Plain lathe gearing for ‘bastard’ metric thread

[Oily RagParticipant @oilyrag]

Hi All,

Just need a confirmation from the members here that what I have geared the plain lathe drop gears will produce the thread I expect.

A little background is that the lathe is a Raglan Little John. This has an 8 tpi leadscrew and in backgear this equates to a 16 tpi leadscrew. for those that do not know the Little John the drop gears are arranged as A/B x X/Y x C/D. A/B gears are fixed centres and must add to 90, likewise X/Y are also fixed centres and must add to 96, whilst C/D can be a floating arrangement able to accommodate a fairly wide range of gear combinations.

The X/Y pairing are listed as 48 x 48 for English threads and 52 x 44 for metric threads.

I am looking to cut a 1.66 p Schaublin buttress thread for a tool holder for my mill. Previously I have used a set up with a 51 tooth gear but for some reason the workshop gremlins have hidden it. In desperation I sat down and tried to formulate the gearing using standard gears. I came up with the following:-

A = 45, B = 45, ~ C = 45, D = 40, ~ X = 44, Y = 52 – in each case it is Driver/Driven. with compounding of B~X and Y~C

this gives the gear ratios as A/B = 1.0; C/D = 1.125; X/Y = 0.846154

this gives 15.2307 tpi, or 1.6676p

I arrived at this combination by juggling the gear set up for 18 tpi which was given as A/B = 45/45, C/D = 32/36 and with the X/Y meshing pair set for imperial at 48/48. by using the metric tooth counts at X/Y (52/44) BUT by reversing them I got the reduction necessary to give the final tpi/pitch required. Note that for reasons outside my control the 36/32 C/D pairing fouls the gear boss so I increased the tooth count to 45/40 to maintain the ratiometrics (9/8 as an improper fraction)

Will it work or have I missed something basic?

[Oily RagParticipant @oilyrag]

[Nigel Graham 2Participant @nigelgraham2]

Well, assuming you've verified the calculations, and the change-wheels they give all mesh properly, you could always take a trial cut on a bit of spare material before risking the part itself, if that's otherwise at an advanced stage.

I keep making that wheel ratio 0.951923 though – but that times 16 does give 15.2308 (actually 15.230768), so unless I keep mis-reading the tooth-counts I think you might have mis-copied the X/Y answer to here.

I am not familiar with the Raglan lathes, so did not know the back-gear also halves the lead-screw's effective lead. I think that's rather unusual!

[DC31kParticipant @dc31k]

As an alternative, instead of thinking of the Schaublin pitch in imperial or in decimal millimetres, think of it as a fraction: 5/3 x 1 or 5/6 x 2.

The lathe surely will have published information for 1mm and 2mm pitches. Start with these and then factor the given gears by the appropriate ratio.

Are you aware of the Raglan manual and the formula for metric screwcutting at:

https://www.oldmansshop.com/raglan/rag-man-p17.php ?

It gives metric pitch = A/B x C/D x 15/8. Plug your required 5/3 into this and it will give you A/B x C/D = 8/9 so a 45:45 on A/B and 40/45 on C/D would do very nicely as you say.

[not done it yetParticipant @notdoneityet]

Back gear will make no difference to the lead screw pitch. Lead screw turns from the spindle. Thread pitch will be the same whether in normal drive or back gear.

[DC31kParticipant @dc31k]

Posted by not done it yet on 21/08/2020 07:37:52:

Back gear will make no difference to the lead screw pitch. Lead screw turns from the spindle. Thread pitch will be the same whether in normal drive or back gear.

In general you are correct. But for a Little John, I do not think so.

Kindly go and look at the manual mentioned above and assess the situation.

[Martin ConnellyParticipant @martinconnelly55370]

Little John manual

This is the link to the manual mentioned above. I have looked through it and it seems to have a typical back gear arrangement that has no bearing on screwcutting pitch. It does state that there is a permanent 2:1 reduction ratio from the spindle giving an effective tpi of 16 to the leadscrew.

