Headstock Collet Chuck – Query about Accuracy

[James AParticipant @jamesalford67616]

[James AParticipant @jamesalford67616]

I have made myself a collet chuck to fit the spindle on my Flexispeed lathe, but it has more run-out than I had anticipated: I am not sure whether I am expecting too much or whether my figures are typical. Measuring the shaft of a slot drill held in the collet, I get a 4 thou total run out.

I made the tool by threading a piece of brass and cutting a register to fit the spindle. I then span this piece onto the spindle and bored it out to fit the collet chuck. I then pressed the collet chuck into the bored-out hole. I have used the very end piece of one of these ,

cutting it off just below the widest part to minimise overhang.

As a precaution, I checked the runout on the register on the spindle and it is a near to zero as makes no difference, the needle just flickering as I rotated the spindle.

So I suppose the question is: am I expecting too much or does the accuracy of my work need to improve? I suspect the latter.

I would appreciate any suggestions.

 

Edited By James Alford on 05/06/2016 15:51:18

Edited By James Alford on 05/06/2016 15:52:01

Edited By James Alford on 05/06/2016 15:52:25

Edited By James Alford on 05/06/2016 15:52:50

[Thor 🇳🇴Participant @thor]

Hi James,

I have an ER-32 chuck very similar the the one in your photo, when I put it in the MT3 spindle in my Compact 8 lathe and a ground tool in a collet I get a TIR of around 0.01mm close to the chuck. Since the MT3 based collet chuck will not accept long work pieces I made an ER-32 collet chuck, on this homemade chuck the TIR was twice of what I got on the purchased chuck. (Too many thumbs I guess)

Thor

 

Edited By Thor on 05/06/2016 16:14:04

[James AParticipant @jamesalford67616]

Thank you, Thor. I looked at the specifications for the collet chuck that I bought, but it did not give the runout. However, I doubt that it differs much from your example.

I have wondered whether some of the runout has been cuased by me pressing the chuck into the piece of brass, resulting in it being slightly squint. I am tempted to try making another version, this time in aluminium. That way, I can make the hole for the collet chuck slightly smaller than it needs to be. I can then but heat it gently to expand the aluminium. As it cools, I hope that it will grip the collet squarely.

Failing that, I shall try something like the example that you have made, although not having turned a known taper before, I can envisage even more inaccuracies at the end of the exercise.

Regards,

James.

[JasonBModerator @jasonb]

James if you have pressed the main body of the chuck into your block then that is quite unlikely to be concentric to the ER taper as there is no need for it to be when the chuck is held by its oroginal taper.

[Neil WyattModerator @neilwyatt]

One idea might be to skim the body using a carbide tool while it's mounted by its taper to get a concentric surface.

Neil

[James AParticipant @jamesalford67616]

Jason,

That is a good point that I had not considered. The collet chuck that I used actually had a straight shaft, not a tapered one, but your point remains valid. Oh, well, start again, this time with a fresh collet chuck and using a short length of the main shaft, not the top end.

Regards,

James.

[James AParticipant @jamesalford67616]

After cleaning out some swarf that had made its way into the threads and by tightening the collet closer more fully, I have managed to reduce the runout to 1.7 thou (0.045mm), but it is still excessive.

[JasonBModerator @jasonb]

James, if you do start agian then check the chuck runout before you take a saw to it, at least you will know if it's you or the chuck that is to blame if it comes out worse.

[Bernard WrightParticipant @bernardwright25932]

James,

Have you checked the mouth of the chuck for runout with no collet in, before resorting to making another.

Did you check the OD of the chuck to be concentric to the taper bore, before making your carrier body?

Also check the nut for radial play on the threads of the body.

I've read elsewhere that others have gained better accuracy, by tapping the nut after tightening the collet onto whatever is gripped, similar to tapping true a chuck on an undersized register.

Bernard.

Edited By Bernard Wright on 06/06/2016 08:28:53

[VicParticipant @vic]

I can't help wondering why you would want to use a milling chuck on a Lathe when ER32 Lathe chucks are available. Unlike a milling chuck stock can be passed through the headstock.

**LINK**

[Russell EberhardtParticipant @russelleberhardt48058]

If you want to get good runout you need to machine the taper after fitting the chuck to the spindle nose. It's not that difficult. I made one for my Atlas lathe over ten years ago and the runout is less that 1/2 thou. It was much easier than chopping up a milling chuck and trying to get it to run true.

Have a look on Harold Hall's site: **LINK**

Russell

[James AParticipant @jamesalford67616]

Thank you for all of the replies. In response to a few questions raised:

Have you checked the mouth of the chuck for runout with no collet in, before resorting to making another.

Yes: it is the same as with a collet installed.

Did you check the OD of the chuck to be concentric to the taper bore, before making your carrier body?

No: I asumed that it would be accurate (possibly a mistake to do so).

Also check the nut for radial play on the threads of the body.

The threads are tight.

I can't help wondering why you would want to use a milling chuck on a Lathe when ER32 Lathe chucks are available. Unlike a milling chuck stock can be passed through the headstock.

To be honest, I had not heard of these. However, those that I can find so far are all too large for my lathe to take comfortably.

I've read elsewhere that others have gained better accuracy, by tapping the nut after tightening the collet onto whatever is gripped, similar to tapping true a chuck on an undersized register.

A good idea that I shall try.

If you want to get good runout you need to machine the taper after fitting the chuck to the spindle nose.

This is what I shall try next: truing the taper to suit the machine.

