Mystery Lathe.

[frogesque]

It's a beautiful little lathe in need of some TLC. Overall length is about 8". Possibly a toolmaker or dockie apprentice piece or a part finished kit for the watch/clockmaker/model railway type market. It even has a miniature 4 way tool post complete with turning, facing and parting tools!

Going to fully restore it over winter. This is it sitting on the bed of my 'enormous' 3.5" Myford ML10

 

[frogesque]

Should add, there are no makers marks or personals tramps on it anywhere

 

[frogesque]

The plot thickens!

Removed the feed screw from the longitudinal slide to clean it up a bit and basically determine the thread size with a view to adding a micromer scale. Primary question, is it metric or imperial?

Its imperial - but - the thread appears to be 5/16 x 25 tpi pitch. So far, so good. Quick rummage around the various online thread charts. Nothing! Not BSF, not Metric fine, not UNF nor UNEF. Desperation, is it a BS Cycle thread - nup.

Various searches via Mr Google and I finally come up with references to 5/16 x 25 tpi non standard threads used on Douglas motor cycles in the 30s, allegedly to keep their spares dept in work and keeping non branded parts being used as replacements.

The question now is, was this lathe made by someone (apprentice?) at the Douglas factory at Kingswood, Bristol?

 

[dickm]

Out of interest, what is the nose thread? If it was the same as the old or newer Unimats, that would open up a lot more accessories for you.

 

[frogesque]

Out of interest, what is the nose thread? If it was the same as the old or newer Unimats, that would open up a lot more accessories for you.

Its another problem, there's only one and a bit threads on the nose and I haven't really l looked at it too close. The original bronze bushes have also been robbed out and the attempted replacemts are far too small on the OD. So bad it doesn't even rattle, more of a sulk! It's a simple hollow spindle so it might be easier just to make a new one with a decent nose and a common thread.

I can get a 2 inch dia 3 jaw scroll chuck for £54 - more than I paid for the lathe. Might just keep it as a working ornament with a little test piece between centres run by driving dog off a catch plate. If I do punt it on it would look good in a machine shop reception area.

Bought it off ebay with a cheeky bid and it's small enough to be a nice little renovation project for testing out the Myford. 25 tpi is 0.040 pitch left hand thread. Basic job with 75 - 24 T change wheels and couple of back flip intermediate gears for LH. There's another odd ball thread for the cross slide but I haven't removed it yet to measure up.

Edit:as an afterthought. It would be fun to drive with a small bench mounted steam engine!

 

[frogesque]

OK, checked the nose as best I could. Od is 0.477", sloppy 1/2" and the pitch is somewhere between 0.060 and 0.065".Assuming it's 0.0625 this would equate to 1/2 BSF. Taps and dies readily available for finishing the thread. As it's a remake I'll settle for that and it will be in keeping with the piece.

The cross slide thread is also weird, the end with the handle and a micrometer scale has 30 divisions. Unfortunately it has broken off. Od of screw is 0.180", small 3/16". Bonkers! But should be possible to make a new screw. 18T driver and 75T driven for an 8tpi lead screw. Will need to get an 18T gear but they are available for under £10. Alternatively a 36T wheel and a 1-2 ratio on intermediate gears might be kinder although cutting fine threads doesn't cause much stress on the train.

 

[J-G]

The plot thickens!

Removed the feed screw from the longitudinal slide to clean it up a bit and basically determine the thread size with a view to adding a micromer scale. Primary question, is it metric or imperial?

Its imperial - but - the thread appears to be 5/16 x 25 tpi pitch. So far, so good. Quick rummage around the various online thread charts. Nothing! Not BSF, not Metric fine, not UNF nor UNEF. Desperation, is it a BS Cycle thread - nup.

Various searches via Mr Google and I finally come up with references to 5/16 x 25 tpi non standard threads used on Douglas motor cycles in the 30s, allegedly to keep their spares dept in work and keeping non branded parts being used as replacements.

The question now is, was this lathe made by someone (apprentice?) at the Douglas factory at Kingswood, Bristol?

View attachment 148238

That image shows that it is 24 tpi (you don't count the first crest) and at 5/16 OD that makes it a Standard UNF thread.

 

[J-G]

The cross slide thread is also weird, the end with the handle and a micrometer scale has 30 divisions. Unfortunately it has broken off. Od of screw is 0.180", small 3/16". Bonkers! But should be possible to make a new screw. 18T driver and 75T driven for an 8tpi lead screw. Will need to get an 18T gear but they are available for under £10. Alternatively a 36T wheel and a 1-2 ratio on intermediate gears might be kinder although cutting fine threads doesn't cause much stress on the train.

