PDA

View Full Version : A Call for Beta Testers for a Lathe Electronic Edge Finder



rgsparber
12-24-2012, 02:42 PM
Mark Cason and I have developed an edge finder that simply connects to many brands of lathes with no modifications to the machine and detects when the tip of the cutter comes in contact with the workpiece. Accuracy is better than +/- 0.0001".

Eventually it will be offered for sale by Mark's daughter. But for now we want to be sure it works reliably and has a good user interface. So we are enlisting beta testers to help us test it out.

You will receive a Lathe Electronic Edge Finder, Model 1. Mark Cason will cover the cost of the device plus shipping within the USA. You will also receive a draft of the user's guide from me via email.

In exchange, you must commit to using the LEEF at least 3 times a week starting in the week you receive it and ending when your results are consistently good. You must take detailed notes on your experience and email these notes to us within a day of each use. If the device fails to work as expected, we reserve the right to get it back for study and will ask you to take data from your lathe.

We also expect you to read each version of the user's guide and supply us comments and suggestions on making it better.

To qualify, you must have a lathe with a spindle resistance greater than 3 ohms. Details on how to take this measurement will be supplied upon request.

Thanks!

Rick Sparber

Paul Alciatore
12-24-2012, 02:51 PM
Rick,

I would really love to test your edge finder, but my shop is still in storage. Rats!

Good luck with it.

Dr Stan
12-24-2012, 03:17 PM
Are you primarily interested in CNC lathes or do you also want to try it out on manual machines? If so I'm interested. I'd also like to know the length of the testing period i.e. a month, 6 weeks, 6 months, a year?

BigJohnT
12-24-2012, 07:33 PM
Is this some kind of contact edge finder?

John

lakeside53
12-24-2012, 07:59 PM
Love to but none of my lathes (Polamco TUM35 and Emco V10P) have a tool post to spindle resistance greater than 0.2ohm (Fluke 87) including the leads... which measure at about 0.15 ohm. What's your measurement methodology? It's hard for me to imagine a decent lathe with greater than 3 ohms... so what am I I missing?

MichaelP
12-24-2012, 09:52 PM
I'm pretty sure the OP meant "3 Ohm or less". Otherwise, it may take quite an effort to find testers. Maybe the Chinese stuff with wooden bearings or loose fit everywhere would work. :)

Lew Hartswick
12-24-2012, 10:02 PM
Give us a brief on the measurement method. I can check the 8 Clausing-Metosa at school but not till school starts again in Jan. I have several Ohmeters, that
should be adequate. It will be interesting to see how this works out.
...Lew...

macona
12-24-2012, 10:08 PM
Mine measure .268 ohms with a HP 34401A 4 wire ohm meter. A lot of meters will not read accurately at low resistances.

lakeside53
12-24-2012, 10:15 PM
They may have meant 3 ohm or more - lot easier to detect, but, that's not going to fly. Of course a rigid non-conductive material around the tooling will solve that, and then we can use a light bulb ;)!

macona
12-24-2012, 10:23 PM
My guess is they are measuring a local change in resistance when the tool comes in contact with the work. That is not going to work for most lathes.

uncle pete
12-24-2012, 10:45 PM
I'm in the dicktatorship of outer Mogol I mean Canaduh so I'm out. But when you can? Post the user manual here. You won't find better input about what it should be and all for free. I'm real interested in how it gets slaved to each different tool point and at those repeatable accuracy levels. If your claims can be obtained by the average user? you'll have a winner.

Pete

RussZHC
12-24-2012, 10:56 PM
Thought I was seeing double, is not an electronic edge finder more or less acting the same as tool positioning in this thread http://bbs.homeshopmachinist.net/threads/57243-Tool-changing-with-CNC and like in the linked video from that thread http://www.youtube.com/watch?v=mWnfioI3G0E&feature=youtu.be, tool tip touches, lights a bulb, enters location...?

+ or - Zero
12-24-2012, 10:59 PM
I've more or less been following Rick's work on this project for quite some time. I know he can measure some very small changes in resistance now and I'm fairly sure it's very low spindle bearing resistance he needs tested --he's made two versions, the first was useful with stuff like old Atlas lathes which quite often have high resistance on the bearings (high being comparative). I think the latest iteration is about very low voltages across very low resistances (zero chance of bearing damage from spark jumping in the bearings, is a major design criteria I believe). He's done some quite impressive stuff from what I've seen of it, and I suspect what he needs tested is even more refined.