Thank you for the link to the Harrold Hall site as well. I am tempted to try making a body from scratch, but I shall need to think about the closer thread. My lathe does not do screw-cutting and I am not sure whether a die-cut thread would be accurate enough.

 

Regards,

James.

Edited By James Alford on 07/06/2016 07:49:14

[James AParticipant @jamesalford67616]

I have finally been able to get back out onto the lathe to have another play with this piece. I have used a small Dremel style grind stone in the cross-slide to try and true the inside taper. I have now managed to get the runout down to 0.01" when measured on the inside of the taper. Measuring the outside of a mill, held in the collet, there is a 0.03" runout.

I shall carry on trying to true it further, but I think that I am unlikely to get it much truer.

Regards,

James.

[James AParticipant @jamesalford67616]

Well, I have given up on this plan. I have managed to get the taper inside of the chuck to run true to within about .005" / 0.0127mm, but the runout on the shank of a slot drill is still excessively large.

I have looked into lathe collet chucks, as suggested by VIc, and propose to buy one of these from ArcEuro.

Am I right in thinking that I "just" need to turn up a back plate with a suitably sized register and then bolt the two together?

 

http://www.arceurotrade.co.uk/Catalogue/Collets/ER-Lathe-Collet-Chucks/ER16-Lathe-Collet-Chuck-62mm-Dia

 

Regards,

James.

Edited By James Alford on 19/06/2016 18:50:48

[Frances IoMParticipant @francesiom58905]

“Am I right in thinking that I “just” need to turn up a back plate with a suitably sized register and then bolt the two together?”
basically yes – I fitted a ER16 collect chuck to my small Perris depends on how your stockstock mounts a faceplate – for my Perris this was a 2.5in by 1/2″ of EN1A which had to be drilled and tapped with a 1/2″ UNF (20tpi) (done on my larger lathe) then fitted onto Perris and skimmed to square on the Perris headstock then the register cut the mounting holes drilled and tapped on the back plate – the chuck was drilled for 5mm cap heads.

(there is a ER25 chuck that has the same register but different fixing holes)

[James AParticipant @jamesalford67616]

Frances,

Thank you. I have a Flexispeed which has a 1/2" BSF thread. I have just the one lathe, so shall have to fabricate the backplate on that.

Regards,

James.

[Nicholas FarrParticipant @nicholasfarr14254]

Hi James, that's exactly what I did with my collet chuck from ArcEuro **LINK**

Regards Nick.

[Bodger BrianParticipant @bodgerbrian]

Posted by Frances IoM on 19/06/2016 19:44:30:

basically yes – I fitted a ER16 collect chuck to my small Perris depends on how your stockstock mounts a faceplate – for my Perris this was a 2.5in by 1/2" of EN1A which had to be drilled and tapped with a 1/2" UNF (20tpi) (done on my larger lathe) then fitted onto Perris and skimmed to square on the Perris headstock then the register cut the mounting holes drilled and tapped on the back plate – the chuck was drilled for 5mm cap heads.

This might be a silly question but in order to get a runout of 1 thou or less (which seems to be what people aim for), would the register on the backplate have to be machined so that the chuck is a pressfit onto it? or am missing something obvious?

Brian

[Frances IoMParticipant @francesiom58905]

I’m sure someone with much more experience will comment but the couple of purchased chucks and backplates appear to have a slight taper in the register – I made a backplate for my small Perris, skimming it and turning the register on the Perris, the ER16 chuck is mounted by cap headed screws through the chuck into the backplate which is screwed onto the spindle but with a small register so fits flush against the rim of the headstock spindle – the run out can be adjusted by tightening these screws – must admit more than I expected and though I can set a very small runout I wonder for how long it will hold

[Neil WyattModerator @neilwyatt]

Posted by Bodger Brian on 24/06/2016 13:46:57:

Posted by Frances IoM on 19/06/2016 19:44:30:

basically yes – I fitted a ER16 collect chuck to my small Perris depends on how your stockstock mounts a faceplate – for my Perris this was a 2.5in by 1/2" of EN1A which had to be drilled and tapped with a 1/2" UNF (20tpi) (done on my larger lathe) then fitted onto Perris and skimmed to square on the Perris headstock then the register cut the mounting holes drilled and tapped on the back plate – the chuck was drilled for 5mm cap heads.

This might be a silly question but in order to get a runout of 1 thou or less (which seems to be what people aim for), would the register on the backplate have to be machined so that the chuck is a pressfit onto it? or am missing something obvious?

Brian

More a 'push fit' – you shouldn't feel any play at all, so as close as you can get without the need to use force really – the fit needs to be closer than your target runout. A tiny chamfer helps with fitting.

Neil

[Bodger BrianParticipant @bodgerbrian]

Thanks for clarifying things chaps. I've never actually seen a collet chuck in the flesh & I must admit that a slight taper on the register hadn't occurred to me as a possibility.

Brian

[john naylor 1Participant @johnnaylor1]

I have looked at all the postings regarding collet hoders.

This is what did.I purchased a piece of 2 1/2" dia round bar,gripped in a 3 jaw chuck and machined the 1.5 x 8 tpi thread to fit the headstock knurled the O/D and rough bored it and then drilled two transverse hole to accomodate a tommy bar .Screwed this on the headstock spindle and did the rest of the machining in situ on the headstock.TRO is about 0.001.I got a detailed drawing for the internal dimensions and taper from a web site which I cannot find now.This item was for an ER32 collet holder.

Edited By john naylor 1 on 06/03/2017 13:19:28

[john naylor 1Participant @johnnaylor1]

PS the nut was puchased separately.

[john naylor 1Participant @johnnaylor1]

[Author]

Posts