Cutting a 5/3 mm pitch thread on a Myford S7 with gearbox

[Adam HarrisParticipant @adamharris13683]

A perhaps only theoretical question, but is it possible to cut a 5/3mm pitch thread (a W20 schaublin/aciera thread) with a Myford Super7 gearbox by selecting off-the-shelf change/tumbler gears? If so, could the responder give the method of calculation please? Many thanks, Adam

[Adam HarrisParticipant @adamharris13683]

[Brian WoodParticipant @brianwood45127]

Hello Adam,,

!.666 mm pitch is very close to 48 DP pitch at 1.661 mm and you can cut that with a 25 T mandrel wheel coupled to a gearbox selection for 16 tpi.

The resulting pitch will be 1.655 mm which is in error by – 0.011 mm [0.0004 inches] that I would imagine to be good enough for the task.

Regards

Brian

[Adam HarrisParticipant @adamharris13683]

Brian, thank you very much! Yes 5/3 mm is 1.667 mm pitch and i agree that 1.655 mm pitch must be close enough and 25T very easy to get (actually i have one already) – can you tell me how you calculated that as i would like to learn ?

[Ady1Participant @ady1]

Gearing explained as simply as possible.

If you use a single simple example as a starting template then the rest of it
can eventually fall into place.
Tables are useful but it would be nice for many users to actually UNDERSTAND
what is going on
Direct gearing is your first port of call.

Golden rule:
Start at the headstock mandrel, and work your way down through the banjo to the
leadscrew.

NON_COMPOUND GEARING
mandrel 20t
middle ones anything you like for a direct drive, not too big IMO, I tend to use
one leg of the banjo and make a vertical 2+3, 40t&55t usually
Leadscrew 63t

63/20=3.15 x 8tpi(leadscrew)=25.2tpi
Which is fine for 1mm metric nuts, 25.4tpi (just don't tell the nut and it'll be
fine)

COMPOUND GEARING (c-gearing)
I use one leg of the banjo for this combination
mandrel 20t (carry forward)

1st-40t/20t (40t is connected to the mandrel 20t)
20/40=0.5 c-gearing
0.5 x 20t(b/f mandrel)=10t equivalent(carry forward to next gear)

2nd-60t/30t (60t is connected to the 20t of first compound gear)
30/60=0.5 c-gearing
0.5 x 10t (brought forward) = 5t equivalent(final "gear" for leadscrew
calculation)

leadscrew 100t
100t/5t=20 x 8tpi = 160 tpi which I find is a nice fine feed.

As long as you use the mandrel->leadscrew (top->bottom) system – until you get
your head around gearing, especially compound gearing – then you won't go far
wrong.

Good luck and happy latheing

=======================

To get the DP, just add 2 to the tooth count and divide by the outside diameter(inches)
of the gear. So, 42 divided by 3.49 is 12

[Roderick JenkinsParticipant @roderickjenkins93242]

That's a great general explanation of change gearing from Ady1. In the particular case of a Myford gear box the intermediate gears between the spindle and the lead screw are essentially fixed. The spindle gear is 24T. If we change this then we alter the tpi selected from the gear box by the ratio of X / 24 where X is the replacement gear. In the solution that Brian has kindly given us we are looking for a 1.667 mm pitch. The closest imperial pitch from the gear box selector is 16 TPI which is 1/16 = 0.0625 inches. Convert this to mm by multiplying by 25.4 and we get 1.588 mm. Changing the spindle gear from 24 to 25 introduces a ratio of 25 / 24=1.0417. Multiply this by the metric equivalent of 16 tpi (1.588) and we get our close approximation of 1.654.

I thought I'd worked out all the useful (and some fairly useless) variations using a spreadsheet and shown in the table below. I've never heard of a 5/3 mm thread before, and a buttress form to boot!

1

Hope this helps,

Rod

Edited By Roderick Jenkins on 23/05/2015 08:41:08

[Adam HarrisParticipant @adamharris13683]

Hi Ady1 and Rod, thank you both very much indeed! That is very helpful. Kind regards, Adam

[Adam HarrisParticipant @adamharris13683]

Rod, I am certainly not the one to explain to you why Schaublin and Aciera went for this external draw bar thread on their "W20" collets! There is quite a lot of info about it on the web and indeed how to cut a buttress thread, but no definitive reason for choosing that particular pitch. If any one knows their decision process for 5/3 mm pitch , I for one would be curious to know it.

