Myford Super7 Leadscrew Accuracy problem

[nick ClarkParticipant @nickclark51930]

Hi to all,

I am having trouble with the rated accuracy of my Super 7 leadscrew (8TPI) as im getting weird figures on the clock.

I have the changwheel type of lathe but I have disconnected all of the gears so its just the leadscrew under manual control (using the leadscrew handwheel).

I am getting figures of 5 thou out over a 1″ dial guage travel.

If this figure were to translate into cutting a long screw ie a foot long then the pitch would be seriously affected.

Can anyone confirm their measurements of the 8TPI leadscrew accuracy using just a clock and turning the leadscrew by handwheel grads.

Any info as to what other people find would be very helpfull.

I have inspected the clasp nuts and they seem to be ok by eye.?

Regards to all

Nick

[nick ClarkParticipant @nickclark51930]

[mgjParticipant @mgj]

Plus or minus? – the error.

 

How old is the leadscrew? Is there a difference between near the headstock and nearer the tailstock?

 

How often are you going to cut a thread a foot long?

 

When you say the clasp nuts are OK, you have physically checked they are not full of swarf, and have adjusted their gibs so you are sure they are engaging fully and without backlash?

 

You have a drag on the carriage, and are not allowing any backlash to alter your reading? You have adjusted out any play at the tailstock end of the leadscrew?

 

What is the error on your clock? Even on Mr Mitutoyos very expensive ones he doesn’t guarantee it beyond a thou – actually against the digital its pretty well spot on, but he doesn’t guarantee it on its spec, so the clock could be out.

[nick ClarkParticipant @nickclark51930]

Hi,

I have done a new test this morning and have checked the clock and for a known movement (on my Mill) it is giving the required reading so I am at the moment not suspecting the clock?. The Myford carriage moves freely and adjusted gibs so no play (bed just reground by Myfords)

The test I have done involves placing the clock on the bed and whilst turning the leadscrew in one direction I expect a movement of 125 thou(1 turn of handwheel).

I got readings like this as progressing along the travel

0.126″

0.125″

0.128″

0.128″

0.127″

0.123″

I am wondering if its the leadscrew worn in places ie the helix angle is changing due to wear causing the plus minus error?

I have been told that myford leadscrews have an accuracy of 2thou / foot so I am getting some pretty bad readings if thats the case.

How I first came to find the problem was my friend cut a 20tpi screw on a myford (super7 box type) and my Super7 changewheel type is finding errors as I run along due to my feed/rev being wrong. If i set for a feed of say 0.500″ I get 0.490″?

This is how I came to suspecting the leadscrew? or nut? as the leadscrew seems to be out.

Any Ideas or further checks would be gratefully received

Best Regards to all.

Nick

[KWILParticipant @kwil]

I have just checked my Myford Super 7 Gearbox PCF type over the full bed length. Carriage is snugly set up for easy movement.  I have needle thrust washers on the leadscrew fixed bracket, set for minimum backlash. Lathe is fitted with Newall Industrial DRO.

 

Handwheel turned in same direction to eliminate backlash with half nuts set to close almost tight, ie half nut stop screw set correctly, handwheel zero point approached slowly to enable more precise setting to cursor pointer. More readings than not were readings spot on,  deviations were 0.0002″ negative and on one occasion 0.0004″ negative. Needless to say I do not have screw cutting problems.

[JasonBModerator @jasonb]

If you actually average out those figures then you get 126.16 so over an inch you are nearer 0.0015 out not the 0.005 in your first post.

 

Jason

[CirclipParticipant @circlip]

Are you actually going to use the lathe to make things Nick or is it just to find how accurate you think it NEEDS to be?? If you’re doing work for NASA, I’m surprised you are being allowed to use an old Myford instead of a Yamazaki ball screw CNC machine.

 

  How does it map over the whole length of the leadscrew??

 

  Regards  Ian.

[KWILParticipant @kwil]

Nick,

 

Read your Messages

 

K

[mgjParticipant @mgj]

Well I am getting similar results to Kwil and my leadscrew is 24 years old. i’m measuring on agood new mitutoyo clock. Like Kwil I have the hemingway roller thrust kit in the fixed end of the leadscrew.

