Imperial Thread Cutting on Metric Mini Lathe

[Howard LewisParticipant @howardlewis46836]

[Howard LewisParticipant @howardlewis46836]

Armed with a 63T gear, are there any tables available, please, giving the Change wheel sets to cut Imperial thread pitches on a Mini lathe with a 1.5mm pitch leadscrew?

(Am too idle to spend days making an Excel spreadsheet, and hope that someone has been there already)

Howard

[Thor 🇳🇴Participant @thor]

Hi Howard,

LMS has ha change gear calculator for the Mini Lathes,have you tried it? You can choose between a metric or imperial leadscrew and the same for the thread you want to cut.

Thor

[Howard LewisParticipant @howardlewis46836]

Hi Thor!

Very Many Thanks.

It looks to work a treat, based on the couple of checks made so far. Will save me HOURS of suck it and see on Excel.

Filed away for future reference!

Kind Regards

Howard

[Brian WoodParticipant @brianwood45127]

Hello Lewis,

​Buy a copy of my book " Gearing of lathes for Screwcutting" ArcEuro are selling them at a discount.

​There are tables in the chapter for mini lathes with both 16 tpi, 1.5 mm and 2 mm leadscrews covering all the usual metric/imperial conversions—BA pitches—-DP and Module. Fill your boots

Regards
​Brian
 

Edited By Brian Wood on 03/08/2017 14:57:07

[Neil WyattModerator @neilwyatt]

Hi Howard,

Check your inbox

Brian's book is still recommended if you want to really explore the subject.

Neil

[Andrew TinsleyParticipant @andrewtinsley63637]

I second that! Brian's book is superbly produced and crammed with screwcutting information on a whole host of lathes.

It really is useful and would also make a good coffee table book for a model engineer! Do yourself a favour and buy one from Arc right away, you won't regret it.

Andrew.

[Bob GitshamParticipant @bobgitsham90361]

It is my uinderstanding that to cut a metric thread on an imperial lathe or vice versa, you need a127 x 120 tooth gear, as the gender bender gear in the train,

The other gears will then give you the pitch you require. there are other combinations that are used but they only give an approximation of the desired thread pitch.

[John MCParticipant @johnmc39344]

My second lathe of many years ago was a round bed Drummond, the almost complete set of change wheels for that had a 63T gear for metric. No chart but worked out the trains soon enough thanks to the subject being taught at technical college, I wonder if its taught now? Interestingly the error in pitch using the 63T rather than a 127T gear was minimal, way smaller numbers than the accuracy the lathe was capable of.

John

[Roderick JenkinsParticipant @roderickjenkins93242]

While Bob's comment is true enough in a purely mathematical sense, the "approximations" will give a result to more accuracy than the lathe is physically capable of producing. They also have the advantage of not requiring a 127 tooth gear or, indeed, finding a way to mount it on small lathes.

Cheers,

Rod

[Bob GitshamParticipant @bobgitsham90361]

Thanks Rod, I hadn't realised that the error was actually that small. I made sure that when I ordered my new lathe LD1216 last year that it came with the 127 x 120 compound gear, and although it is nominally an imperial lathe, Imperial leadscrew etc., all the handwheels are calibrated in both imp and metric, and the screw cutting chart gives both imp and metric threads straighrt from the norton style box., with only haveing to set the compound gear for Imp or metric.

A few months ago I did see a very nice spreadsheet on the internet that gave all the approximations that could be used, It actually listed them with thev actualerror as well. I had a look for it last night but couldn't find it. It had a built in calculator where could list the gears you had plus leadscrew pitch etc and it would calculate your options.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Bob Gitsham on 30/08/2017 11:44:05:

It is my uinderstanding that to cut a metric thread on an imperial lathe or vice versa, you need a127 x 120 tooth gear, as the gender bender gear in the train,

The other gears will then give you the pitch you require. there are other combinations that are used but they only give an approximation of the desired thread pitch.

That's the conventional wisdom Bob, but it's not rock-solid. The magic 127, or slightly less magic 63, toothed gears allow one to calculate a TPI and then convert it to metric, or vice versa. The logic is analogous to 'multiply thou by 25.4 to get mm'.

Calculating ratios with paper and pencil using the method mentioned by John MC is hard work. Having 127 or 63 makes the sums easier. However, we don't have to do it the hard way – a computer can crunch the numbers in nanoseconds and tell you how big the error is. Only if the error was too big would I get a 63 tooth gear. (I would never buy a 127 tooth gear because fitting on the banjo is likely to be problematic – it's too big for comfort!)

Whilst the conversion method might suit the way we do arithmetic, and it gets results, it may not be the only way of producing an acceptable thread. My metric lathe produces more imperial threads than the imperial version and it doesn't have either a 63 or a 127 toothed wheel. What it does have is a set of gears that can be combined in ways that produce thousands of different thread possibilities. Many of them are duplicates or non-standard, but quite a few produce reasonable approximations of imperial TPIs. A quirk of the mathematics means that the gears provided with the imperial model are not quite so productive as it's metric equivalent. Nonetheless, given a calculator like that suggested by Thor, it's well worth checking to see if your existing gear set will do the job without needing a special 63 toothed gear.

Dave

[Martin DowingParticipant @martindowing58466]

63 tooth gear gives rather poor accuracy in metric/imperial translations.

There are far better combinations than that with standard gears of 20-21-25-30…75 teeth.

For superb accuracy you will also need 38 and 42T.

127T gear even if perfect is rather impractical (too large).

Read "Screwcutting in the Lathe" by Martin Cleeve.

You will find nearly everything you need to know (and more) there.

[Roderick JenkinsParticipant @roderickjenkins93242]

Posted by Martin Dowing on 04/09/2017 20:28:55:

63 tooth gear gives rather poor accuracy in metric/imperial translations.