It does seem somewhat 'Bonkers' but it could be a 2BA which is nominally 0.185" OD with a pitch of 0.81mm which equates to 31.3 tpi - so that doesn't really align with 30 divisions on the dial - unless they were being 'approximate' !!

You talk of getting an 18T gear and I may be speaking without knowledge since I know little about the ML10 but the ML7 range use 20DP gears with a 5/8" bore so the smallest gear that can be made for those (under 'normal' conditions) is 20T. You can get a special 12T but only combined with a 30T as the initial drive gear and as far as I'm aware there isn't an 18/30T combination available. If there were I suspect that the cost wold be well north of £10.

I'm also uncomfortable with your statement that an 18T driving a 75T on an 8TPI lead-screw would give you 30tpi (if that is what you are intimating) - I calculate that you would need a 20T driving a 75T for that, and on an ML7 you would then have a 50T idler; but you could also use the 12/30T driving a 45T via a 65/70/75/80T idler.

[EDIT] I've just done some further calculations and 0.81mm pitch (2BA) divided by 30 divisions equates to just over 1 thou. actually 0.00106299" so maybe not quite as 'bonkers' as first thought.

 

[frogesque]

Senior moment! Recount gives 24 tpi. Will check it for definite with a clock on the advance over a few threads.

25tpi would equate to 0.040 pitch, 40 divivisions on the micrometer dial. 24 tpi is 41 and a bit thou pitch (0.0417" which would give an error of 17 thou over 10 turns) which is odd considering the exact 30 thou pitch on the cross slide. Maybe the maker was aware of the error and lived with it.

 

[frogesque]

J-G

Ah! Not 30 tpi, it's 0.030 pitch. 0.125 pitch on the lead screw/0.030 pitch on the job.

Ratio 125/30 = 250/60 = 25/6. times 3 top and bottom gives 75/18. (could also use 75/36 with a 2/1 iteremediate compound but currently I can't source a 36T). Should note:driver is 18T and the driven 75T with whatever intermediate suits.

The 18T gear was gone by the time I got back to it but I've sourced one made of Tufnol. It will probably only be used once so should be adequate. Normal bore on change wheels for an ML10 is 5/8". If it comes with a smaller bore I will fudge it somehow. Seems they are produced, normally as pairs, for a back flip tumbler to make LH threads.

 

[J-G]

J-G

Ah! Not 30 tpi, it's 0.030 pitch. 0.125 pitch on the lead screw/0.030 pitch on the job.

Ratio 125/30 = 250/60 = 25/6. times 3 top and bottom gives 75/18. (could also use 75/36 with a 2/1 iteremediate compound but currently I can't source a 36T). Should note:driver is 18T and the driven 75T with whatever intermediate suits.

The 18T gear was gone by the time I got back to it but I've sourced one made of Tufnol. It will probably only be used once so should be adequate. Normal bore on change wheels for an ML10 is 5/8". If it comes with a smaller bore I will fudge it somehow. Seems they are produced, normally as pairs, for a back flip tumbler to make LH threads.

I assumed that I'd got something wrong but couldn't really fathom why. Your reference to 2:1 ratio confused me somewhat but I now consider that you are referring to the idler gear ratio which actually means nothing. The idler can be any convenient size that will link the driver to the driven and fit on the banjo.

Now, an 18T gear at 20DP is 1" OD and the Root is 0.775"; couple that with the fact that the drive keyway needs to be 0.076" deep and I'm sure you will appreciate that there is no way that the gear can stay in one piece with a 5/8" bore. Here is a drawing which proves the point.

There's no doubt that you may be able to find some way to mangle a connection to the 30T gear (with a smaller bore) at the top of the train, but it would be quite a challenge. It would be a much easier option to use a 24T driving a 100T which will still give you a 0.030 feed.

Incidentally, there is nothing untoward about Tufnol, I have made many gears using it and never had a failure in use.

 

[frogesque]

Thanks so much for taking the time to look into this. Looking at the photo on Evilbay, the Tufnol gear is plain with no keyway. As tumble gears they are only used to reverse direction for LH threads


Originally Myford used small shear pins, think 1/16 dia to transmit power through the train and I can easily drill a small hole to accommodate this. The few gears that came with it have this arrangement. The more modern ones I've sourced have the keyway. The full gear set for the ML10 was 2 x20T, 25 through to 80T by 5 step intervals. I've built up a full set + some odd ones but don't have a 100T, 125T or 127T (for metric). They start to get expensive at that size.

