Lathe line-up

[Horwig]

I was never sure if my problems were the Micky Mouse lathe or the Micky Mouse trying to use it!

However recently, I compared the jaws by winding the cutting tool in and turning the empty chuck by hand, lo and behold it touched one jaw well before the others, no wonder my jobs were never accurate.

Question is, how should I machine the jaws, is there a correct way of doing it?

 

Huw

[Peter Drew]

Are the jaws "soft" or hardened  ?  

[Horwig]

They are soft, I can mark them with a file

[Peter Drew]

The  usual method is to hold a piece of ground steel bar and skim the outer details on the jaws and then hold a sturdy ring on the outside of the jaws and lightly bore out the inner details.   Make sure that the jaws have been fitted in the right sequence and that the same sequence is always followed. 

[Gina]

I never rely on the chuck being accurate - mine certainly isn't.  If I want an accurate two ended turned item I use a turned mandrel.  The actual item is drilled through then a mandrel is turned so that it's a close fit in the hole.  A thread turned on the free end allows a nut to be screwed on to hold the job.  This way, the job has a virtually perfectly aligned hole and outside.

[Lonestar70]

Hi Huw,

Checking the jaws with the chuck unloaded will not give you a true indications since the jaws are free to float a little in that state.

Try gripping a piece of ground steel in the chuck and test against the ground steel, ideally you should use a dial test indicator for this which will give you an accurate figure for the run out, however, if you are careful you can get a pretty good idea by: -

1/ bringing the tool gently up to the ground steel and take note of the dial reading on your cross slide hand wheel.

2/ Back the tool off again a couple of turns then rotate the chuck 90 deg and wind the tool back in gently until it touches the ground steel... take a note of the reading on the hand wheel dial.

3/ Repeat step 2 a further 2 times rotating the chuck by 90 deg each step.

You can now compare the 4 obtained numbers... subtract the lowest from the highest and you will have the total amount of run out.

Most 3 jaw chucks have run out figures of a few thousands of an inch and not much can be done about it.

The only exception being a 'grip true chuck' which can be adjusted to zero run out.

When you say your work never comes out accurate, what exactly is the problem?

The way the lathe is mounted can have quite a lot to do with how accurate it is.

I recommend you do this before even considering machining the jaws... there are several things that can cause poor results from a lathe.

If you need further help then just ask.

Best regards.

 

[Horwig]

Thanks all. I'm probably trying to do too much with a small lathe. It's a Chester Conquest, one of the 7x13 familly of Chinese products.

At the moment I'm working on the adaptor between the camera and the Wynne corrector for my new focuser. I'm machining a piece of 4 inch ali tube section to a 16mm length by mounting it on the shoulders of a 4 inch chuck.

It's the ali collar on top of the Wynne here:

When I measured the old one on a surface plate with a dial gauge it was about 250 microns out, not a good situation. Considering using a small grindstone on a router mounted somehow on the cross slide of the lathe to true up the shoulders on the chuck.

[Lonestar70]

Hi Huw,

You might be better of making an expanding mandrel to hold the tube on the inside bore.

You can then machine either 1 or both ends at the same setting ensuring they are truly parallel to each other.

This article shows how to make one... just change the outside diameter of the split part to suit your tube.

http://www.cartertools.com/mandrel1.html

I would suggest that you make the split part first but make it a bit oversize to start with and don't split it yet.

Then make the expanding cone followed by the main tapered body... leave this in the chuck.

slip the split part onto the body followed by the expanding cone and add a washer and nut and tighten down to grip the split part.

Turn the outside to a very close fit in your tube.

Undo and remove the nut and washer... then remove the split body.

Use a hacksaw and put a single split along one side of the part all the way through into the bore.

de-bur and reassemble with the nut and washer but don't tighten it.

Slip your piece of tube onto the expanding body and tighten the nut to expand the split part until it firmly grips the tube.

Face each end to correct thickness... Job Done.

Hope this helps.

[fozzybear]

hi I agree with Lonestar70 are we referring to a 3 or 4 jaw chuck there are different ways to set them up sorry to arrive late on the project?  I use a tormach15l at work

If a 3 jaw then how old is the lathe the jaws might be worn and you can grind down the offending one if not adjustable. not knowing your lathe

Andy

 

hang on the conquest is your's the 4 jaw independent or not?