Depends on whether you are using it as a driver or driven.

**LINK**

Rod

[HopperParticipant @hopper]

I turn metric threads on my imperial lathe without any special gears. Using Martin Cleeves' book Screwcutting on the Lathe. He has the charts to use the original gears set-up as compound gearing, giving an accuracy usually in the range of 1 in 3,000 or so. Close enough for my shed. The book has the charts for imperial on metric also.

[Howard LewisParticipant @howardlewis46836]

Thanks for all the replies.

Have down loaded a chart, giving the combinations and the errors.

As already said, the ideal solution is 127/120 compound gear., (as on the big lathe) Might have a job fitting it in, but will ponder it!

Howard

[Neil WyattModerator @neilwyatt]

Posted by Martin Dowing on 04/09/2017 20:28:55:

63 tooth gear gives rather poor accuracy in metric/imperial translations.

What? 1 part in 16,000 not good enough for you?

That's about 1/3 of a wavelength of red light in a millimeter or the length of a cricket pitch in 100 miles…

Neil

Edited By Neil Wyatt on 07/09/2017 19:53:07

[Bob GitshamParticipant @bobgitsham90361]

How is this figure calculated? It would seem to me that as 63 is half of 126 that is 1 full tooth per revolution, short of 127, and given that the 63 tooth gear is run with a 60 T gear it is an attempt at approximating 127 x 120 gear setup which is, and has been, used on larger lathes for a very long time.

Therefore the error occurs to me to simply be 1/127th. Ok, not a lot, and particularly in short lengths of thread such as the average nut and bolt., where the thread engagement lengthy is typically about one diameter. But it would become a problem in any attempts to make a long running thread such as a lead screw. Sure the mating nut could be made to match, but the error would be significant in the total length travelled over many turns.

Bob.

[John OlsenParticipant @johnolsen79199]

You don't actually use the 63 tooth in the same way as the 127, so the error is much less than your estimate. Check out some of the tables referred to by others above, the errors are quite acceptable for most purposes.

I have a 127 tooth gear to fit the Myford, but have never used it since it would need a different banjo and a slot cut in the stand to actually get it on. It is about the size of a bread and butter plate. Looks good hanging on the wall though.

John

[Michael GilliganParticipant @michaelgilligan61133]

Bob,

Our esteemed Editor's article on the subject is available for download: **LINK**

http://www.arceurotrade.co.uk/reviews/MEW204-P31_P36-Thread-Cutting.pdf

… this should reveal the magic of 63

MichaelG.

[Bob GitshamParticipant @bobgitsham90361]

Posted by Martin Dowing on 04/09/2017 20:28:55:

63 tooth gear gives rather poor accuracy in metric/imperial translations.

There are far better combinations than that with standard gears of 20-21-25-30…75 teeth.

For superb accuracy you will also need 38 and 42T.

127T gear even if perfect is rather impractical (too large).

Read "Screwcutting in the Lathe" by Martin Cleeve.

You will find nearly everything you need to know (and more) there.

It's only too large on small machines, most medium to large machines come with the 127×120 combo.

My LD1216 from Taiwan came with this set, plus a variety of other gears, and the built in Norton style gearbox. it also ha a thread chart for use with the Norton box that covers almost all imp. and metric threads without even using half the gears supplied.

So, I guess even more strange threads could be achieved with them. I'm not sure I'll ever need them, But maybe one day I'll see what can be done.

Bob.

[Neil WyattModerator @neilwyatt]

Brief example cutting 1mm pitch on a 1/16" leadscrew mini lathe:

1/16 * 25.4 = 1.5875mm pitch (exactly).

So to step this down to 1mm pitch you need to divide the gear train by 1/1.5875

1/1.5875=0.62992125984 (approximately)

So if you use the ratio 63:100 in the gear train the error is:

63-0.62992 = 0.00007874

As a percentage that's 0.008% which is close enough for rough work.

Visiting this site I see their precision metric leadscrews have an accuracy of 0.08mm per 300mm.

That's 0.026%. Which is three times the error given by using a 63 tooth gear – over a foot!

Any questions?

Neil

[Neil WyattModerator @neilwyatt]

Posted by John Olsen on 08/09/2017 07:08:26:

the errors are quite acceptable for most purposes.

Indeed, such as making precision leadscrews – see above

Sorry John, I know you are on my side!

Neil

[Anonymous]

Posted by Neil Wyatt on 08/09/2017 09:47:50:

Visiting this site I see their precision metric leadscrews have an accuracy of 0.08mm per 300mm.

Hmmm, and further down on the same page they say the lead error over the length of the rod is 0.3mm per 300mm, rather different. I would expect precision leadscrews to be ground rather than rolled.

Of course that doesn't change the fact that a 63 tooth gear is perfectly satisfactory for cutting metric pitches with the correct gear train. Personally I say hooray for quick change gearboxes; imperial, metric, DP and module all at the change of some levers, and one gear.

Andrew

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Andrew Johnston on 08/09/2017 10:05:02:

Posted by Neil Wyatt on 08/09/2017 09:47:50:

Visiting this site I see their precision metric leadscrews have an accuracy of 0.08mm per 300mm.

Hmmm, and further down on the same page they say the lead error over the length of the rod is 0.3mm per 300mm, rather different. I would expect precision leadscrews to be ground rather than rolled.

< etc. >

.

+1

To quote Jason's recent comment on the Soba thread:

"Problem is one mans Shoddy is another mans will do the job perfectly well."

The situation seems similar with 'Precision Leadscrews'

… or indeed anything "Precision"

MichaelG.

Edited By Michael Gilligan on 08/09/2017 10:50:17

[Author]

Posts