AutoCAD - as I found out when I was doing minor work in AutoLISP on my AutoCAD quite some years ago. It's amazing how much of it there is - even or especially - yet.

Next, I tried the link to your site at www.cnccookbook.com and was not able to access the site. I just got the "The page cannot be displayed" message on my screen. Any reason? I'd like to have a look at it.

You have very nicely pointed out how many parts there are to be considered and set to work to get a fully functional CNC system up and running.

I would have no problems with the "mechanicals" if needs be - "ball-screws" and all. Possibly the same for minor editing to G-code but I'd need the CAM software to translate it to g-code in the first instance.

I will explore the wizards in Mach as you suggest in the initial stages at least.

I'd get my local computer wiz to do the "computer bits" as he is very god indeed with that. His charges are very reasonable and he can source stuff cheaper than I can. He has a net-work of people that most of us could only dream about.

I am not "married" to AutoCAD and would be quite at ease with moving to a "matched" CAD/CAM set-up. What I intend doing (at this stage anyway) is just simple stuff as individual items. The "assembly" bit can come later.

I always try to crawl before I walk - "hasten slowly" if you like. I don't need a cacophony of "bells and whistles" at this initial stage.

I just want to explore what CNC is to me and what it and I can do in small incremental steps on a progressive basis.



My recommendation would be to really look for a community college CNC class or hopefully a buddy that's all over it. I could definitely see a case for a club or small group of machinists to go in together and share the software. You only use it for a short part of the machining process and it costs a lot of money.

I can see the sense and logic in that but I cannot go out on my own at night (most days too) and leave my wife at home as she has "problems" and gets "worried" - "stress" is a euphemism. She would worry herself stiff if any people she did not know well were in the house or the work-shop.

The nearest College that does this sort of course is 15 miles, next 2 are 30 miles away. There are no relevant "clubs" within reasonable distance either.

So that's out - I've got to "go it alone".

I have a lot to think about, as you say.

It is all "coming together" as regards my appreciation of the many things to be considered and implemented.

Thanks to your advice and that of others, I can more clearly see that getting the Seig 4X - even in a "turn-key" set-up is only a part of a road that needs to be started.

Getting to the end of the road to where the Seig CNC is all set up and it and I are "ready to go" is quite a journey which is well under way with this thread and others and people like yourself.

Thank you.

lazlo
11-16-2007, 10:27 PM
I would hope that resolving the "4,000" matter that my intended use of the Seig vertical CNC mill as a lathe (and perhaps as a cylindrical grinder) - see previous posts on this thread -will be realised sooner rather than later.

No. If you want to hob the gears, you still need a powered workhead to drive the gear blank.

John built the workhead for his electronic hobber from a 40:1 right-angle gearbox. He replaced the bearings with precision angular contact bearings, and drives the gearbox with a 1/4 HP stepper motor, driven by a Gecko G201 stepper driver:

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/hob%20indexer1.jpg

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/hob%20indexer2.jpg

..and here's John's setup on the Victoria, showing the powered workhead driving the gear blank at 300 RPM, while the mill is driving the hob against it:

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/hob%20indexer4.jpg

lazlo
11-16-2007, 10:31 PM
By the way Tiffie, I'm not sure why you want a complicated gear hobbing setup like John built.
If you just want to cut gears in your shop, you can just use a rotab or dividing head to index the gear blank like Shorty's Youtube video.

It's not going to generate a perfect involute form like you get with John's gear hobber, but I doubt you'll be able to tell the difference in a home shop.

S_J_H
11-16-2007, 10:57 PM
Tiffie, I use Sheetcam for 2.5D work(http://sheetcam.com/. It is very good and cheap and the author is constantly improving it. I give it major thumbs up.
I have also tried deskproto ( http://www.deskproto.com) which can do 4 axis work. It is also very good.
Meshcam is another cheap 3d cam package and also has free trials.(http://meshcam.com/)
Cambam is at the moment totally free and looks real good,(http://www.brusselsprout.org/CAMBAM/)

For cad use which is what you draw or design the part with, I use many free programs.
The easiest cad program I have ever seen is from emachine shop.( http://www.emachineshop.com/) It is very limited though. But for super quick stuff I use it first as it is so fast to use.