[Ady1Participant @ady1]

The compounding formula to try and remember to multiply by is output over input

output/input = gearing ratio

 

Edited By Ady1 on 23/05/2015 10:25:44

[Adam HarrisParticipant @adamharris13683]

Brian/Rod, following Rod's methodolgy I have run up my own spreadsheet calculating backwards the best fit tumbler drive gear for each TPI position, and find that using a 22T tumbler drive gear by selecting TPI 14 gearbox position, I get an error of only 0.21628% which is the closest possible out of all (using a small drive gear of less than 40T) and better than 25T with TPI 16 by a factor of 3. Now to find or make a 22T gear…..

Edited By Adam Harris on 23/05/2015 10:31:21

[Adam HarrisParticipant @adamharris13683]

22T gear from Myford themselves (designed for inside gearbox so may or may not fit tumbler drive gear position) at £32.

[Ady1Participant @ady1]

If any one knows their decision process for 5/3 mm pitch , I for one would be curious to know it.

Only a guess but perhaps it was a move towards metrification (15.24 vs 16tpi)

5/3 is 50/30 teeth or 60/36 teeth on a metric leadscrew so its no biggie

Issues only arise if you approach the thread from an imperial leadscrew perspective

16 tpi is a good "middle thread size" with good strength on an imperial lathe

I have used it a lot myself 60/30 on an 8tpi leadscrew kinda thing when I need to throw something out

Edited By Ady1 on 23/05/2015 10:53:17

[Brian WoodParticipant @brianwood45127]

Hello Adam,

I would be a little wary of using a gearbox 22 T wheell, they are quite narrow at about 1/4 inch wide so that the cone of gears fits in the space available. They are also hardened, not in itself a fault of course, but you will not be able to open the bore to fit the tumbler stud as I am pretty sure the bore is less than 16 mm. The keyway will also be shallow if that is the case.

The 25T gear is a standard size, cheap, and gives an error of only 0.0004 inches [4/10s] I don't believe the collet threads are made to anything like that level of accuracy, the buttress thread is only engaged on one side anyway and total thread engagement is maybe 10 pitches at the very most.

There are other ways of gearing for that pitch as you have found, but I firmly believe the pursuit of accuracy in micron values is grossly misplaced for this application.

If this were a leadscrew thread for positioning I would begin to agree with you, but even tool room lathes have to make compromises with pitch accuracy.

Save your money and buy the cheaper wheel, you will never detect the error in this case

Brian

Edited By Brian Wood on 23/05/2015 13:53:57

Edited By Brian Wood on 23/05/2015 13:54:15

[Adam HarrisParticipant @adamharris13683]

Brian, many thanks – your point is well taken and I will therefore go with the 25T. Kind regards, Adam

[Brian WoodParticipant @brianwood45127]

Hello Adam,

Good luck with making your drawbar. Rod gave a very neat expression for working out the pitch from a [known] starting point.

Last word if I may

I use the word [known] in brackets because in the train between work and cutting tool there are a whole variety of errors involved in producing the result you want. The outside gear pitch errors from the indexing equipment when they were cut, leadscrew error from manufacture, pitch errors on those gears within the gearbox are those you have no control of whatsoever, either in magnitude, direction, nor accumulative effects

Your skill is the only thing you can control and even when you have made your drawbar as well as you are able, the final gremlins are pitch errors in the collets you hope to engage with!!

It is a deep subject, one I have studied in some detail, and one thing I have learned all too clearly is that striving for the ultimate agreement with the stated pitch of a thread is largely futile over short engagement distances.

Commercial threads on stuff like all thread, a rolledl thread material made by the mile, is that pitch error varies within +/- 1% or more of stated pitch and even normal lathe lead screws will build up a 4.5 thou error in 44 inches of travel

Best wishes

Brian

Edited By Brian Wood on 23/05/2015 15:40:54

[Adam HarrisParticipant @adamharris13683]

Many thanks Brian

[Author]

Posts