[Stub Mandrel]

WAIT!!!!

Swap two of the jaws over.

If it is 'well before the others' it isn't wear and it isn't poor quality, it may be, ahem, user failure... You might have the jaws in the wrong order.

I have two (quality) 4" chucks, one of Chinese and one of Indian origin and, despite what the Burnerd-worshippers will tell you, both of these will hold a bar with no more than 0.02mm runout (if they are scrupulously clean).

Even a cheap imported chuck shoudl be accurate to within a few thou when new.

 

If the jaw swap makes it worse, then revert and test the ACTUAL runout with a DTI and report back.

Don't do anything drastic yet.

[Chriske]

10 hours ago, Stub Mandrel said:

WAIT!!!!

Swap two of the jaws over.

 

 

 

That's not it. If you have mounted the 3 jaws not in the correct order there will be a severe wobble, not just a fraction of a millimeter.

Best is to regrind the inner side of the jaws. That is the only way to have the chuck running perfect again.  If you're a frequent user of that lathe you should perform that task at least once every year.
The only tool you need is a Dremel with a small grinding stone. 

You should first prepare the jaws before you proceed..!
The Dremel is fixed in the toolpost of the lathe. Lathe is running at (about) 300rpm, Dremel at highest rpm. Very slowly start grinding the inner side of the chuck until all three jaws are done. You should use automatic feeding mechanism to do this.

After that grinding procedure the lathe's chuck will perform as new, the wobble is completely gone..!

[Lonestar70]

1 hour ago, Chriske said:

That's not it. If you have mounted the 3 jaws not in the correct order there will be a severe wobble, not just a fraction of a millimeter.

Best is to regrind the inner side of the jaws. That is the only way to have the chuck running perfect again.  If you're a frequent user of that lathe you should perform that task at least once every year.
 

You should first prepare the jaws before you proceed..!
 You should use automatic feeding mechanism to do this.

After that grinding procedure the lathe's chuck will perform as new, the wobble is completely gone..!

I agree with you regarding the placement of the chuck jaws, however, I disagree with the need to grind the jaws every year.

I use my lathe daily for private business turning, a lot of it quite heavy, and have never needed to grind my chuck jaws in over 15 years.

It had an average run out of 0.003" over it's full diametrical capacity when new and it maintains that figure within a tenth or so to this day despite extensive use.

The problem with grinding the jaws is that they will only be dead accurate at the diameter you grind them at... at any other diameter then errors in the scroll mechanism, and yes there will be some regardless of the make of chuck, will introduce run out errors... such is the way with scroll chucks.

Huws' problem is not caused by the inside faces of the chuck jaws since he is using the outside stepped faces to locate his workpiece... grinding these outer faces with a piece of ground steel held in the chuck will help a little, however, the same scroll errors will also effect the outside of the jaws at any other diameter than that at which they were ground.

The only way to be sure of accurate work would be to use an expanding mandrel made specially for the job without removing the mandrel body from the chuck after it has been turned.

Removing the mandrel body and then replacing it in the chuck will not guarantee the same accuracy even if you mark the mandrel body for the number 1 jaw position.

Holding it in a collet would be more accurate for multiple useage, but even they have a small amount of run out, however, I don't think Huw has the luxury of such things... but I could be wrong on this aspect.

Huw gave a figure of 250 microns error (0.0098"), however, what we don't know is if this was between the end faces or if it was a taper over the outside dia... both could have different causes.

More detailed information would be helpful in determining where the cause of this error might be... grinding chuck jaws would be way down the list for things to check/do.

Keep happy.

 

 

[Horwig]

Thanks all

 

Firstly, refering to Neil (Stub)'s post, jaws on my chuck and the chuck itself are numbered, so they should be in the correct order.

I involved my cousin in the problem, he knows a thing or two about metalwork, he has a lovely Viceroy machine.