Alibre is nice for 3d work and they have a powerful free 3d package( http://www.alibre.com/)
I have tried many others as well. Just find something that you like.
I looked at the dolphin cam package that John likes. It also looks real nice. I might even consider buying it down the road.
I find the cam programming easy. The better you are at machining manually the better you will be with cam. You'll already know what feed rates and DOC are possible. You'll know how the machine will react to certain types of cuts and that different materials will require different feed rates and doc.
I sort of disagree with Bob on this subject. For me the cad design is tough. But I have no problems with cam. If in doubt just be conservative.
Yes any good machinist can convert a machine to cnc or even build one from scratch. But a few subjects need to be learned such as 0 backlash methods and bearing block configurations. If using linear rails then even more study will be needed. Use of proper couplings and if using belt drives you need to know which styles work the best. But it's all straight forward work.
The electronics can be tricky for beginners like myself. But it's not to bad.
Mach3 is complicated at first. It will take time to learn it. But once you get the hang of it it gets easier and easier to the point you'll wonder why you ever thought it was difficult.
Steve

oldtiffie
11-17-2007, 02:57 AM
No. If you want to hob the gears, you still need a powered workhead to drive the gear blank.

John built the workhead for his electronic hobber from a 40:1 right-angle gearbox. He replaced the bearings with precision angular contact bearings, and drives the gearbox with a 1/4 HP stepper motor, driven by a Gecko G201 stepper driver:

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/hob%20indexer1.jpg

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/hob%20indexer2.jpg

..and here's John's setup on the Victoria, showing the powered workhead driving the gear blank at 300 RPM, while the mill is driving the hob against it:

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/hob%20indexer4.jpg


Thanks lazlo.

You got it.

I've had that set-up of John's in mind since he first posted it.

I knew about the "synch-ing" of the hobbing cutter and the work-head (dividing-head) that John had rigged up - work of art really - to emulate the pure mechanical (gearing, chain, linkage, universal joint) set-up that we used on a similar mill many years ago.



John built the workhead for his electronic hobber from a 40:1 right-angle gearbox. He replaced the bearings with precision angular contact bearings, and drives the gearbox with a 1/4 HP stepper motor, driven by a Gecko G201 stepper driver:

I guessed that set-up with the device driven by the cogged belt from the mill arbor was to do with synchronising the arbor and the stepper motor driving the head drive set-up.

I didn't know then why he didn't "CNC" it but apparently it could not be done by pure CNC at the time - but the result for hobbing was achieved.

I realised that the electrical/electronic device that was driven by the mill arbor via the cogged belt was to do with the speed/synchronisation of/with the drive for the dividing-head.

It was all there.

I thought the "CNC-hobbing"'s time had "arrived" when I saw Ken's lathe "CNC-ing" that thread.

But, alas, not so - yet.

I knew that Ken had the spindle speed as a very critical element as regards absolute speed (500 +/- 1 rpm) but I didn't know why.

The "Frog" post/s and John's advice re- the "4,000 encoder samplings/rev (maximum)" provided the answer/s to that.

John said that the optimum "Encoder rate" was about 4,000 per rev of the/a spindle but that it was not "do-able" yet - but well on the way to "being there".

I had no problem with any of the setting parameters at all.

I also reasoned that pending advice to the contrary that it should/may be possible to synchronise the lathe head-stock spindle with the "X" and "Y" feeds by sampling the encoder rate on the "lathe" spindle. The rest just fell into place. I realise that the "4,000" requirement must apply in that application.

I also reasoned that if the "Lathe" set-up was feasible - with or without "screw-cutting" that the same set-up could be used as a cylindrical grinder mounted in a similar manner as I had proposed for the "tool post" when used as a lathe.