But I'm sure he will be back and speak for himself. I'll just go so far as to say he's quite gifted and seems to be honorable and honest about what he's doing.

+ or - Zero
12-24-2012, 11:08 PM
here's a youtube of an earlier version he was working on (I don't know if this is the latest thing or not):
http://www.youtube.com/watch?v=4ilY9Pse7TU

Paul Alciatore
12-24-2012, 11:29 PM
To those of you who question the spindle resistance requirement, have you measured it while the spindle is turning? I suspect a spindle, like my SB, that has a journal bearing will have substantively larger resistance when spinning than at rest. At rest, the weight of the spindle will force the oil out of the area between it's bottom and the bearing surface and metal to metal contact will exist. When it is up to speed, the spindle rides on a film of oil, as in non-conductive oil and the resistance will go up. If there is no load on the thrust bearing, the spindle resistance can be fairly high - in relative terms at least.

lakeside53
12-25-2012, 01:08 AM
I can see that as a possibility on a plain bearing lathe with no change gears. I'll test that tomorrow but with 8 bearings (radial, anglar contact and thrust types) in my 14x40 spindle train, plus three internal gears and the external change gears as a path, I'm betting it won't matter. My Emco has only two angular contact bearings, but also has a bunch of gears in contact.

rgsparber
12-25-2012, 11:59 AM
Are you primarily interested in CNC lathes or do you also want to try it out on manual machines? If so I'm interested. I'd also like to know the length of the testing period i.e. a month, 6 weeks, 6 months, a year?

We are mostly interested in manual machines because they are more likely to have the higher resistance. The testing period should be less than a month although the testers may continue to learn things and we would still want to hear from them.

Contact me off list if you are still interested.

Thanks!

Rick Sparber
Rgsparber@aol.com

rgsparber
12-25-2012, 12:01 PM
Is this some kind of contact edge finder?

John

John,

Yes. It uses electronics to detect when the lathe cutter touches the workpiece. The lathe must be stopped during the procedure.

Rick

rgsparber
12-25-2012, 12:05 PM
Love to but none of my lathes (Polamco TUM35 and Emco V10P) have a tool post to spindle resistance greater than 0.2ohm (Fluke 87) including the leads... which measure at about 0.15 ohm. What's your measurement methodology? It's hard for me to imagine a decent lathe with greater than 3 ohms... so what am I I missing?

My lathe and mill have a resistance greater than 3 ohms and serve me well. We do have a model that works with machines with resistances of greater than 0.2 ohms and one that works down to 0.01 ohms. We have plenty of beta testers for those models.

Rick

rgsparber
12-25-2012, 12:11 PM
You can read about lathe EEFs at

http://rick.sparber.org/ma.htm#6

Look near the bottom of this section. There are a few articles and versions of the EEF.

I also wrote an article on how to measure resistances below 1 ohms using a low cost meter:

http://rick.sparber.org/electronics/kelvin.pdf

Rick

rgsparber
12-25-2012, 12:20 PM
I'm in the dicktatorship of outer Mogol I mean Canaduh so I'm out. But when you can? Post the user manual here. You won't find better input about what it should be and all for free. I'm real interested in how it gets slaved to each different tool point and at those repeatable accuracy levels. If your claims can be obtained by the average user? you'll have a winner.

Pete

Pete,

Please see the link I just posted. The last article (Model 2) presents everything I learned while making the 0.01 ohm version. It is a monster at over 100 pages but you don't have to read much of it to make one.

I am working on Model 2.1 which is more automated. But there will be no change is sensitivity.

Rick

rgsparber
12-25-2012, 12:59 PM
Mine measure .268 ohms with a HP 34401A 4 wire ohm meter. A lot of meters will not read accurately at low resistances.

Right you are. Without a Kelvin connection, it is near impossible to get readings in this range. Furthermore, many of these meters put out a lot of current. Some people are concerned about currents above 1 amp arcing their bearings. In my procedure for measuring below 1 ohm, I use 20 mA. Not as noise free as 1 amp but a lot less contraversial.

At 0.268 ohms, you could easily use the Model 2. The Model 1.5 has a threshold of 0.2 ohms and might work but it would be close.