We firstly measured runout, it was fairly acceptable. What really worried me was the stepped shoulders of the jaws. I was mounting a 4 inch tube section in a 100 mm chuck after all. The difference in protrusion of the steps varied by the quoted 1/4 mm compared to the face of the chuck. After much playing, tea and discussion, he took the chuck back to his workshop, mounted it in his lathe and faced the jaws on mine, job was a good one.

I've re-finished the ali collar I was working on, now it's within 20 microns, reckon that's good enough for me.

 

Huw

[Chriske]

All depends on the accuracy you're working with on that lathe.
Indeed, if you load rather heavy axis in your lathe, there probably no need to grind the jaws. Everyone should decide that for himself.

After grinding I put a straight and round polished rod in my lathe. I can assure you, putting a dial gauge against the polished rod, I see no movement at all on my scale.
But again, all depends on the needed accuracy of the parts you're producing.
To be clear, every time I grind these jaws, I only remove very little. Most of the time I only remove just 1,2 maybe 3/100 of a millimeter. I like my jaws(lathe) to be absolutely perfect.
For my personal needs this correction is needed. I like both ends of every part I'm working on perfectly 'in line'.

[Stub Mandrel]

44 minutes ago, Chriske said:

All depends on the accuracy you're working with on that lathe.
Indeed, if you load rather heavy axis in your lathe, there probably no need to grind the jaws. Everyone should decide that for himself.

After grinding I put a straight and round polished rod in my lathe. I can assure you, putting a dial gauge against the polished rod, I see no movement at all on my scale.
But again, all depends on the needed accuracy of the parts you're producing.
To be clear, every time I grind these jaws, I only remove very little. Most of the time I only remove just 1,2 maybe 3/100 of a millimeter. I like my jaws(lathe) to be absolutely perfect.
For my personal needs this correction is needed. I like both ends of every part I'm working on perfectly 'in line'.

If I want that level of accuracy I plan y work so I can do it all at one setting turn between centres, use a four-jaw chuck or use a collet chuck.

 

 

[Chriske]

Yeah I know, that a possibility too between centres, but I don't like it...

[Horwig]

I eventually sorted the lathe out, with a lot of help from my cousin.

Lessons learnt here, never take anything for granted.

I was persuaded to buy a new magnetic stand for my dial test indicator, makes life a lot easier, now I can REALLY repeatably measure things.

First of all, I noticed that when measuring the face of the chuck I could vary the reading by loading it with a bit pressed from the tail stock against an end mill reverse mounted in the chuck to load it, so looked at the preload on the back of the main shaft, it was loose, so tighten it up and things were much better, but now my runout on the end mill mounted in the chuck was well off.

Above I said that the jaws were numbered, so MUST be in the correct location, as the slots were also numbered.

Here is the sound of me eating my own words, re-inserted the jaws one slot out, and runout was now about a tenth of the reading before.

I'm now starting to machine things that make sense, mind you, the lathe is still operated by Micky Mouse.

 

Huw

[Peter Drew]

3 Jaw chucks usually have three options for tightening with the key, these can vary slightly so find which one gives the least runout and always use this one to tighten.   

[fozzybear]

At work we have RIGID 300 C POWERDRIVE COMPLETE THREADER  and a common trick is to put in reverse you cannot get a thread on anything. ok of topic but a workshop prank...  I came foul of this a few times. Doooo.

[Horwig]

On 31/03/2018 at 23:22, fozzybear said:

At work we have RIGID 300 C POWERDRIVE COMPLETE THREADER  and a common trick is to put in reverse you cannot get a thread on anything. ok of topic but a workshop prank...  I came foul of this a few times. Doooo.

The prank I really like was from a very respected broadcast equipment manufacturer. They would prepare a special treat in the quality control section of the plant for any new engineer. It involved hiding a 12volt bulb wired across the switched side of the main 240v power switch and then wait for the poor trainee to switch it on, huge flash and a bang, and as they say adrenaline is brown.

[fozzybear]

8 hours ago, Horwig said:

The prank I really like was from a very respected broadcast equipment manufacturer. They would prepare a special treat in the quality control section of the plant for any new engineer. It involved hiding a 12volt bulb wired across the switched side of the main 240v power switch and then wait for the poor trainee to switch it on, huge flash and a bang, and as they say adrenaline is brown.

reminds me of the song golden brown by the stranglers.....