The beauty of the "grinder" set-up is that I can set the "traverse" ("X") and "in-feed" ("Y") very accurately as regards rates and limits/stops.

The "grinder" set-up can be implemented on the Seig as it is now as no synchronisation is needed - I could use a stock-standard mini-lathe head and tail-stock as it is in a "bolt-on" setting. I could make the cylindrical grinder into a universal item by mounting it on a graduated "swiveling base" as on a "real" universal grinder or as on a lathe taper-turning attachment. The head-stock could be driven by the motor and manual speed-controller (220/240v) from LMS:
http://littlemachineshop.com/products/product_view.php?ProductID=3060&category=
http://littlemachineshop.com/products/product_view.php?ProductID=3061&category=

I already have the tool-post grinder from Littlemachineshop.com -
http://littlemachineshop.com/products/product_view.php?ProductID=2722&category=
which has 2 speed ranges and a mountable Jacobs chuck for drilling/milling if needs be.

I have the "Proxxon" grinder from LMS.
http://littlemachineshop.com/products/product_view.php?ProductID=2891

I will make a tool-post adaptor for the Proxxon similar to this LMS item: http://littlemachineshop.com/Products/Images/480/480.2935.jpg for internal TP grinding or where a smaller TPG is required.

If I can get the "screw-cutting with a lathe" set-up going, I might consider using the TPG as a thread grinder. I would "tilt" the TPG by an amount = thread helix angle so that the wheel was aligned to the thread. This would solve afew problems as well.

O have the dressing tool for dressing grinder wheels at specific angles to a fairly high order of accuracy. I could use it on my Tool and Cutter grinder or on the mill.
http://cdcotools.com/item.php?itemid=283
http://cdcotools.com/item.php?itemid=27
http://cdcotools.com/item.php?itemid=205
http://cdcotools.com/item.php?itemid=206

I hope this explains "where I'm at" and "where I want to be".

oldtiffie
11-17-2007, 05:15 AM
Tiffie, I use Sheetcam for 2.5D work(http://sheetcam.com/. It is very good and cheap and the author is constantly improving it. I give it major thumbs up.
I have also tried deskproto ( http://www.deskproto.com) which can do 4 axis work. It is also very good.
Meshcam is another cheap 3d cam package and also has free trials.(http://meshcam.com/)
Cambam is at the moment totally free and looks real good,(http://www.brusselsprout.org/CAMBAM/)

For cad use which is what you draw or design the part with, I use many free programs.
The easiest cad program I have ever seen is from emachine shop.( http://www.emachineshop.com/) It is very limited though. But for super quick stuff I use it first as it is so fast to use.

Alibre is nice for 3d work and they have a powerful free 3d package( http://www.alibre.com/)
I have tried many others as well. Just find something that you like.
I looked at the dolphin cam package that John likes. It also looks real nice. I might even consider buying it down the road.
I find the cam programming easy. The better you are at machining manually the better you will be with cam. You'll already know what feed rates and DOC are possible. You'll know how the machine will react to certain types of cuts and that different materials will require different feed rates and doc.
I sort of disagree with Bob on this subject. For me the cad design is tough. But I have no problems with cam. If in doubt just be conservative.
Yes any good machinist can convert a machine to cnc or even build one from scratch. But a few subjects need to be learned such as 0 backlash methods and bearing block configurations. If using linear rails then even more study will be needed. Use of proper couplings and if using belt drives you need to know which styles work the best. But it's all straight forward work.
The electronics can be tricky for beginners like myself. But it's not to bad.
Mach3 is complicated at first. It will take time to learn it. But once you get the hang of it it gets easier and easier to the point you'll wonder why you ever thought it was difficult.
Steve

Thanks Steve.

Another great informative post.

I just hope that all who are participating in this thread and those that are "looking in" are getting some value out of this as there have been some very knowledgeable, busy and generous people who have given freely of and with their skills and knowledge for the benefit of others when they might quite understandably just ignored it all and gone on with something else. But, to their credit, and my gratitude, they have not.

