A Little Quiz

[Sub MandrelParticipant @submandrel]

Ok, a Mini-Lathe can be bought with either a 16 tpi leadscrew or a 1.5mm pitch

leadscrew.

 

I fit a leadscrew mounted handwheel and want to be able to measure movements to a thou over short distances (<1″). What’s the best combination of leadscrew and index dial?

 

Neil

[Sub MandrelParticipant @submandrel]

One for the grey cells

[Steve GarnettParticipant @stevegarnett62550]

Posted by Stub Mandrel on 12/10/2010 20:02:45:

What’s the best combination of leadscrew and index dial?

 

Best? Neither is particularly good, but at least the 16 tpi thread would give you 62.5 thou/revolution. If you calibrated the dial with 125 steps, each would represent half a thou, and at least you could calibrate right around on this basis, even though you’d have to find some way of indicating how many times you’d been around, and of course you’d reach a half-thou point when you’d been 360 degrees with the wheel.

 

The metric thread works out at 16.93 tpi, and on the basis of trying to measure thou, this is just about a non-starter, giving you 59.113 thou per revolution. 

 

If you really wanted it accurate in thou, then probably the thing to do is to use the topslide rotated at an angle so that sensible turning of its handle through a whole number of divisions per turn moved the tool in by the correct amount. But like the metric thread, insanity lies in this direction (unless you are cutting threads, of course…)

 

So the 16 tpi is the least worst, but the best solution is to use the ‘dial’ on a DRO!

[JasonBModerator @jasonb]

My Emco Ecomat had a 62.5 handwheel marked in 0.0005 increments. worked well on the odd occasion I used it.

 

If you are only working over an inch length then a plunger type dial indicator would be a lot easier and eliminate any backlash

 

Jason

Edited By JasonB on 12/10/2010 20:54:51

[John OlsenParticipant @johnolsen79199]

I think I agree with the DRO answer. This does not have to be a permanent one either, you can come up with a way of mounting a digital caliper, which might be a cheaper solution. The dial indicator would be a good plan B.

 

I did have a feed once with a 13 tpi leadscrew. 

 

regards

John

[Steve GarnettParticipant @stevegarnett62550]

Posted by Stub Mandrel on 12/10/2010 20:02:45:

I fit a leadscrew mounted handwheel …

Of course, there is another answer as well. If you are going to create the handwheel from scratch, do a proper job and gear it (like they did on some toolroom lathes) so that the dial revolutions come out in a whole number of sensible units/rev. Doesn’t matter what leadscrew you use at all, then.

 

This is what Smart and Brown did on the 1024, and… I never use it!

 

– because, of course, it has a DRO.

[Sam StonesParticipant @samstones42903]

 

I bought a mill/drill which had leadscrews with 3mm pitch. That wasn’t too bad except the handwheels were calibrated so that 1mm was equal to 6.66 divisions !!!

 

It was quite useless until I re-calibrated to a more sensible number of divisions.

 

A dial indicator as Jason advocates is a very simple option for short <1″ movement, and isolates backlash etc.

 

Sam

[Steve GarnettParticipant @stevegarnett62550]

Yeah, a bit of a strange quiz really, because the question is wrong! Sorry Stub, but I think that it would be better worded as “Given that both of the leadscrew ratios available on this lathe’s cross-slide would give inconvenient calibration divisions on a directly driven dial, what would be the best way of accurately indicating position in thou’s over an inch or so?”

 

And I agree with Sam and Jason – the dial indicator seems to have a lot going for it in terms of simplicity and a good result in the short term, but if this is going on for any length of time, a DRO is still the way to go.

[Gordon WParticipant @gordonw]

My take is stop thinking in thous. if you have a metric leadscrew, other is make the biggest dial, so the divs. are well spaced. For real accuracy use dial gauge as suggested.

[Sub MandrelParticipant @submandrel]

You are a gippy bunch!

 

I expected some interesting observations and got them!

 

My answer is that for convenience of use a 60 division index on a metric leadscrew is the closest you can get to 1 thou=1 division while not having an inconvenient number of divisions. 125 half-thous are too many and hard to follow and 63 or 62 with the imperial leadscrew are less accurate than the ‘metric thou’. In short all options are ‘wrong’ but the ‘least worst’ could be the metric option.

 

So what have I got? My index is divided into 64 and I tend to use it for fractional distances or to put on small cuts to an accuracy of ‘about a thou’ . Like David said elsewhere, I tend to measure until I get close anyway.

 

Imagine if instead of 25.40 mm = 1.000 inches we had settled on 25.6mm to the inch?

 

My beloved fractions would all convert easily to millimetres:

 

1/2″ = 12.8, 1/4″ = 6.4 down to 1/128″=0.2mm exactly!

 

Even better the ‘metric thou’ of 0.025mm at 1024 to the inch would be very CAD friendly.

 

Alas my idea is borne too late, if only I had thought of it 200 years ago the war of 1812 and all subsequent European wars might have been avoided.

 

Sorry, in a lighthearted mood to see practical engineering getting those Chilean miners to the surface (although many commentators seem to think it was God who bored the hole and lifted them out, not guts and brains).

 

 

Neil

 

[JasonBModerator @jasonb]

Posted by Stub Mandrel on 13/10/2010 19:54:19:

 

 

My answer is that for convenience of use a 60 division index on a metric leadscrew is the closest you can get to 1 thou=1 division

 

Well it may be close for 1 thou but the comulative (sp?) error is quite large.

 

say you want to move your tool 7/8″, using your method it will move 21.875mm (875×0.025) but using the method I suggested or teh dial gauge you will move the correct 0.0875″ or 22.225mm thats 0.35mm or 14 thou error

 

Jason

Edited By JasonB on 13/10/2010 20:11:40

[Steve GarnettParticipant @stevegarnett62550]

I agree with Jason; that’s a bigger error than I’d be happy with.

[adyParticipant @ady]

From les pultra manuel.

 

Most fine work is now executed on metric dimensions.

 

For those engaged on work with english dimensioning of ordinary accuracy it will often be quite satisfactory to regard 1/100mm as 0.0004″ although the correct figure is 0.00039″

[Sub MandrelParticipant @submandrel]

That’s why I said ‘least worst’ not ‘most accurate’

 

Neil

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Do what a real engineer would do and replace the leadscrew .

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

 

Or do what a real engineer would do if the feedscrew could not easily be replaced :

 

Divide the index dial into ANY reasonable number of divisions , work out what
increment each division represents and use visual interpolation to get sub values .

 

For instance with a 16tpi feedscrew chose 25 index divisions each of which
represents two and one half thou and use visual interpolation to get readings of

one thou or less within the main division . This method is universal and was much
used by our ancestors when faced with many and curious feedscrew threads .

 

Note that this method is EXACT – one turn is 1/16th inch and no compromise .

 

Note also that for this specific case four divisions gives ten thou and forty divisions

gives one hundred thou – all easy to use values .

 

You could set up a vernier sub scale to do the interpolation but you will find

that the visual method is MORE accurate and much more facile .

 

[Author]

Posts