I've a rotory table on which I made a 100 hole division plate that can fit at the back of the chuck so conceivably could make one but that's for a different day. Stand by for a home made shadow graph!

As an aside, reasonably cheap 36T circular saw blades are available. (Wikes £13) Very simple job to make a ratchet to suit for indexing. Maybe need to look out those old brass lever door handles and melt them down for a gear blank! Old cast brass plumbing fittings are also available.

 

[J-G]

Thanks so much for taking the time to look into this.

You're welcome - I like to excercise my brain with anything Gear or Thread related

Looking at the photo on Evilbay, the Tufnol gear is plain with no keyway. As tumble gears they are only used to reverse direction for LH threads

Not quite correct - the Tumbler Gears are to facilitate reversing the direction of the lead-screw. You may have to use it for left-hand threads but it will depend upon how many gears are in the train.

Originally Myford used small shear pins, think 1/16 dia to transmit power through the train and I can easily drill a small hole to accommodate this.

Measuring the image you attached I calculate that the drive pin hole is more likely to be 1/8"Ø - though if you can drill a 1/16" or even a 1.5mm hole with the 30T and the 18T in alignment then it would probably be OK for a short run.

Here's the same drawing with a 1/16" hole in the best position - it still weakens the specific tooth but you'd probably get away with it.

The few gears that came with it have this arrangement. The more modern ones I've sourced have the keyway. The full gear set for the ML10 was 2 x20T, 25 through to 80T by 5 step intervals. I've built up a full set + some odd ones but don't have a 100T, 125T or 127T (for metric). They start to get expensive at that size.

I bought a 127T for £15 - though that was about 8 years ago, before I bought my Hobber.

I do have a 100T - but only one - which I made from Alum and I could consider making available as a loan but we'll have to take this discussion to PM to sort out details. I could also make another from Tufnol, though I don't have a sheet of 10mm thick on the shelf at the moment.

I've a rotory table on which I made a 100 hole division plate that can fit at the back of the chuck so conceivably could make one but that's for a different day. Stand by for a home made shadow graph!

Hmm... I think we do need to start a PM - I've recently written a windows program to create Shodowgraph images of Involute gear-teeth accurately scaled both on screen and output to a printer. It does require Screen & Printer calibration - which is built in.

Another program (unfinished but operational) takes a TPI entry, calculates the change gears required and shows them in position on the ML7 head-stock. It currently only handles simple trains, Complex trains are planned but time is an enemy!

 

[frogesque]

Thanks again!

Before the topic drift gets too far, the shear pin holes don't go right through the gear body which would strengthening the tooth you indicate.

I may also have a solution to the 100T wheel but thanks for your offer.

Regarding a shadowgraph. To keep it simple with no mirrors or prisim, a 35mm lens with the job back-lit at the focal plane will project a 10 times image at roughly 350mm. Altenatively in a well blacked out room 100 times at a distance of 3.5m (11.5ft). Good quality threaded Pentax K lenses are readily available and cheap as chips because they are no use for modern cameras. Even with adaptors (Nikon or Canon etc) the camera won't even know its got a lens attached. Tried it and it works, just needs some work for a stand and focusing arrangement and fixtures for work holding.

Getting back on topic to the wee lathe, I've taken the headstock apart to look at bearings etc. Think it will be a lot easier just to remake the spindle and bearings to suit. Currently the end play nut and spindle thread are 7/16 x 32 tpi (a standard ME thread) and the nose, what's left of it, is 1/2 BSF.

Use stainless steel for the spindle or hardened an tempered tool steel. The existing one is pretty soft, probably annealed high carbon.

 

[J-G]

Regarding a shadowgraph. To keep it simple with no mirrors or prisim, a 35mm lens with the job back-lit at the focal plane will project a 10 times image at roughly 350mm. Altenatively in a well blacked out room 100 times at a distance of 3.5m (11.5ft). Good quality threaded Pentax K lenses are readily available and cheap as chips because they are no use for modern cameras. Even with adaptors (Nikon or Canon etc) the camera won't even know its got a lens attached. Tried it and it works, just needs some work for a stand and focusing arrangement and fixtures for work holding.

The physical creation of the Shadow is no real problem - and of course my program doesn't do that, it creates the true involute outline for any tooth count and DP / MOD with 14½°, 20° or 30° PA, against which the shadow is compared.

I used to create these manually at 50:1 when I was an apprentice and each one took about a day! My program creates any magnification that will fit on an A4 sheet at the click of a mouse (plus the time to output to the printer!)