You have again provided a very impressive list of info and links etc. that I, somehow or another, have to deal with.

My sincere thanks.

oldtiffie
11-17-2007, 05:54 AM
By the way Tiffie, I'm not sure why you want a complicated gear hobbing setup like John built.
If you just want to cut gears in your shop, you can just use a rotab or dividing head to index the gear blank like Shorty's Youtube video.

It's not going to generate a perfect involute form like you get with John's gear hobber, but I doubt you'll be able to tell the difference in a home shop.


Thanks lazlo.


By the way Tiffie, I'm not sure why you want a complicated gear hobbing setup like John built.


I'm not sure either!! Other than that "I want to and I think I can" there is no rhyme or reason at all!!

I'm not sure I need it either - in fact, to be honest, I don't!! .

The CNC option, when implemented, will be a lot less complicated than John S's one on his "Victoria" mill. If needs be, as an interim measure, I might just emulate John's set-up as it is "do-able".

I've wanted to be able to hob gears in my HSM shop for the best part of 50 years. It is just that for a whole host of reasons that I was not - or just plain did not - do it or get around to it. And the equipment required was entirely out of reach for a whole lot of reasons.

I upgraded my shop over the last 2 years as I am running out of time and I've stared the end in the face once too often recently (cancer - but OK so far - I hope). So my wife and I decided to get it done - which I've largely done.

The gap was CNC in a usable "turn-key" "ready-to-go" set-up. The Seig 4X seems to fit the bill - so to speak.

That set-up on John S's "Victoria" mill was the catalyst really.

I could see true hobbing working - but I didn't know how - until the last week.

Steve and Evan were the next catalysts - and you know the rest of the story. The rest, as "they" say, is history.



If you just want to cut gears in your shop, you can just use a rotab or dividing head to index the gear blank like Shorty's Youtube video.

It's not going to generate a perfect involute form like you get with John's gear hobber, but I doubt you'll be able to tell the difference in a home shop.

Yep - dead right. I can do that now - that is the "Plan A" option until the "CNC "4,000" matter on the Seig 4X" - in which gears are truly hobbed - is "sorted" and then it reverts to being "Plan B".

At that stage I will have both options open to me.

In the meantime I will have the traditional gear-cutting option available to me as it is now on the machines that I have - as you say, the "Shorty's Youtube video" method.

That anything other than hobbing will not generate/produce a perfect involute is 100% right - I'd be silly to even try to say otherwise - and I won't.

As you say, "cut" - as opposed to "hobbed" gears - which are less accurate than hobbed gears, are more than adequate for at least 90+% of anything that I'm likely to do or need.

I guess that I am trying to see what the Seig can be made to do. Hence the interest in "screw-cutting", "Turning" and "cylindrical grinding".

If I can get these items "up and running" - fine. If not - tough.

If I have to wait? I wait.

Will I use them if I get them going? Possibly not.

I'm more interested in the journey than the destination (I don't know if there is one!!!).

Time? Who knows?

John Stevenson
11-17-2007, 06:25 AM
It takes all sorts, some are more adapted to mechanical issues and some can handle electronics.
Cad / Cam I feel comes in between them.

There isn't any instant point once and click program although VCarve from Vectric is very close on simple work.

First off unless it's something simple that you can use a wizard for like a circular pocket or bolt circle where you just filling the boxes then you need a CAD program so you have accurate way points to work to [ turn right for 3" at this corner :D ]

There are many programs out there, some free and some bloody expensive but they all do the same thing in a fashion.
Which one is right is rather like religion, you find the one that suits you and stay with it :D

YOU NEED CAD, if you can't be arsed learning CAD then the CNC route just isn't for you - FULL STOP.

From CAD you then save as a DXF which stands for Drawing eXchange Format which is an industry standard to allow drawings to be shared between applications.