Rick

rgsparber
12-25-2012, 01:05 PM
I've more or less been following Rick's work on this project for quite some time. I know he can measure some very small changes in resistance now and I'm fairly sure it's very low spindle bearing resistance he needs tested --he's made two versions, the first was useful with stuff like old Atlas lathes which quite often have high resistance on the bearings (high being comparative). I think the latest iteration is about very low voltages across very low resistances (zero chance of bearing damage from spark jumping in the bearings, is a major design criteria I believe). He's done some quite impressive stuff from what I've seen of it, and I suspect what he needs tested is even more refined.

But I'm sure he will be back and speak for himself. I'll just go so far as to say he's quite gifted and seems to be honorable and honest about what he's doing.

I do try hard to be "honorable and honest" ;-)

Do note that we are offering beta testers a free unit and we pay postage. In exchange we expect information. We are NOT asking for your credit card number or first born. Furthermore, if you want, you can go to my articles, find the schematic and circuit board artwork, and make your own. My work is given away with no strings attached. If you were to make the edge finder and make a ton of money, I would be glad to have helped advance the art. I would not want nor accept a penny.

Rick

rgsparber
12-25-2012, 01:07 PM
They may have meant 3 ohm or more - lot easier to detect, but, that's not going to fly. Of course a rigid non-conductive material around the tooling will solve that, and then we can use a light bulb ;)!

Early on I did play with wrapping cutters in one layer of paper and using a simple continuity tester. That didn't work well with my parting tool. Others I talked to said it was too much hassle.

Rick

rgsparber
12-25-2012, 01:13 PM
To those of you who question the spindle resistance requirement, have you measured it while the spindle is turning? I suspect a spindle, like my SB, that has a journal bearing will have substantively larger resistance when spinning than at rest. At rest, the weight of the spindle will force the oil out of the area between it's bottom and the bearing surface and metal to metal contact will exist. When it is up to speed, the spindle rides on a film of oil, as in non-conductive oil and the resistance will go up. If there is no load on the thrust bearing, the spindle resistance can be fairly high - in relative terms at least.

I found a catch-22 trying to detect touchdown with the lathe running. The instant the circuit detected the cutter touching the workpiece, it sheared off the high spot. Only with the lathe stopped can I have the cutter touch the workpiece and not remove metal.

Rick

lakeside53
12-25-2012, 01:43 PM
My lathe and mill have a resistance greater than 3 ohms and serve me well. We do have a model that works with machines with resistances of greater than 0.2 ohms and one that works down to 0.01 ohms. We have plenty of beta testers for those models.

Rick

Sorry Rick, I meant no offense with "decent lathe". I missed any prior references or posts about this his topic and it wasn't clear to me whether you thought all lathes were typically greater than 3 ohms, but I see you've done your homework!

I'm reading your papers right now -excellent work!


I'm certainly interested in a "low resistance" kit or finished unit when they become available. I can see that it would be a good replacement for my $$$ Acu-rite probe on my cnc'd BP - I tend not to use it as it a pita to unload the tool just to load the edge finder. I'd be in heaven if is could use the existing tooling but I can see it's a little tricker to get the tooling cutting flutes lined up on a mill. Do you turn the spindle by hand as you approach, or even by hand in reverse (carefully)?

BigJohnT
12-25-2012, 02:42 PM
You can read about lathe EEFs at

http://rick.sparber.org/ma.htm#6

Look near the bottom of this section. There are a few articles and versions of the EEF.

I also wrote an article on how to measure resistances below 1 ohms using a low cost meter:

http://rick.sparber.org/electronics/kelvin.pdf

Rick

Rick,

That is quite a web site full of knowledge! I'm the same way with my web site (http://gnipsel.com/) it's all to help others learn from what I've learned and well a place to look back to when I forget... nap time here so I'll read more later.

John

rgsparber
12-25-2012, 03:02 PM
Rick,

That is quite a web site full of knowledge! I'm the same way with my web site (http://gnipsel.com/) it's all to help others learn from what I've learned and well a place to look back to when I forget... nap time here so I'll read more later.

John

John,

I see that you play with the Arduino. I looked into using the ATTiny85 to control my Model 2 edge finder but it was too complex compared to a simple resistor-capacitor timer and a quad op amp. Too bad. So I'm still looking for a good applicatioin for a very cool device.

Rick

rgsparber
12-25-2012, 03:21 PM
Sorry Rick, I meant no offense with "decent lathe". I missed any prior references or posts about this his topic and it wasn't clear to me whether you thought all lathes were typically greater than 3 ohms, but I see you've done your homework!

