PDA

View Full Version : Pulley choice for modified lathe:poly vee or timing belt?



Greg Q
01-04-2010, 05:40 AM
I am about to restart the slow restoration and modification of my 10 X 20 lathe. Originally it was powered with a 3 hp, two range 415v commutator motor with a rare and obsolete electronic variable speed. (Phillips Motronic on a 1965 AI Hembrug DR-1M)(think:7/8 scale Monarch 10EE)
http://www.lathes.co.uk/hembrug/

A bit of history: The machine was owned by the phone company and used in a research facility. The original drive system had failed and been replaced by an obviously hand-rolled more modern electronic speed control. The replacement drive had also failed, and the lathe was greased up and stuck in storage in 1983.

I wanted to keep as much of the lathe's original speed range as possible, but at 45-4000 rpm and no gearing, this was a big ask. I was powerfully impressed by Macona's servo motor modification of his Monarch so went the same way.

I now have a suitable 2.5 kW servo motor and drive for the spindle and need to think about pulleys. The characteristics of the servo are rapid acceleration, 3000 max rpm, rapid electronic braking, stable speed control.
The servo motor does not have a long shaft-maybe 40mm or so I need a compact power transfer solution.

The lathe does have a 4:1 back gear in the headstock.

I'm sure I should pull the spindle to address cleaning and servicing the bearings, so while in there I can modify the drive pulley

My choices are:
1.Using the existing B-series double vee belt pulley with one belt
2.Making an L-series timing belt* spindle pulley, or
3.Making a poly-vee pulley.

The timing belt pulley would allow using the servo's step/dir for indexing which might be a neat feature, but I have no planned application yet.

I recently read that 1500 or so is the max for smooth running with timing belts. Smooth is more important to me than the indexing, hence the serpentine belt option which is smooth and transmits more power than a similar width standard vee.

Can you guys pile on with your opinions please?

*only because ebay coughed up a new, cheap L-series tooth cutting wheel.

Thanks
Greg

Bguns
01-04-2010, 06:06 AM
The Poly Vee will be quieter and smoother..

My Harley (timing) secondary drive belt is pretty quiet, but rpms are not as high as you need.. With over a 100 hp its holding up fine, but much heavier than a Poly Vee of same width.

With a 150 amp truck Poly Vee drive alternator needing ~ 3 HP to reach full output, a Poly Vee should easily handle Lathe drive duty (and having backgear just makes life better)

You should be able to add an encoder wheel now or later if needed.

These guys sell the inserts to make it a snap...:

http://www.toolflo.com/cats/Automotive.pdf

oldtiffie
01-04-2010, 06:33 AM
If the servo motor with a maximum speed of 3,000RPM is direct-coupled to the lathe spindle with a required top speed of 4,000RPM the servo:spindle speed ratio is 3,000:4,000 = 3:4 both of which exceed the 1,500RPM of the timing belt - so that option is out.

"V"-belts can be "lumpy".

I'd opt for a micro-vee serpentine belt as it will be smoother. There are a vast range of them at local motor stores directly (off the shelf) or by special order from the catalogues. These belts will have power capacity to spare if chosen correctly - the catalogues and manufacturers web sites will help there.

I've done the math for servo motor and spindle speeds for direct and back-gear - assuming no slip and no speed over-lap between bottom of "HI" range and top of "LO" range.

That is a bit impractical of course and will need some re-working but it is a good start or "ball-park" to start from.

I will tidy up the math and will scan and up-load it in the next day or so.

Evan
01-04-2010, 07:02 AM
Top speed of 1500 RPM for a timing belt? They will go much faster than that.

However, slip is your friend in a crash. Use a poly vee belt.

Greg Q
01-04-2010, 07:54 AM
Thanks guys. I guess its unanimous then?

I hadn't thought of the crash scenario Evan, but that sobering point has been taken. Thanks.

Tiffie: Thanks for that. I am guessing at the useable lowest rpm of the servo, but think I should be safe at 200, allowing a 1:1 motor/spindle ratio. This gives me 50-750 in back gear and 200-3000 in direct. This would be a first approximation which I could in future tweak if required. It might be possible to increase that ratio to 1.3:1 if the servo makes good torque at 150 rpm.

(Motor & drive are Allen Bradley ultra series with encoder feedback)

Bguns, thanks for the link on the insert tooling, but I'll be using a friends Rivet 608 for the machining and that is going to be an HSS job I think. Machinery's handbook has an extensive section on multi-vee dimensions and specifications. Enough at least to allow grinding tooling and having a stab at it.