This is the file that then needs to go into CAM [ of whatever form ] to get converted to G-Code which is what the machine understands.
Steve posted a good post with links in post #84

Mach3 also has a simple one built into it called Lazycam, I haven't used it as I have Dolphin and VCarve to work with, remember this is a religion, once you find something that works you stick with it.

Once the drawing is in whatever CAM [ Tm ] most of the work is done as it's displayed as a 2D drawing, i.e. in the flat, no depth.
You then have to decide how to cut this, in what order and add depths to it together with speeds, feeds and tool information.

I have made this sound easier than it is because different programs work different ways and it's not worth going into detail.

Once all this information is entered a decent program will give you a preview in some way. Some are just the toolpath on the screen, some are graphical models.
This is at the point you can play "What if " and alter the cutting sequence, number of passes and in some case the toolpaths to make it better / faster etc.

When you are happy with the on screen presentation you can then post process the drawing.

The post processor is a conversion file that take the information and writes the code for ONE machine. In our case we select the post for Mach 3 and it will write the code for that part.
If we were using a machine with say TurboCNC controlling it then the post processor would be different as there are small differences between controllers.
Mach 3 can read M03 M04 P2.0 which means switch the spindle on [M03] and pause [M04] for 2 seconds [P2.0]to get up to speed.
Turbo CNC has to have this spelt out as
M03
M04 P20.0
Because it can't handle more than one M code per line. it's small differences like this that mean the post processor has to be special for the controller.

I think that goes far enough in this post for explaining CAM.
I will be more than willing to take this to a new post and work a file thru if someones comes up with a simple part or drawing.

I can work it on say two different free CAD programs and two free CAM programs.

.

oldtiffie
11-17-2007, 07:39 AM
It takes all sorts, some are more adapted to mechanical issues and some can handle electronics.
Cad / Cam I feel comes in between them.

There isn't any instant point once and click program although VCarve from Vectric is very close on simple work.

First off unless it's something simple that you can use a wizard for like a circular pocket or bolt circle where you just filling the boxes then you need a CAD program so you have accurate way points to work to [ turn right for 3" at this corner :D ]

There are many programs out there, some free and some bloody expensive but they all do the same thing in a fashion.
Which one is right is rather like religion, you find the one that suits you and stay with it :D

YOU NEED CAD, if you can't be arsed learning CAD then the CNC route just isn't for you - FULL STOP.

From CAD you then save as a DXF which stands for Drawing eXchange Format which is an industry standard to allow drawings to be shared between applications.

This is the file that then needs to go into CAM [ of whatever form ] to get converted to G-Code which is what the machine understands.
Steve posted a good post with links in post #84

Mach3 also has a simple one built into it called Lazycam, I haven't used it as I have Dolphin and VCarve to work with, remember this is a religion, once you find something that works you stick with it.

Once the drawing is in whatever CAM [ Tm ] most of the work is done as it's displayed as a 2D drawing, i.e. in the flat, no depth.
You then have to decide how to cut this, in what order and add depths to it together with speeds, feeds and tool information.

I have made this sound easier than it is because different programs work different ways and it's not worth going into detail.

Once all this information is entered a decent program will give you a preview in some way. Some are just the toolpath on the screen, some are graphical models.
This is at the point you can play "What if " and alter the cutting sequence, number of passes and in some case the toolpaths to make it better / faster etc.

When you are happy with the on screen presentation you can then post process the drawing.

The post processor is a conversion file that take the information and writes the code for ONE machine. In our case we select the post for Mach 3 and it will write the code for that part.
If we were using a machine with say TurboCNC controlling it then the post processor would be different as there are small differences between controllers.
Mach 3 can read M03 M04 P2.0 which means switch the spindle on [M03] and pause [M04] for 2 seconds [P2.0]to get up to speed.
Turbo CNC has to have this spelt out as
M03
M04 P20.0
Because it can't handle more than one M code per line. it's small differences like this that mean the post processor has to be special for the controller.

I think that goes far enough in this post for explaining CAM.
I will be more than willing to take this to a new post and work a file thru if someones comes up with a simple part or drawing.