I'm reading your papers right now -excellent work!


I'm certainly interested in a "low resistance" kit or finished unit when they become available. I can see that it would be a good replacement for my $$$ Acu-rite probe on my cnc'd BP - I tend not to use it as it a pita to unload the tool just to load the edge finder. I'd be in heaven if it's could use the existing tooling but I can see it's a little tricker to get the tooling cutting flutes lined up on a mill. Do you turn the spindle by hand as you approach, or in reverse (carefully)?

No offense taken. My lathe is certainly not commercial grade but good enough for me.

It is very likely that Home Shop Machinist will run an article written by me on the three Lathe Electronic Edge Finder and it will also mention where you can buy them. I just have to be careful not to make it sound like an advertisement.

I assume you understand that the Lathe Electronic Edge Finder is intended for use on a lathe because there is nothing else like it. For a mill you have other choices like a simple Electronic Edge Finder that is based on a continuity checker.

Rick

lakeside53
12-25-2012, 04:05 PM
I assume you understand that the Lathe Electronic Edge Finder is intended for use on a lathe because there is nothing else like it. For a mill you have other choices like a simple Electronic Edge Finder that is based on a continuity checker.

Rick

I have a simple (but expensive for what it is) electronic edge finder probe for my mill.. but... I have to take out the tooling and reload it. For most of my work I'd MUCH rather have the same system you propose for the lathe that would measure an edge relative to the tooling already installed.

BigJohnT
12-25-2012, 05:10 PM
John,

I see that you play with the Arduino. I looked into using the ATTiny85 to control my Model 2 edge finder but it was too complex compared to a simple resistor-capacitor timer and a quad op amp. Too bad. So I'm still looking for a good applicatioin for a very cool device.

Rick

Rick,

Yes, I did mess with the arduino some but never did get it to do what I wanted... pretty cool anyway.

Are you going to make a demo video for your edge finder? I'm still trying to grasp in my mind how that would be better, faster, safer than traditional methods? I built a couple of DRO350's quite some time back but never did install one on the Samson lathe so a 6" scale and a 3/8" dowel is fast and easy tool setting for me. I even use the dowel method on my CNC lathe as it is fast and accurate and no possibility of damaging the edge of a tool.

John

rgsparber
12-25-2012, 07:00 PM
I have a simple (but expensive for what it is) electronic edge finder probe for my mill.. but... I have to take out the tooling and reload it. For most of my work I'd MUCH rather have the same system you propose for the lathe that would measure an edge relative to the tooling already installed.

Using an end mill as a probe works fine for the Z axis but presents problems for the X and Y. The cutting edge is not uniform assuming you have a helix shaped flute. So you must orient the cutter so the touchdown point is at the maximum diameter from the center of rotation. If you are willing to do that, then my EEF will work for you.

Rick

BigJohnT
12-25-2012, 07:12 PM
Using an end mill as a probe works fine for the Z axis but presents problems for the X and Y. The cutting edge is not uniform assuming you have a helix shaped flute. So you must orient the cutter so the touchdown point is at the maximum diameter from the center of rotation. If you are willing to do that, then my EEF will work for you.

Rick

I'm guessing a dowel of the same diameter could be used...

John

lakeside53
12-25-2012, 07:38 PM
Using an end mill as a probe works fine for the Z axis but presents problems for the X and Y. The cutting edge is not uniform assuming you have a helix shaped flute. So you must orient the cutter so the touchdown point is at the maximum diameter from the center of rotation. If you are willing to do that, then my EEF will work for you.

Rick



No problem, I face that every day when I'm eyeballing, listening to the cutter touch, watching for "witness marks", scuffing cigarette paper or any other short cut. I think rotating carefully by hand in reverse will make good enough contact.

When do you think either the finished items or kits will be available?

rgsparber
12-25-2012, 08:55 PM
Rick,

Yes, I did mess with the arduino some but never did get it to do what I wanted... pretty cool anyway.

Are you going to make a demo video for your edge finder? I'm still trying to grasp in my mind how that would be better, faster, safer than traditional methods? I built a couple of DRO350's quite some time back but never did install one on the Samson lathe so a 6" scale and a 3/8" dowel is fast and easy tool setting for me. I even use the dowel method on my CNC lathe as it is fast and accurate and no possibility of damaging the edge of a tool.

John

If you go to YouTube and search on "rgsparber1" you will find many videos showing me using the EEF.