Greg

Bguns
01-04-2010, 08:11 AM
So you are going to cut it one groove at a time ....

Just to save a Dime :)

Nice lathe ...

Understand about Inserts for 1 or 2 offs, of course.. That HSS will work fine...

Evan
01-04-2010, 08:19 AM
Why not cut one groove at a time?

http://metalshopborealis.ca/pics/sheave5.jpg

Greg Q
01-04-2010, 08:24 AM
Gosh no, not to save a dime as buying more tools is never a sad occasion for me. It's just that the Rivet is only set up for traditional tooling right now as it too is a work in progress. Neat lathe though-chock full of gratuitous beauty in its castings, form and detail. If it was mine it would be in the living room.

Oh wait-I'm married. But you get the idea.

Greg

Greg Q
01-04-2010, 08:28 AM
Material? I have some 90mm dia 6061 T6 aluminum cut-offs that would be a good size. Is there any reason to choose a different material?

(Evan, that looks like steel in your photo. Is it, and if so was there a compelling reason for it?)

SpyGuy
01-04-2010, 09:17 AM
Material? I have some 90mm dia 6061 T6 aluminum cut-offs that would be a good size. Is there any reason to choose a different material?

I think the aluminum would need to be hard anodized to avoid wear. I don't think the anodization will appreciably affect the dimensions of the grooves.

Stainless would be nice! :D

Evan
01-04-2010, 09:57 AM
It's aluminum and aluminum wears just fine in such an application. Look how long Zamac V belt pulleys last. A poly vee belt put on a lot less wear on the pulley simply because the contact surface is greater and there is less differential slip due to the thinness of the belt.

mixdenny
01-04-2010, 11:35 AM
I recently read that 1500 or so is the max for smooth running with timing belts.

Are these the same timing belts that are on millions of automobile engines that reach 6500 rpm?

Dennis

Black_Moons
01-04-2010, 11:48 AM
Hehe. I agree with other posters, 1500RPM sounds like a crock.

If anything id suspect it has a SFM limit not an RPM limit, ie a surface feet per min, ie RPM * diamiter * PI (/12" of course if your diamiter is in inchs)
And id wager that somewhat relies on how much tension the belt is under, especialy under the slacked 'return' part of the belt.

Smaller pullys let you reduce SFM at the expense of more torque required on the belt and more belt tension required to achive that torque.

Question then becomes will belt X transmit 3hp and not require a pully too small to be under its critical SFM, or will that diamiter result in too much torque, And is that diamiter pratical for your application.

Evan
01-04-2010, 12:03 PM
Timing belts don't have a fixed maximum rpm. It depends on load, pulley sizes and the particular type of belt. They can run at 10K in some cases. The faster they go the more horsepower they can transmit since the limit on horsepower is the torque required to rip the teeth off. More rpm means less torque for a given horsepower.

DR
01-04-2010, 02:12 PM
A flux vector drive with associated 3 phase motor should do fine with the rpm range needed.

If there's any possibility of future use with spindle positioning and such I suggest a timing belt drive.

oldtiffie
01-04-2010, 03:13 PM
Thanks guys. I guess its unanimous then?

I hadn't thought of the crash scenario Evan, but that sobering point has been taken. Thanks.

Tiffie: Thanks for that. I am guessing at the useable lowest rpm of the servo, but think I should be safe at 200, allowing a 1:1 motor/spindle ratio. This gives me 50-750 in back gear and 200-3000 in direct. This would be a first approximation which I could in future tweak if required. It might be possible to increase that ratio to 1.3:1 if the servo makes good torque at 150 rpm.
(Motor & drive are Allen Bradley ultra series with encoder feedback)

Bguns, thanks for the link on the insert tooling, but I'll be using a friends Rivet 608 for the machining and that is going to be an HSS job I think. Machinery's handbook has an extensive section on multi-vee dimensions and specifications. Enough at least to allow grinding tooling and having a stab at it.

Greg

Greg.

My math shows the following RPM spindle and motor relationship (assuming 100% efficiency and zero belt slippage:

Main speed:
HI: 4,000:3,000
LO: 160:120

Back gear (4:1)
HI: 160:120
LO: 120:40

So your motor speeds range (RPM) is:
HI: 3,000
LO: 120

Your lathe spindle speed range (RPM) is:
HI: 4,000
LO: 40

There will be a speed over-lap between the spindle speeds of 120RPM.