I can work it on say two different free CAD programs and two free CAM programs.

.

Thanks John.

Huge reply - huge offer.

I hope some one can and will put up the couple of files you require as I'd love to see it.

I am no stranger to CAD - AutoCAD most recently (some years ago).

Exporting in/to DXF format is not a problem either.

Getting the right post-processor that is compatible with the CAD and mach3 is the vital link if I understand you and several others. I will take advice on that.

If choice is related to religion - no problems - as to put it mildly - I have a very eclectic outlook in that regard - anywhere between none at all (default condition) and "what-ever suits at the time - but only for as long as necessary" otherwise.

My requirements will be kept simple at first at least.

I anticipate just using the "canned" and built-in features and wizards on the mill for a while so that I can get the "feel" of using a CNC machine.

Then I will:
- "brush up" on CAD;
- export DXF;
- get to grips with the post processor;
- see what it looks like on the mach screen;
- try the options;
- a series of "dry runs" on the mill before cutting anything; then
- work it up on a progressive basis.

bob ward
11-17-2007, 09:56 AM
YOU NEED CAD, if you can't be arsed learning CAD then the CNC route just isn't for you - FULL STOP.

The promise of a plug & play CNC mill provides the incentive to make me want come to grips with CAD.

A few doodles on the drawing board have always got me by so far, but no longer I fear.

John Stevenson
11-17-2007, 11:51 AM
Bob,
Not as hard as it seems as there are two main ways to draw, however if you take a course they only teach you one way. OK for the youngsters who have had no exposure to drawing but us dinosaurs struggle. ;)

I taught a class some years ago that had a mixture of sexes and ages but they were all beginners.

Without explaining both methods so as not to make anyone biast I got them to draw a simple shape of a rectangle with various corners and arcs.
I had them do it the textbook method which I'll explain shortly then I had then repeat the same job in mirror image so as not to be the same using the patented Stevo "Old Fogie " method.

Then I asked who was better with method 1 and who was better with method 2, The youngsters preferred method 1, the older folk preferred method 2.,

Method 1 as taught by the textbooks consists of starting from 0,0 and drawing each and every entity [ lines, circles, arc etc are called entities in CAD ] from that point. This is know as the Absolute method as everything is absolute from 0,0

Method 2 is for people used to a drawing board and involves drawing a horizontal line and a vertical line and off setting all entities from these to lines then joining all the points up, basically just as we did in the old days with construction lines. This is know as the Relative method as everything is relative to the last entity drawn.
You need a CAD system that can offset entities easily, some make you jump thru hoops to do this.

You choose whichever method suits the way you work.

.

BobWarfield
11-17-2007, 02:08 PM
Bob,
Not as hard as it seems as there are two main ways to draw, however if you take a course they only teach you one way. OK for the youngsters who have had no exposure to drawing but us dinosaurs struggle. ;)
.

I like Rhino3D, which does both relative and absolute. However, it isn't parameteric (perhaps that's where you were leaning with relative?). I have actually spent a fair amount of time with Alibre as well, and ultimately gave up and went back to Rhino. I've also spent time with OneCNC's built in CAD, which is also not bad, but I still prefer Rhino.

CAD really is one of those things were you'll spend a lot of time learning one, and once you do, you won't want to use the others. It just takes too long to relearn a new set of habits. Get a couple and play with them to see which one you grab on to quicker. I'd say at least try Rhino and Alibre. If you already know AutoCAD, you're fine, but I think there is better to be had these days.

My problem with the CAM end is that it's even harder for complex CAM tasks to come to grips with the way the program wants you to think. I played with Dolphin, SheetCAM, the one that plugs into Rhino, and OneCNC. I've looked at LazyCAM, but had already bought OneCNC so didn't try it too much.

I remember when I first started fooling with it, I wanted to hand it a drawing and have it just figure out how to whittle away the raw stock to get to what I'd drawn. No such luck!