Rick

DR
12-25-2012, 10:12 PM
This all sounds interesting...

I'm not sure of the practical application of it though. You have a piece of round stock chucked in your lathe, you move the tool in to touch the stock. The edge finder lights up on contact.

What does that tell you? Assuming the stock is, say, nominally 1" diameter, that tells you where the tool tip is for accurate turning, huh?

No, the location of the tool tip is not really representative of the diameter it'll turn for a number of reasons. The roundness of the stock, the accuracy of the lathe chucking device, the real possibility of play in the lathe slide not being taken up since there's no force on the non cutting tool tip.

Not to rain on the parade too much, but why is this type edge finder better than the old method of taking a skim cut and measuring to verify tool location under real cutting conditions?

In my experience the skim cut is the best way, that's what we do on CNC lathes.

MetalMunger
12-25-2012, 10:25 PM
hmmmm....
How many packs of Bugle Boy cigarette paper can I buy for the price of one of these electronic thingies?
2 lathes so it will be X2 for cost.

MichaelP
12-25-2012, 10:27 PM
This all sounds interesting...

I'm not sure of the practical application of it though. You have a piece of round stock chucked in your lathe, you move the tool in to touch the stock. The edge finder lights up on contact.

What does that tell you? Assuming the stock is, say, nominally 1" diameter, that tells you where the tool tip is for accurate turning, huh?

No, the location of the tool tip is not really representative of the diameter it'll turn for a number of reasons. The roundness of the stock, the accuracy of the lathe chucking device, the real possibility of play in the lathe slide not being taken up since there's no force on the non cutting tool tip.

Not to rain on the parade too much, but why is this type edge finder better than the old method of taking a skim cut and measuring to verify tool location under real cutting conditions?

In my experience the skim cut is the best way, that's what we do on CNC lathes. That's how I see it too. Besides, there is an oxide layer that can alter resistance significantly.


It can be used later in the process, however. Let's say before the final cut(s) if you need to re-zero your tool for any reason (new cutter inserted, lost position, etc.). Hopefully, the precision of such measurement will be sufficient. If it's comparable with those milling machine electronic edge finders in this respect, it wouldn't be really practical for a lathe, IMHO.

rgsparber
12-25-2012, 10:28 PM
This all sounds interesting...

I'm not sure of the practical application of it though. You have a piece of round stock chucked in your lathe, you move the tool in to touch the stock. The edge finder lights up on contact.

What does that tell you? Assuming the stock is, say, nominally 1" diameter, that tells you where the tool tip is for accurate turning, huh?

No, the location of the tool tip is not really representative of the diameter it'll turn for a number of reasons. The roundness of the stock, the accuracy of the lathe chucking device, the real possibility of play in the lathe slide not being taken up since there's no force on the non cutting tool tip.

Not to rain on the parade too much, but why is this type edge finder better than the old method of taking a skim cut and measuring to verify tool location under real cutting conditions?

In my experience the skim cut is the best way, that's what we do on CNC lathes.

DR,

The EEF is intended to be used for tool changes on a lathe. You chuck up some stock and turn a reference surface. Then you remove that tool and put in another one. For example, changing from right hand cutter to left hand cutter. The EEF lets you find zero for the new cutter.

I do agree that there is some uncontrolled error due to slide play. Yet when I cut a wide groove using a Diamond RH and then LH cutter, I found no perceptible ridge between them on the OD.

I also agree that a skim cut is the best way to establish zero. But what do you do if you are touching down on a reference surface and cannot make a skim cut?

Rick

rgsparber
12-25-2012, 10:35 PM
hmmmm....
How many packs of Bugle Boy cigarette paper can I buy for the price of one of these electronic thingies?
2 lathes so it will be X2 for cost.

MetalMunger,

I have used cigarette paper for more than 20 years. Still on my first pack of papers. Can't beat them for low cost.

There are two cases where the papers give me problems. One is trying to detect touchdown with a boring bar down a deep hole. The EEF can do that. The other is when I want my best accuracy. The EEF can detect touchdown to better than 0.0001". I know there are other factors influencing accuracy but just comparing paper to EEF, I am unable to get that kind of accuracy with my paper.

As far as cost goes, if you have a spindle resistance of greater than 2 ohms, you can build the EEF for about $3.50 in parts. So maybe it is 35 times more money than paper. ;-)

Rick

rgsparber
12-25-2012, 10:41 PM
hmmmm....
How many packs of Bugle Boy cigarette paper can I buy for the price of one of these electronic thingies?
2 lathes so it will be X2 for cost.