 

I would dig out the feed nut(s) and degunge them, and then fit and adjust properly. I restored an old 7 like that. A lot of brass and iron dust had built up in the nut threads, so it was only engaging on the tips – and had worn there too. Once they were back to full engagement all the errors disappeared. Not as good a fix as a new clasp nut, but that is what the owner wanted.

 

You can take a pair of calipers and measure the feed screw at different places along its length. Over 3″ or so such errors will show.

 

After that I’d REALLY REALLY check the adjustment of the saddle to make sure there is no lost movement. ie if the saddle does move (I know you have adjusted it), you can easily pick up a free thou or two depending on exactly where the clock is placed. What happens when you bore a hole – do you get a lot of extra cut on the powered outward leg. If so you might consider that saddle adjustment across the shears.

 

If none of those worked I’d ring Myfords, but probably its the nut. They are built so they wear faster than the screw, being easier and relatively cheap to replace. Also if someone has had a big jam up it may have got warped or strained and its jumping forwards a touch within the thread – a possibility in a lathe that has needed a regrind.

 

 

Edited By meyrick griffith-jones on 29/12/2009 18:22:23

[Les Jones 1Participant @lesjones1]

One way the pitch over several turns could be measured would be to use a digital caliper  with one jaw clamped in the toolpost and the other jaw (Or end of the depth gauge.) clamped to the headstock or tailstock. Care would have to be taken that the caliper beam was parallel to the axis of the lathe. This could be set up using a dial gauge.

Les.

[KWILParticipant @kwil]

The real problem in using a Digital Caliper is that you do NOT know their accuracy, they only tell you the resolution, ie it appears to read in 0.001″ etc  I have yet to see any accuracy  figures stated for these devices.

[RichmondParticipant @richmond]

Accuracy repeatability are issues.

 

I beleive that in fact for repeatability ( as in open and close the scale to the same point repeatedly ) it is something like 0.1mm whereas resolution is 0.01mm.

 

This info is rarely mentioned  in the sales blurb for ANY capacitive scale…… which in essence is the standard digital caliper.

 

I have no clear info on accuracy, but an extension is that infact a measument could deviate by upto 0.1mm from norm….. so in theory not that accurate at all.

[JasonBModerator @jasonb]

Mitutoyo give the accuracy of their digital mics and callipers

 

A Calliper will typically be 0.02mm over 200m, 0.03mm over 300mm, thats just over 0.001″ per foot.

 

Their mics are 1 micron which is 0.001mm or  0.00005″

 

Mitutoyo site if you want a look.

 

Don’t know what a cheap imported digital calliper would be though

 

Jason

[KWILParticipant @kwil]

I agree with Richmond, the common digital calipers are only useful for an indication. The internal scale is only a printed circuit board.  Checking metal in a rack when you cannot get close in, they might tell you whether it is 10mm or 0.375″ [3/8″]..  Even the Mitutoyo at 0.02mm per 200mm (not 200m Jason!) is a whole 0.0008″ out. nowhere near a micrometer standard.

[nick ClarkParticipant @nickclark51930]

Hi,

Thanks for all your replies on this problem, I could not see what was wrong so I read some of your replies out to my mate over the phone and he offered to let me try a new (spare) set of nuts for it which he had. we tried them in the lathe today and found the following.

We looked at the new and the old and the only difference was a slight bend in them only visable against a new set which is why it could not have been getting full engagement in the root of the leadscrew without having the adjustment slightly less engaged.

I can now confirm that the leadscrew is ok as I have checked it and the pitch error has gone. Also with the new nuts the handle is in a lower position than before confirming a lower engagement in the leadscrew.

Thanks for all your help in this problem I can now cut perfect threads with a bit of luck!

I hope you all have a Happy New Year

Regards

Nick

 

[mgjParticipant @mgj]

Thats great BUT- no thread is EVER perfect. There is always a tolerance!!!

[Author]

Posts