In my experience, the CAM programs would almost like to see a series of drawings that are focused on particular machining operations and contain no additional detail to confuse them. They're all about laying out toolpaths. They can simulate the effects of various cutters to a greater or lesser degree.

CAM for mills seems way easier than lathes. Likewise, CAM that's just cutting outlines, for example CAM for a CNC plasma cutter, is easier still. Programs like VCarve are powerful for doing engraving and very easy.

In any event, there is a significant learning curve, and you do need to know how to manually use the machine for best effect, although lots of folks will tell you that you don't. I don't believe it, especially on a homemade CNC machine and cheap CAM software. It might be one thing to have high end CAM choose a bunch of parameters and do the right thing on a Haas or Fadal, but it's quite another on a little home machine.

BTW, you can play with CNC real cheap and cheerful. Here's a CNC mini-router I have that cost me about $600. It's been a lot of fun and a great learning tool:

http://www.thewarfields.com/cnccookbook/CCCNCMiniRouter.html

I can haul it around to show other folks too. You learn a lot about the foibles trying to cut aluminum on a machine like that!

Cheers,

BW

dp
11-17-2007, 02:38 PM
Bob - on the EZClamps job it looks like most of the time cutting was taking out material that had nothing to do with liberating the clamp hidden in the stock. Why the rectangular cutout?

Michael Moore
11-17-2007, 05:14 PM
I'm on the other side of the CAD fence from Bob Warfield.

I started out with ACADr10 which required (IIRC) that to draw a line I pretty much had to say "start here and go at this angle for this length". I could freehand something and trim overlapping lines, but then I was pretty much stuck with whatever length they ended up being.

Rhino 3.0 (I haven't tucked into 4.0 yet) seemed very much of that style where I had to know how long of a line and at what angle I wanted it before I drew it. Once I got a few stable points I could start snapping to features and things would speed up a bit.

It took me a little bit of time to get comfortable with Alibre but I feel a lot more at home with it than Rhino. I can draw a few lines and circles and get a rough outline. Then I come back and set the properties for the different entities. "Line A is 2.562" long, and the end point here is 2" from the X axis and 3 " from the Y axis and I want it parallel to line B over yonder".

And with the parametric feature I can go back up to the middle of the process and change Hole B from 1.5" ID to 1.510" and have that (usually!) successfully generate down through all the rest of the drawing.

Since I tend to have an "oops I should have done this back then" working style that is really handy for me.

A friend really likes RhinoCAM where Visual Mill runs inside of Rhino. He's a lot more knowledgeable about CAD and CAM (he's a real machinist, unlike me). I sent him a message to tell him that when I paid for my VM6.0 upgrade the other day I also paid to have the dongle modified so I could run VM inside of Rhino 4.0. He wrote back today and mentioned that Rhino "is difficult to learn but sooo powerful."

Just what I need, yet another software package that is so powerful it can only be used for good or evil. :)

It appears that having all the Rhino features (if you can figure them out) to set up machining boundaries on surfaces can be a big help so I'll see what I can do to get somewhat comfortable with Rhino.

But right now, I suspect that I'm likely to do the modeling in Alibre and then export the finished model to Rhino. Rhino may be very helpful in healing small discontinuities or in finishing surfaces that may be more complex than Alibre wants to work with (the way I see it Alibre is 3D CAD while Rhino is a surfacing/modeling program so pick the right horse for the course) and RhinoCAM may be a more powerful way to do the CAM process (though that may not be an issue when I see what the VM 6.0 upgrades are).

The important thing is to try a few different CAD programs and hope that the one that you feel most comfortable with will do what you need. Life is too short to spend much of it wrestling with balky software that doesn't think the way you do.


I remember when I first started fooling with it, I wanted to hand it a drawing and have it just figure out how to whittle away the raw stock to get to what I'd drawn. No such luck!