MetalMunger,

Why did you say "times 2 for cost"? The EEF is just a box with two wires sticking out of it. You can move it to any machine as long as the spindle resistance is within its range.

Rick

rgsparber
12-25-2012, 11:02 PM
This all sounds interesting...

I'm not sure of the practical application of it though. You have a piece of round stock chucked in your lathe, you move the tool in to touch the stock. The edge finder lights up on contact.

What does that tell you? Assuming the stock is, say, nominally 1" diameter, that tells you where the tool tip is for accurate turning, huh?

No, the location of the tool tip is not really representative of the diameter it'll turn for a number of reasons. The roundness of the stock, the accuracy of the lathe chucking device, the real possibility of play in the lathe slide not being taken up since there's no force on the non cutting tool tip.

Not to rain on the parade too much, but why is this type edge finder better than the old method of taking a skim cut and measuring to verify tool location under real cutting conditions?

In my experience the skim cut is the best way, that's what we do on CNC lathes.

DR,

If you go to

http://rick.sparber.org/sceef.pdf

and look at the text starting on page 10, I show some shop experience with the EEF. This is an older version of the EEF but the physics is the same.

Rick

DR
12-25-2012, 11:04 PM
DR,

The EEF is intended to be used for tool changes on a lathe. You chuck up some stock and turn a reference surface. Then you remove that tool and put in another one. For example, changing from right hand cutter to left hand cutter. The EEF lets you find zero for the new cutter.

I do agree that there is some uncontrolled error due to slide play. Yet when I cut a wide groove using a Diamond RH and then LH cutter, I found no perceptible ridge between them on the OD.

I also agree that a skim cut is the best way to establish zero. But what do you do if you are touching down on a reference surface and cannot make a skim cut?

Rick

Well if it works for you, then I can't argue with success.

With a reference surface that can't be skimmed I would use paper or best is a piece of brass shim stock of known thickness.


Getting a machine to cut accurately to the needed diameter can be difficult sometimes, especially on lower end lathes. Even occasionally on a quality CNC. A number of times I've had the situation on parts where there are multiple diameters that all come out right on and there might be one diameter that's not right. There could be some quirk about that one diameter and the relationship of cutter geometry to diameter making it cut over or undersize. Typically in the case of that one "bad" diameter you'd change the programmed diameter to compensate. My newest lathe that's just being installed has a secondary tool table where you compensate for the irregular diameter there and still program the part to the print dimensions.

Anyway, I'm getting a bit off the topic of edge finders. So, again, if it works more power to you.

rgsparber
12-25-2012, 11:12 PM
Well if it works for you, then I can't argue with success.

With a reference surface that can't be skimmed I would use paper or best is a piece of brass shim stock of known thickness.


Getting a machine to cut accurately to the needed diameter can be difficult sometimes, especially on lower end lathes. Even occasionally on a quality CNC. A number of times I've had the situation on parts where there are multiple diameters that all come out right on and there might be one diameter that's not right. There could be some quirk about that one diameter and the relationship of cutter geometry to diameter making it cut over or undersize. Typically in the case of that one "bad" diameter you'd change the programmed diameter to compensate. My newest lathe that's just being installed has a secondary tool table where you compensate for the irregular diameter there and still program the part to the print dimensions.

Anyway, I'm getting a bit off the topic of edge finders. So, again, if it works more power to you.

DR,

I do hope you go to

http://rick.sparber.org/sceef.pdf

and look at the text starting on page 10. If i am deluding myself, I REALLY want to know it before I offer "snake oil" to others.

Rick

tyrone shewlaces
12-26-2012, 12:10 AM
No, the location of the tool tip is not really representative of the diameter it'll turn for a number of reasons.
Personally, I won't trust dials or DROs on anything critical either. I only trust my final measurement, which often needs a little filing or polishing to make it good anyway.


Not to rain on the parade too much, but why is this type edge finder better than the old method of taking a skim cut and measuring to verify tool location under real cutting conditions?
In my experience the skim cut is the best way, that's what we do on CNC lathes.
There are tool probes on nicer CNC lathes. I've used them and they are extremely accurate. But even there, tool pressure alone can make every tool slightly different than the others. Setting up the dials or DRO or CNC tool offsets is always only a starting point. Only the final cut gives you a verdict. If the final cut is slightly deeper or shallower than your skim cut and at a different diameter, you may not be where you're supposed to be either.