Back-gear LO range with motor speed at 120 RPM should not be a problem as - assuming no losses - the spindle torque will be 4 X motor torque because of the back-gear 1:4 ratio.

Providing that you don't really load up that lathe and hog it, my guess is that it will run any reasonable load using most common materials and HSS/TC cutters.

That 40 RPM LO back-gear spindle speed should be ideal for screw-cutting, form-tool work, larger drills and machine reaming.

I saw a good range of serpentine automotive belts in Super-Cheap Auto (Frankston - OZ) recently. The selection guides in the catalogues seemed pretty good too. They had the manufatures web addresses as well.

If the local press/media here is anything to go by, then you will be one of the few who has come out in front with Telstra. I am quite happy with them as my phone provider - as I don't like BigPond as an ISP. I use Chariot as my ISP.

That seems to be a nice lathe.

I will post the math work-sheet later in the day after I've tidied it up.

derekm
01-04-2010, 03:31 PM
Top speed of 1500 RPM for a timing belt? They will go much faster than that.

However, slip is your friend in a crash. Use a poly vee belt.

And dont overspec the belt - get one that will slip if over driven. So use a narrow a belt as you can, and tension it properly.

Greg Q
01-04-2010, 03:59 PM
Hi all.

The 1500 speed on a timing belt was the speed at which the user perceived spindle vibrations from the belt. I don't know what accuracy was used in aligning the two pulleys, nor do I know if an idler pulley was used. It could be that whip or a harmonic was developing in the machine in question.

In no way did I mean to imply that 1500 was a max speed, just that I wanted to avoid even minute vibrations and feared that the timing belt would be likely to induce them. (As much as I still think the indexing ability would useful if I ever make a tool post spindle)

In the course of this thread the idea of belt slippage as a safety feature makes me think the poly-vee trumps the toothed belt,

Machinery's has a formula that suggests a 1" multi-vee will handle 5 hp...I wanted maybe a bit wider to allow for the sudden braking loads the resistors are capable of delivering. I wonder now though with the slippage mentioned here.

What say ye? an intentional 30% undersize sound about right?

Thanks again to all,

Greg

Greg Q
01-04-2010, 04:09 PM
Tiffie wrote:If the local press/media here is anything to go by, then you will be one of the few who has come out in front with Telstra.

Well not exactly. A NSW machinery dealer got it first, and he stitched me up pretty good. Still, it should be a great machine when completed. Plus I have learned some handy skills like industrial wiring, scraping, machine tool testing, spray painting, precision machine survey and used machinery dealer psychology 101.

Greg

S_J_H
01-04-2010, 04:17 PM
Don't forget the 60' wedge polyflex belts. They are great for machines/lathes and run very smooth and can handle a lot of rpm and power.
http://www.gates.com/brochure.cfm?brochure=2790&location_id=3292

I used a 3mm polyflex belt to drive my 45,000 rpm high speed spindle.

Steve

John Stevenson
01-04-2010, 05:35 PM
So you are going to cut it one groove at a time ....

Just to save a Dime :)

Nice lathe ...

Understand about Inserts for 1 or 2 offs, of course.. That HSS will work fine...

Why not cut it all in one go ?

For pence?

http://www.stevenson-engineers.co.uk/files/poly%20vee%20tool.jpg

This is an insert from a Coventry diehead of 11 tpi, standard pipe thread, it's very, very close to a Polyvee J series.

Purists will say that the angle is wrong but it's soon taken care of after a short while running.

I have some cutter spindles that run 8 hours a day since 1999 at 9,000 rpm with no problems and only a couple of belt changes in that time.

.

oldtiffie
01-04-2010, 10:28 PM
Tiffie wrote:If the local press/media here is anything to go by, then you will be one of the few who has come out in front with Telstra.

Well not exactly. A NSW machinery dealer got it first, and he stitched me up pretty good. Still, it should be a great machine when completed. Plus I have learned some handy skills like industrial wiring, scraping, machine tool testing, spray painting, precision machine survey and used machinery dealer psychology 101.

Greg

Thanks Greg.

Quite a learning curve - ain't it?

But you seem to have come through it pretty well.

Well done.

Here is the math work-sheet that I was talking about.

Your back-gear speed range is 40 > 1,000 RPM.

While I understand the need for dynamic electronic braking, excessively heavy pulleys will not assist the braking because of the potential energy in them. Excessive pulley weight requires more torque to accelerate to the required speed.

There is no good reason that I can see why the pulleys should not be made of aluminium as opposed to cast iron.

