lathe set up

[firebird]

Hi

I have a Myford ML7 lathe that has done everything I have asked of it very well. I have some work to do that will require machining to very tight tolerances so I bought a test bar to set up accurately. Checking between centres the tailstock was out by about .008. which I have cured by setting the tailstock over with the adjusting screws. Its now ok in that way. See video.



However when I set the dial on the top it reads that the tailstock end is lower by .007. See video.



How do I/is it possible to correct this.

Cheers

Rich

 

[Bluechip]

Hi F/B

If you have Raising blocks, slacken off the tailstock end securing nuts.
If the DTI moves, the bed is twisted, and you do the set-up in the Myford maual from square 1 ..

I had to do it with my S7. In fact it has to be done an any Myford installation.

Sad fact of life, if the bed is not ' neutral ' to the raising blocks & jack screws everything after is pretty much futile.

Dave

 

[firebird]

Hi Dave

Thanks for the prompt reply. My lathe does not sit on raising blocks, it is however on a wooden bench so I presume if the bench isn't flat the lathe won't be either. Is it a case of playing about with packing under the the four hold down bolts. Would it be worth my while getting some raising blocks? Do they have jacks incorporated in them?

Cheers

Rich

 

[Bluechip]

Rich
My first Myford was a ML7/T on a drip tray with raising blocks on a wooden bench. Wood wanders about all over the shop with temp and humidity. I had no end of grief with the thing, I had to set the thing up about every 6 months. PITA. Yes, I suggest you do get the blocks. I believe RDG do em or maybe Chronos. If you've never done it, it can be a protracted business with raising blocks and jack screws, I wouldn't like to do it with shims.
If you want, email me with your phone # for a natter about it. I get calls for nowt at weekends....

Dave

 

[firebird]

Hi dave

Iv'e just checked, Chronos and RDG stock them. I'm going to Harrogate in two weeks so I think I will get some. That looks the easiest way to do it. When they are in place if I get stuck I will take you up on your kind offer.

Cheers

Rich

 

[Bluechip]

Rich

I think you may still have the problem, the blocks wont get round the fact that timber warps, not a lot but it was enough to upset mine. Got it more or less sorted in the end by mounting the drip tray on a piece of kitchen work top. This was 'attached' to the bench on a bed of RTV silicon. About 10mm thick. The worktop was screwed down on the un-cured RTV, when it was cured, the screws were removed. This more or less let the bench top timber move without affecting the lathe too much. I don't recall having to set the lathe up again anyway. The drip tray was, of course, bolted to the worktop.

Just looked at the video, if the tailstock is not horizontally axial, the DTI will be ' riding down the curve of the tailstock barrel' as it were.
May be the error is not in the vertical plane. Or had you determined that ?

How good is the test bar ? I have one that's not absolutely accurate. Found that out when I borrowed one made by Arrand.
Put it in the spindle and clock it. Mine is not staight by .0009 or so at the end when I rotate it. The Arrand one says <.0002 so I doubt it's the Myford headstock taper. In fact that seems to have zero error with a clock inside it. As does the chuck register

Have fun

Dave

 

[firebird]

Hi Dave

I see what you mean about the tail stock. It could be that the barrel tips down slightly. I have had to come in now, its not sociable to spend the afternoon in your workshop (according the the law of SWMBO) so I'll check that another day. My bench tops are kitchen worktop and the lathe has a drip tray. The raising blocks may still be a good idea though, when stuff rolls under the lathe it can be difficult retrieving it.

Cheers

Rich

 

[Holescreek]

Very cool! I wish my talking lathe sounded like a lawn mower when it moved. :big:

 

[Bluechip]