You have to set up with something. I use different methods depending on the situation and don't see anything wrong with having another way to pick from (if it works right). This might be kind of handy from time to time.

BigJohnT
12-26-2012, 08:50 AM
If the "device" could be fitted with a spring loaded plunger and had an output contacts that could stand 24vdc that would be a nice Z offset setting device for my CNC lathe... thinking a bit more if the output could be selected to be either N/O or N/C it could be used with a renishaw type of probe and connected to the probe input of my CNC mills, again very nice.

John

rgsparber
12-26-2012, 09:19 AM
If the "device" could be fitted with a spring loaded plunger and had an output contacts that could stand 24vdc that would be a nice Z offset setting device for my CNC lathe... thinking a bit more if the output could be selected to be either N/O or N/C it could be used with a renishaw type of probe and connected to the probe input of my CNC mills, again very nice.

John

Driving a relay with a Model 1 or Model 1.5 version would be easy although I would prefer to use an opto isolator which would take up less space and should work fine for CNC.

What I am having trouble understanding is how you would use a spring loadded plunger. It seems to me it would not be accurate.

One beta tester uses an 1/8" diameter by 1" long piece of drill rod in his CNC mill as a probe. The rod has some spring if there is a small crash yet does not hurt the machine if the crash is big. Worst case he tosses the rod and put in another peice. He has yet to damage the rod.

Rick

BigJohnT
12-26-2012, 09:39 AM
Driving a relay with a Model 1 or Model 1.5 version would be easy although I would prefer to use an opto isolator which would take up less space and should work fine for CNC.

What I am having trouble understanding is how you would use a spring loaded plunger. It seems to me it would not be accurate.

One beta tester uses an 1/8" diameter by 1" long piece of drill rod in his CNC mill as a probe. The rod has some spring if there is a small crash yet does not hurt the machine if the crash is big. Worst case he tosses the rod and put in another piece. He has yet to damage the rod.

Rick

The spring would prevent any damage if there was some over travel. The same thing could be used to touch off Z on the mill to set the fixture offset for the tool table and if made short enough placed on top of the material for Z using a probing routine.

The output would be going to an input on an I/O card so no real current is needed like a relay.

I have a ton of work to finish before the end of the year but in between making parts I'll study your design. I am very interested in making a tool touch off for my CNC mill and a probe.

John

rgsparber
12-26-2012, 12:00 PM
The spring would prevent any damage if there was some over travel. The same thing could be used to touch off Z on the mill to set the fixture offset for the tool table and if made short enough placed on top of the material for Z using a probing routine.

The output would be going to an input on an I/O card so no real current is needed like a relay.

I have a ton of work to finish before the end of the year but in between making parts I'll study your design. I am very interested in making a tool touch off for my CNC mill and a probe.

John

Ive had bad luck with spring loaded probes. The spring does not always return the reference surface to the same place so I get additional error. It will be interesting to see how your design works.

Rick

lakeside53
12-26-2012, 12:46 PM
My cnc probe input expects a NO closure. Opto might work if the source or sink currrent and polarity is correct (no idea what is is from my mill) but I'd prefer a relay contact output for sure, and it would be more flexible for carrying machine to machine. There are some really tiny relay packages available.

Forestgnome
12-26-2012, 05:08 PM
I wasn't going to comment, as everyone has their own way of doing things, but here's a little feedback for the inventor. This just isn't something I would use, as it seems unnecessarily complicated. By that I mean when I want to zero my infeed, I grab a magic marker and scratch through the ink with my tool. Fast and simple. On the mill i usually use the cigarette paper trick. The thing is both techniques eliminate error from runout of material, chuck, or spindle. Your tool doesn't do that. Again, maybe some people will find it useful.

BigJohnT
12-26-2012, 05:29 PM
Ive had bad luck with spring loaded probes. The spring does not always return the reference surface to the same place so I get additional error. It will be interesting to see how your design works.

Rick

I did not think of that but it gives me something to ponder while making parts... probably make it like the renishaw probe with 3 arms and 6 balls.

John

DR
12-26-2012, 06:02 PM
DR,

I do hope you go to

http://rick.sparber.org/sceef.pdf

and look at the text starting on page 10. If i am deluding myself, I REALLY want to know it before I offer "snake oil" to others.