Reducing the amount of material in the pulleys will help as well.

Reducing the pulley diameters will certainly assist but needs to be balanced against the minimum pulley diameters and contact angles required by the belt manufacturer.

There may well be a compromise required with belt tension as while "slip" may be required in case of a spindle "crash" it may work against the minimum belt tension required for efficient dynamic braking.

The "belt-shriek" and more frequent belt replacements may not please you either.

Its very much a question of making informed "best guesses" and trying it out and adjusting if and as required.

I don't know how your set-up works, but on my two Sieg mills (X3 and a Super X3), I have to return the spindle speed setting to zero and wind it back up again as required. I cannot (re)start it in a mid-speed range.

As I said previously, I have assumed 100% efficiency in terms of "slip" in the servo motor and the belt drive. You may may need to make your own judgments. I have assumed that you don't have a "screw-on" lathe chuck but a "bolt/key on" type.

If you do have a screw-on type, please consider making sure that the chuuck will/does not unscrew under severe braking. If it happens, it is not an experience that you will either forget or want to repeat any time soon!!!

If it were me, I'd err on the side of reducing top speed (pulley ratios) to get lower speeds and higher torques at the lathe spindle as well as perhaps easier starts on the motor.

I hope this sheet helps.

http://i200.photobucket.com/albums/aa294/oldtiffie/Sketches/Lathespeedratios1.jpg

Greg Q
01-04-2010, 11:09 PM
Thanks Greg.

Quite a learning curve - ain't it?


While I understand the need for dynamic electronic braking, excessively heavy pulleys will not assist the braking because of the potential energy in them. Excessive pulley weight requires more torque to accelerate to the required speed.

I don't know how your set-up works, but on my two Sieg mills (X3 and a Super X3), I have to return the spindle speed setting to zero and wind it back up again as required. I cannot (re)start it in a mid-speed range.

As I said previously, I have assumed 100% efficiency in terms of "slip" in the servo motor and the belt drive. You may may need to make your own judgments. I have assumed that you don't have a "screw-on" lathe chuck but a "bolt/key on" type.

If it were me, I'd err on the side of reducing top speed (pulley ratios) to get lower speeds and higher torques at the lathe spindle as well as perhaps easier starts on the motor.


Thanks for the worksheet old son, I appreciate it.

I will use this 6061 that I have here (genuine billet!;) ). I will check, but I'm pretty sure the polyvee will run on a 2 1/2" pulley. I have about 4 1/2" of room inside the headstock cover.

Thanks for your concern on the spindle nose. As it happens it is a D1-4 camlock, and I have three chucks to fit as well as a collet set-up, so no worries about stopping/reversing.

As far as the motor control goes, I am going to try to implement a two range control with lock-out on high range and potentiometer control of the speeds within the ranges. 2000 rpm will be max in low.
I don't want to inadvertently twist a dial and overspeed a chuck, hence my concern.
The servo drive allows for this, as well as an e-stop circuit and limit stops-handy for preventing newbie crashes. I think I'll rig up an electronic carriage stop to implement this.

On slip: I assume that the encoder feedback makes for very accurate servo speed stability so don't anticipate problems on that score.

(The servo package kind of fell into my lap via ebay, otherwise I was going to buy a 5 hp 3 phase induction motor/vector vfd package.)

Finally, I may be asking too much for both 40 and 4000 rpm, so I might aim for a bit lower initially. Luckily the motor mounting has about 8" of adjustment travel so I can always upsize the motor pulley if required.


Greg

oldtiffie
01-05-2010, 12:35 AM
Thanks Greg


As far as the motor control goes, I am going to try to implement a two range control with lock-out on high range and potentiometer control of the speeds within the ranges. 2000 rpm will be max in low.
I don't want to inadvertently twist a dial and overspeed a chuck, hence my concern.
The servo drive allows for this, as well as an e-stop circuit and limit stops-handy for preventing newbie crashes. I think I'll rig up an electronic carriage stop to implement this.

While many of the "more macho" types will "rubbish" this, I don't like 3 and 4 jaw chucks spinning above 3,000 RPM max. It does not improve the bearings etc. if there is any even minor "run-out" or "out of balance".

I'd have a good look at the rated speed for any chuck - and stay under it.

Centripetal force is a PITA.