Hi Rich

If you are on something like worktops it would be as well to find out if there IS any twist on the lathe bed. Maybe not. Mine was on 10 x 2 joists, through bolted. This was just building timber, certainly not stable. Also it was in an insulated shed, not heated all the time, so the temp. and humidity would vary a great deal. Your arrangement is much better than the one I had. The old Geezer who sorted mine out was an ex-Rolls Royce millwright. Spent a lifetime installing and fixing machine tools, so I guess he knew what he was doing.The first thing he did on mine was to put a 1" dia. bar in the chuck, set a dial guage about 6" - 8" from the jaws, locked the carriage and cross slide, set the guage to zero, then slackened the securing nuts about a flat or two in turn while watching the guage.
He also locked the mandrel, as mine was a Tri-Lever, this could be done by selecting two speeds at once. Motor not running of course ;D
It moved, about .003". as I recall.
So, the bed was twisted, Bugger ...
That was it. Then followed a protracted bout of fooling about with the jack-screws, and it did come in.
The guage was set to zero with no bolts holding the bed, so this was the 'Neutral' ie as it was machined.
Just keep nipping things up and altering the jack screws until it gets locked down on all four bolts with the guage on zero still.
Once it was done, it's fairly easy to keep a check on it, I checked about every 3 months or so. It's no problem then, as it will be close anyway.
There has been an excellent write up in ME by Tubal Cain on the subject IIRC. Some time ago obviously.
Just need a bit of patience I think, not easy to do the first time.

Dave

 

[putputman]

Rich, it is always nice to have everything "dead true" on these machines, but from a practical look, when you adjusted the tailstock to run true in the horizontal path, I think you solve your problem for turning accuracy between centers. The .007" in 3-4 inches of travel on the vertical plane will be almost unmeasurable when you turn a shaft between center. It will change the tool height in relation to the shaft throughout the length of the shaft.

I bet Marv could come up with a formula that would tell the difference in dia/inch based on change of tool height.

Now if you are turning just using a chuck or collets, you might have a whole new set of problems. Perhaps you should chuck up your test mandrel and check that also.

Just another thought.

 

[mklotz]

Your wish is my command...

Imagine you've set your lathe tool on center so that, on a perfectly aligned lathe, it would make a cut with a radius of 'r'. Now suppose that the TS is high by the (presumed small) amount 'e'. As shown in the diagram, it will now cut a slightly larger radius, denoted by 's'.

The error introduced by the TS offset, then, is merely the difference between 's' and 'r', which turns out to be approximately e^2/(2*r). The diameter error is twice that, or e^2/r.

Two observations...

This error is tiny compared to the effect of an equivalent error in the fore/aft direction.
For e=0.005" and r=0.5", the error is 50 millionths of an inch. It's not unusual for new precision lathes to be delivered with the TS a few thou high so they will stay in spec for twice as long as they wear.

Since r occurs in the denominator, the error gets worse as the size of the workpiece decreases. Of course, as shown above, the error really isn't something you need to worry about unless the TS is wildly out of vertical alignment.

 

[putputman]

If I follow you right Marv, assuming that he had .007" misalignment in 4 inches and was turning a 1" dia shaft, the the difference in diameter from end to end (4") would be 0.0000979.

Should be close enough for any model I have ever seen.

 

[mklotz]

Your calculation is correct...

0.007^2/0.5 = 9.8E-5 = 98 millionths

 

[firebird]

Hi

Hey Marv, PutPutman, how does your brain do that. I can't get mine to do it. ??? ???

I tried the test again with the tail stock now wound further back into its housing. The reading is now down to .005 so I'm assuming there is a little play in the tailstock. as you say though thats near enough for me.

Cheers

Rich

 

[mklotz]

Hi

Hey Marv, PutPutman, how does your brain do that. I can't get mine to do it. ??? ???

Ah, another one who decided he'd never have any need for math in the real world.

I tried the test again with the tail stock now wound further back into its housing. The reading is now down to .005 so I'm assuming there is a little play in the tailstock. as you say though thats near enough for me.

I wouldn't be comfortable measuring the TS droop with a DI mounted on the carriage - too many other error sources that could affect the measurement.

Mount a DTI in a collet in the headstock and run the tip around a dead center mounted in the TS. That will give you a measurement of TS error more directly relevant to turning work supported by the TS.

Does your TS have a key or, more likely, dog point setscrew which prevents the barrel from rotating? If so, check its adjustment. Also, make all your measurements with the barrel lock locked down. (If you've already attended to these things, just ignore me and soldier on.)

And finally, remember that, whatever droop you measure, it's pretty much irrelevant.

 

[firebird]

Ah, another one who decided he'd never have any need for math in the real world.

Hi Marv

Yeah your so right. Didn't listen, didn't see any reason to learn. One thing I have learnt, the older you get the harder it is to learn.

Cheers

Rich

 

[putputman]

Rich, the main thing here is that you understand that the 0.007 of vertical movement in approx 4" has very little to do with your accuracy. When you corrected the horizontal offset, you essentially solved your problems.

Beat of luck on your new project.