Martin C

[HopperParticipant @hopper]

Easy way to check gear train is to rotate the chuck exactly 1 revolution and measure carriage movement with a dial indicator to see it matches desired pitch.

[not done it yetParticipant @notdoneityet]

I’ve got a Raglan 5” and used to use a LJ. Thread cutting is the same in both speed ranges. Both mine have been QCGB models. Far easier to dial in obscure pitches with those.🙂

[Brian WoodParticipant @brianwood45127]

If it is of any help, the basic change wheel gearing for any generic 8 tpi leadscrew lathe [say Myford for example] is a straight 23/44 reduction. The pitch outcome is 1.6597 mm

That might fit on a Raglan as C=23 and D=44

Regards Brian

[Oily RagParticipant @oilyrag]

Thank you everyone who responded.

The confusing part about the leadscrew pitch is that it gives the formulae and finalises the tpi by a multiplication of 16, the tumbler gear assembly does give a 2:1 ratio as mentioned by Martin Connelly; this rather than my incorrect assumption that it was achieved via the back gear, which I accept increases torque available by slowing the spindle speed rather than affecting anything else (a Simpson moment! Doh!!).

As the job is mounted in a 4 jaw and not wishing to break it all down, having taken all bar the finishing cuts to clean, up I was loathe to break it down for a trial cut. Then it struck me that I could mount a 'Sharpie' fine point felt pen on the toolpost and 'scribe' a trial thread on the component. Sure enough the thread was wrong – I had transposed the C/D gear wheels and managed to get a finer pitch thread! Getting them the right way around gave the correct pitch.

Thank you again everyone who responded – it helps to be able to discuss issues like this when you are a one man in a shed with only those who know how to bake delicious cakes in the household!

[Howard LewisParticipant @howardlewis46836]

A 16 tpi Leadscrew has a pitch of 1.5875 mm.

The ratio between that and your desired 1.66 mm is 1.0456696, so you need to find a gear train which provides that ratio as opposed to one giving a 1:1.

A 22:23 gives 1.0454545, so the error is in the 4th place of decimals. Will That matter? The problem is likely to be finding 22T and 23T gears, unless you have the means of (or a friend ) to cut them.

If you have a 38T gear , meshed with a 55, it will give a 1.4473684 ratio.

You may need to insert an Idler, or use tumbler reverse, to ensure that the resulting thread is right hand!

HTH

Howard

[Brian WoodParticipant @brianwood45127]

Hello Oily Rag,

I have used pencils and even chalk in the past to 'prove' the gearing for a thread pitch before cutting metal. It is an old trick really but jolly useful.

I'm glad your result was as you had hoped although I must confess I couldn't follow the logic. Success in the end which is what mattered to you

Regards Brian

[Nigel Graham 2Participant @nigelgraham2]

Oooh, that trick of using a pen or pencil to prove the gearing might be old but it's new to me, so thank you Oily Rag and Brian!

Re Hopper: I use several, even 10, complete spindle turns to give a much larger advance, hence more readily measured accurately.

[not done it yetParticipant @notdoneityet]

While the ‘sharpie trick‘ is pertinent, an easy enough alternative could have been to dismount the chuck and replace with another for any trial threading operation(s).

Refitting the chuck should present no problem with registration of the chuck and no ‘breakdown of the workpiece from the chuck’ would occur.

Most modellers do have a 3 jaw chuck available.

[Michael GilliganParticipant @michaelgilligan61133]

The job is done, and Oily Rag is content, so I mention this only for ‘academic’ interest:

The Schaublin buttress thread may not be exactly 1.66mm pitch anyway [*]

Some useful links in this previous thread: **LINK**

https://www.model-engineer.co.uk/forums/postings.asp?th=123693

MichaelG.

.

[*] the truth may be out there, somewhere …

…

[Author]

Posts