I resemble that remark! I wish it actually worked that way.

cheers,
Michael

BillH
11-17-2007, 08:23 PM
John, you really need a nicer camera :)

John Stevenson
11-17-2007, 08:33 PM
??
Sorry Bill that one went over my head. Mind you so did most footballs :D

.

oldtiffie
11-18-2007, 03:41 AM
I down-loaded mach3 today and installed it on my W2K computer. I hope to install it on my XP machine tomorrow.

Haven't had time to play with it yet.

Will pay for and register it tonight or tomorrow.

Also paid for the CD with all of the videos etc. on it. Should be here in a week or so - I will have a look at it then. There was too much to down-load as my ADSL is a slow as .......................

I have AutoCAD 2004 on my computer but might download one of the lighter CAD applications to draw and export DXF files to the mach3 CAM software.

I'd like to give it a run in slow time to get used to it before the Seig comes out.

Many many sincere thanks to all the people who contributed to this and other threads to get me to the stage I'm at.

I am very grateful and very humbled by the help I've had.

Many thanks.

John Stevenson
12-02-2007, 04:48 AM
Well here's a write up if the KX1's big brother, the KX3, and this is already on US soil although it only looks like a preview machine.

http://www.mini-lathe.com/KX3/KX3-2.htm

.

oldtiffie
12-02-2007, 06:44 AM
Well here's a write up if the KX1's big brother, the KX3, and this is already on US soil although it only looks like a preview machine.

http://www.mini-lathe.com/KX3/KX3-2.htm

.

Thanks John.

I read that write-up with great interest.

It ties together very well with what you have said on this thread.

"Guessed" cost in USA is about US$6,000/0.85 ~ AU$7,000 which with the usual 10>15% extra here compared to the rest of the civilised world takes it to say AU$7,000 x 1.15 ~ AU$8,000

That seems to be "do-able" - subject to the usual reviews and constraints (no - not my wife as she is quite OK spending on stuff in the shop providing that adequate provision is made for all the usual things - which has been done).

It sees to be a true "turnkey"/"all-in" package such that I only have to supply the computer - which is OK as well.

Is there (to be) any on-site set-up and/or assistance by the local agent/s or does it all just arrive in a series of boxes to be assembled and set to work by the customer (me?)?

I have bought the licence for Mach3 and have it installed on my computers. I read quite a bit of the stuff on the Mach3 site - very informative and helpful.

I bought the Mach3 CD with the demo software, Tutorial videos and documentation (pdf's) on it. It only arrived on Friday and I've been busy all week-end re-configuring my net-worked computers. I will get into it shortly as I'm keen to have a reasonably good "grip" on and appreciation of CNC, Mach3 (mill and lathe) and coding etc. So a lead time of 6 months before the new Seig mill comes here will do me fine.

Again, many thanks for your info, patience end perseverance as it is all very much appreciated.

tmarks11
12-29-2007, 12:38 PM
Sieg and the big box players like Grizzly have identified there is a market out there for an off the shelf machine and these two have been produced to fill that gap.
rats. You got my hopes up that I would open the 2008 Grizzly catalog and see these babies available.

No such luck; currently it looks like Smithy is the only US source (since Syil is not going to pick up the Sieg KX3, instead producing their own updated version in a different facory). I have some doubts about buying from Smithy, partially because I have a hunch they have a heck of a mark-up on these machines.

BobWarfield
12-29-2007, 01:41 PM
Bob - on the EZClamps job it looks like most of the time cutting was taking out material that had nothing to do with liberating the clamp hidden in the stock. Why the rectangular cutout?

Chip clearance. The first attempt involved just milling the outline without the cutout. 3 broken end mills later, that plan was abandoned. It's quite a deep cut relative to diameter and no amount of compressed air or feedrate fiddling seemed to help. At a particular depth the endmill would jam with chips and snap.

Once the clearance was provided, the issue went away. You'll see later attempts involved drilling holes and then milling the outline. That worked better and was faster.

The reality is that while the little mill cuts metal, it isn't particularly happy about it. I need to get my big Industrial Hobbies mill converted and then life will be good.

Cheers,

BW