Rick

Okay, last evening I did download your PDF. Maybe I OD'd on Christmas cookies, but I couldn't figure any of it out.

The part about infeeding .005" and only seeing a .003" difference in material diameter, that would be expected on a very light duty machine like a 12" Atlas. A good reason why the test skim cut is my preference.

Your device is not likely to be something that'd be used in my shop so my input may be irrelevant to your project. If it works for you that's all that matters.

rgsparber
12-28-2012, 10:43 AM
My cnc probe input expects a NO closure. Opto might work if the source or sink currrent and polarity is correct (no idea what is is from my mill) but I'd prefer a relay contact output for sure, and it would be more flexible for carrying machine to machine. There are some really tiny relay packages available.

The Model 1 and 1.5 can drive a relay or opto. Either one would be off of the main board and connected with wires.

Rick

rgsparber
12-28-2012, 10:50 AM
I wasn't going to comment, as everyone has their own way of doing things, but here's a little feedback for the inventor. This just isn't something I would use, as it seems unnecessarily complicated. By that I mean when I want to zero my infeed, I grab a magic marker and scratch through the ink with my tool. Fast and simple. On the mill i usually use the cigarette paper trick. The thing is both techniques eliminate error from runout of material, chuck, or spindle. Your tool doesn't do that. Again, maybe some people will find it useful.

I'm glad you did replay. There any many ways to do just about anything in machining. I've used a marker and I've used cigarette paper. Both work well as long as you have access to the reference surface. Not so easy to do if using a boring bar down a blind hole. How often is that needed? Not very often.

The marker and cigarette paper are also difficult to interface to CNC ;-)

What I like about using the EEF is that I just slap on the two probes and feed in the cutter until I see the red LED. Then I zero my dial. I remove the probes and start making chips. For me, it is faster than using paper and is more accurate.

You said the marker and paper eliminate runout error. By this do you mean that you use the paper with the machine running? If so, you are right, my EEF can only be used when the machine is off.

Rick

rgsparber
12-28-2012, 11:07 AM
Okay, last evening I did download your PDF. Maybe I OD'd on Christmas cookies, but I couldn't figure any of it out.

The part about infeeding .005" and only seeing a .003" difference in material diameter, that would be expected on a very light duty machine like a 12" Atlas. A good reason why the test skim cut is my preference.

Your device is not likely to be something that'd be used in my shop so my input may be irrelevant to your project. If it works for you that's all that matters.

What I was demonstrating with the .005" versus .003" is calibrating the cutter. It is my undestanding that this is done with CNC too. You must dial in the correction factor for any cutter. I was trying to demonstrate that by using the EEF I was able to calcuate the correction factor so my cuts canceled the fixed error of my machine. This is the "time invarient" part of the error. As long as the cutter does not dull too much, I can set my dial to .005" and will get .0036" on average. If I wanted to get, say .005" as my average cut, I would need to calibrate to a larger in-feed. It would be approximately 0.0069" given the correction factor of 5/(3.5).


The variation related to this depth of cut is my "time varient" error. That is the random error associated with my lathe and comes from things like play in my bearings. I cannot control for it. But this error was +/- 0.0003" which is a lot better than I would expect from my light weight lathe.

What you see in my data are a series of skim cuts. A single skim cut will not give you the average and deviation.

The bottom line in all of this is that my EEF gives accurate and repeatable results. If I ran this test with cigarette paper and my level of skill with this paper, I would expect a time varient error greater than +/- 0.0003". How much this matters depends on the machinist.

Forestgnome
12-29-2012, 11:16 AM
I'm glad you did replay. There any many ways to do just about anything in machining. I've used a marker and I've used cigarette paper. Both work well as long as you have access to the reference surface. Not so easy to do if using a boring bar down a blind hole. How often is that needed? Not very often.

The marker and cigarette paper are also difficult to interface to CNC ;-)

What I like about using the EEF is that I just slap on the two probes and feed in the cutter until I see the red LED. Then I zero my dial. I remove the probes and start making chips. For me, it is faster than using paper and is more accurate.

You said the marker and paper eliminate runout error. By this do you mean that you use the paper with the machine running? If so, you are right, my EEF can only be used when the machine is off.

Rick

Yes, I use the paper for milling with the cutter turning. I don't use paper on the lathe. I could see where the device might be handy in the boring situation you mentioned. If the reference surface in boring isn't available or usuable at the opening of the hole such as a beveled entry. In that case I reference off the outer surface of the piece.