Here is the formula (simplistically) from:


Formula
The magnitude of the centripetal force on an object of mass m moving at a speed v along a path with radius of curvature r is:[4]

F = mv^2/r

The direction of the force is toward the center of the circle in which the object is moving, or the osculating circle, the circle that best fits the local path of the object, if the path is not circular.[5] This force is also sometimes written in terms of the angular velocity ω of the object about the center of the circle:

F = mrω2

From:
http://en.wikipedia.org/wiki/Centripetal_force

F = mrω2

ω is the angular velocity - a direct derivative of the spindle and chuck speed in RPM

F is the centrifugal force.

M is the mass of the chuck and job

r is the radius of the mass from the chuck/spindle centre.

F = mrω2

So, if we double either m or r we will double F

If we double both m and r we will increase F x 4

If we double ω (speed in RPM) we will increase F x 4

If we treble ω (speed in RPM) we will increase F x 9!!!!

If we increase ω x 4 (speed in RPM) we will increase F x 16!!!

Etc.

The job in the chuck sticks out from the head-stock bearing a lot further than it does in the collets.

Those collets are inherently better balanced as well.


The servo drive allows for this, as well as an e-stop circuit and limit stops-handy for preventing newbie crashes. I think I'll rig up an electronic carriage stop to implement this.

An E-stop is a MUST. It is mandated in a commercial shop under OH&S. There are quite a variety available - try Middendorps ("Middies").

I'd have a physical carriage stop PLUS the E-stop if you like. The mechanical one will ensure a physical stop of the carriage without stopping the spindle. If the mechanical stop fails the E-stop will catch it.

Greg Q
01-05-2010, 01:06 AM
An E-stop is a MUST. It is mandated in a commercial shop under OH&S. There are quite a variety available - try Middendorps ("Middies").

I'd have a physical carriage stop PLUS the E-stop if you like. The mechanical one will ensure a physical stop of the carriage without stopping the spindle. If the mechanical stop fails the E-stop will catch it.

I think the received wisdom on most iron bodied chuck revs is something like 2200 rpm. I have seen ratings for the semi-steel go up to 3000 in some high $$ models.

The stop issue is interesting: Current machines must have e-stops, and foot brakes too above a certain size. It is my aim to incorporate a foot pedal and two red button stops. I'd like the machine to comply with all current standards, and am going to some pains to ensure making a CE compliant electrical installation. I have pretty much outgrown shade tree mechanical shortcuts despite having been raised on them.

The z-feed stop will be a limit stop set to avoid crashes. The Hembrug has an elegant clutch design in the apron which I am sure to screw up until I've set it a few times. The achilles heel of this machine is that the lead screw drive system incorporates two fibre gears in the tumbler reverse for crash protection. The genuine gears are silly money from Holland (I do have two pairs of spares purchased from Jon Leyden in Long Beach.). They take about four hours to replace. No shear pins anywhere in the system.

Middies: I have never had anything but rude "service" from them, and their prices are crazy. I prefer to search online for reputable brands like Kraus Naimer, Seimens, AB, Sprecher & Schuh. I bought a dozen new Telemechanique e-stop mushroom button switches for less than the Middies price for one. (and I will use them all too eventually)

Greg

oldtiffie
01-05-2010, 01:39 AM
Greg.

You are pretty well on top of it.

My blood runs cold when I read here of people running two - or more - machines concurrently and leaving one unattended - even mills and surface grinders!!

I can imagine the panic when something goes wrong and the E-stop (if fitted) is out of reach - as is anyone else if you are in the shop on your own.

I wait at every machine I use - one at a time - I switch it on when I get to it and off before I leave it - every time - band-saw included.

I'm surprised about your local "Middies" as mine is tops - I get "trade" on everything no matter how small - and I don't even run an account there. They go to a lot of trouble to get it right. If I want anything from them it is for "now" (two days latest) and I get it. I can't comment on the prices from other sources, but they beat the local traders here - Bunnings et al included.

I have heard several really pi$$ed off accounts from the Trade about other "Middies" branches. Perhaps its a local staffing issue that central management needs to get to grips with.

I have had several reports of similar "service" at Hare & Forbes and it seems that as it too is "trade-focused" in that anyone not perceived to be "Trade" gets a lesser (or any?) "service". I must say that I get treated very well there as I do at "Total Tools" - at both of which I get the "right price".

I do just as well if not better at our local Steel/metal/welding supplier and fabricator and bearing supplier too.

Other than H&F I usually only buy relatively small amounts at irregular intervals.

Same with out computers.

I pay hard cash or EFT - never credit - right there and then - before I leave - every time.

My guess is that it would/might? be different if I were shopping in Frankston, Dandenong or Mornington.