 

[frogesque]

The physical creation of the Shadow is no real problem - and of course my program doesn't do that, it creates the true involute outline for any tooth count and DP / MOD with 14½°, 20° or 30° PA, against which the shadow is compared.

I used to create these manually at 50:1 when I was an apprentice and each one took about a day! My program creates any magnification that will fit on an A4 sheet at the click of a mouse (plus the time to output to the printer!)

I come at this more from a clockmaking and resoration angle. Hand built antique clocks usually use a cycloid profile, rather than involute. Generally it was a case of repairing the wheels or duplicating them. Hand filed form tools in either silver steel or gauge plate, hardendend and tempered then used to cut the brass to an existing profile that might or probably not be very exact. Likewise screws.made with a screw plate every maker had their own and there was no standard as such.

Matching teeth to a theoretical profile via a shadow graph sound good! My idea would be to copy the profile off say a 35T wheel to make a 36T. Might be noisy but the ultimate test is, "If it works, it's good enough!" that said, I do realise the profile is more critical the less teeth you have. My dentist would probably agree

 

[dickm]

Good quality threaded Pentax K lenses are readily available and cheap as chips because they are no use for modern cameras.

Being a flinty souled pedant, will point out that Pentax K lenses are bayonet fitting. The old 42mm threaded ones are usually called Pentax screw or thread. And they aren't "no use"; just occasionally they can solve a problem even on digital Pentaxes. Says the guy who still has an MX.

 

[frogesque]

Being a flinty souled pedant, will point out that Pentax K lenses are bayonet fitting. The old 42mm threaded ones are usually called Pentax screw or thread. And they aren't "no use"; just occasionally they can solve a problem even on digital Pentaxes. Says the guy who still has an MX.

Oops! Yeah, it was the 42mm screw on lenses I was meaning. I used to do some ultra close up photography with 42mm screw lenses and extension tubes. + adaptor. Worked well with my old Canon Rebel but my newer Canon 750d has to talk to the lens so it won't work., just throws up an error.

As far as glass goes, some of the old lenses were exceptional quality and far cheaper than modern glass.

 

[J-G]

I come at this more from a clockmaking and resoration angle. Hand built antique clocks usually use a cycloid profile, rather than involute. Generally it was a case of repairing the wheels or duplicating them. Hand filed form tools in either silver steel or gauge plate, hardendend and tempered then used to cut the brass to an existing profile that might or probably not be very exact. Likewise screws.made with a screw plate every maker had their own and there was no standard as such.

As it happens I am also 'into' clocks and was gifted a very rare beast - a 0.5MOD Cyclodal form Hob.

Matching teeth to a theoretical profile via a shadow graph sound good! My idea would be to copy the profile off say a 35T wheel to make a 36T. Might be noisy but the ultimate test is, "If it works, it's good enough!" that said, I do realise the profile is more critical the less teeth you have. My dentist would probably agree

Hmm... the difference between the profiles of 35T & 36T - 20Dp, 20°PA Involute form gears is negligible. Here's the output from my program with a transparent 35T image overlaid - but offset - onto the 36T image

As you can see, this is at 40:1 and I suggest that if you can decern any difference then your eyes are vastly superior to mine - the only 'tell-tale' is the dotted line showing the location of the 'Base Circle'.

I haven't done the same comparison between (say) 10T and 11T but assume that the difference would be conciderable.

This just shows why it is possible to cut all tooth counts with only 8 different gear cutters for each DP-MOD/PA - though I use hobs so only need one for each.

 

[frogesque]

As it happens I am also 'into' clocks and was gifted a very rare beast - a 0.5MOD Cyclodal form Hob.

Hmm... the difference between the profiles of 35T & 36T - 20Dp, 20°PA Involute form gears is negligible. Here's the output from my program with a transparent 35T image overlaid - but offset - onto the 36T image
View attachment 148484
As you can see, this is at 40:1 and I suggest that if you can decern any difference then your eyes are vastly superior to mine - the only 'tell-tale' is the dotted line showing the location of the 'Base Circle'.

I haven't done the same comparison between (say) 10T and 11T but assume that the difference would be conciderable.

This just shows why it is possible to cut all tooth counts with only 8 different gear cutters for each DP-MOD/PA - though I use hobs so only need one for each.

That is one superb program you have written! We are not worthy!

Not had much chance to do much work on things the last few days but I have managed to source a 100T change wheel, £15 + p&p. Pretty reasonable considering a Mehanite blank would probably cost as much.

The lathe does appear to have had 2 different folk working on it. The tailstock has features that are different to the main part and I suspect a later modification or addition. My main job now will be to make a thread dial indicator for the Myford to assist in cutting the quirky threads.