Off the shelf E.R collet chucks worth it?

[MWParticipant @mw27036]

[MWParticipant @mw27036]

Hi,

I've not posted a question for a while now, but I've had quite a lot of other things to do in between.

I've been insulating my workshop roof and i'm on the eve of finishing a powered ventilation system for the building as well.

But turning back to the lathe after all this is said and done, I wanted to start on a simple "back plate for a collet chuck project". An E.R 40 one to be precise.

The recess doesn't fit my lathe, so I wanted to take all the due time and care making a good job of turning a cast iron back plate for the bugger to run true to my lathe.

Then I found this guy on youtube, who by the sounds of it, is trying the same thing, albeit with a different machine;

**LINK**

So a lot of people, including myself, buy these collet chucks, with the aim of using them to eliminate run-out. However, long story short, this guy has found that his one is not up to the job of holding stock completely true. In particular it has an overall runout of 0.05mm, and this is after all his extremely thorough measures to get this true, If you can bare watching all 3 parts that is.

He has even skimmed the back face of the chuck itself. But anyway, he puts the errors down to two things;

one, the taper of the bore isn't entirely true, and he doesn't have the means to correct this, but it's about 0.02mm out.

two, the collets themselves(are unbranded), probably account for the remaining 0.03mm runout.

Now this is okay if all you want to do is hold milling cutters, but my 3 jaw will give a better reading than 0.05mm, other than holding the stock more gently than a 3 jaw, this doesn't cut the mustard for me.

The most I could possibly accept even after all the fiddling round would be 0.01mm anything above 0.03, at chuck face, is worse than my 3 jaw.

So I don't really want to turn a lot of black dust, and the best I can hope for being that.

Do you think I would be better off making spindle collets, because I can at least guarantee that the bore I am turning, will be entirely concentric to the machine (or atleast around 0.01mm)?

Food for thought.

Michael W

[Nick HulmeParticipant @nickhulme30114]

He could have started out by lobbing a nice big chunk of known, free machining round stock in the chuck, carefully machining and checking a taper plug to fit the collet chuck taper properly (blue & all that to check fit). Drilling and tapping the centre of the plug would then allow him to clamp the chuck on a known true & central taper to indicate, and if required, machine the rear face and register (the only bits that matter).

It looks like he "Cheaped Out" on his initial purchase, should have gone to specsavers, better still Arc Eurotrade who do a nice ER40 lathe chuck. 

Edited By Nick Hulme on 14/04/2018 11:46:29

[MWParticipant @mw27036]

Your suggestion Nick. Thats exactly what he did. With aluminium stock though. And blued the taper before skimming the face.

[John HaineParticipant @johnhaine32865]

ER tapers are easy to make as they're quite short. The ER40 closing thread is 1.5 mm pitch. Closing nuts are easily obtainable even with a ball bearing. Why not make your own chuck turned in situ on the lathe? Might not be hardened but when/it wears you can always recut it.

Edited By John Haine on 14/04/2018 12:36:27

[JourneymanParticipant @journeyman]

Out of curiosity I just measured the runout on my Warco bought ER25 lathe chuck. As best I can measure it the internal taper is .005mm out front and back. I think that will do me nicely. I machined the backplate to fit on the lathe but used cap screws from the rear rather than have bolt heads on the front. Wrote it up *** HERE *** if anyone is interested.

John

[VicParticipant @vic]

I suppose it depends what you’re doing and why Michael. I’m a hobbyist and for what I’ve done to date a few thou either way makes no difference. I know there are some professional machinists on here for which these errors are too much but in such cases I wonder why they post on forums like this anyway. If what you’re doing is critical then I think you’ve answered your own question, you need to spend more money and buy the top quality stuff. I do sometimes wonder though how long these super accurate collets stay that way given the way they get deformed in use? Regular checks needed methinks! Good luck with whatever you choose!

[VicParticipant @vic]

I should have said, I think I checked my lathe collet chuck not long after I bought it and found it was a little bit off. I have to say though that it’s been a great piece of equipment to have and has been very useful.

[Anonymous]

Posted by Vic on 14/04/2018 15:51:39:

I know there are some professional machinists on here for which these errors are too much but in such cases I wonder why they post on forums like this anyway.

With an attitude like that they're probably wondering the same thing too.

Andrew

[James AlfordParticipant @jamesalford67616]

Michael W,

I bought a collet chuck from Arc Euro and needed to make a back plate for my lathe. This thread covers it and might help. I had real trouble getting rid of run-out initially, but it was my technique and skill, not the collet chuck at fault. I did get run-out down to 0.0005", a mere flicker on the needle.

http://www.model-engineer.co.uk/forums/postings.asp?th=120081&p=2

James.

[not done it yetParticipant @notdoneityet]

The’professionals’ and ‘experienced’ amateurs comments are very welcome by me, for one.

OK, most of my work is not that intolerant of the odd couple of thous or 0.05mm on overall dimensions, but how to do it better, or by an alternative method, or not wear out my cutters prematurely, is all good advice for someone who is a jack of all trades.

For example, out-of-round facing cutters which cut mostly on one side (like the banggood one that doubleboost raved about, so he gets offered more freebees) do not impress me – I much prefer those cutters that actually sound ‘sweet’ when in work and will provide a better finish plus work efficiently for much longer.

I usually cut slots with a smaller sized cutter, but how much better if I had confidence that using the nominal sized cutter would not result in an oversized slot? Agreed, I have time to cut the slot and then finish it to size, but I still admire those that just do it with the one (finishing) cut.

[Neil WyattModerator @neilwyatt]

0.015mm is typical for a standard set, so you are unlikely to meet your <0.001 requirement without spending $$$.

Also consider:

How accurate is your spindle bore/spindle nose register?

Your 3-jaw will give you 0.03mm runout now, how will it fare in five years time?

[JasonBModerator @jasonb]

You have added an extra "0" behind the point Neil!

[MWParticipant @mw27036]

Thanks for the contributions

Hmm, yes 0.01 is what i'm aiming for.

I've tested the spindle face and the taper runout albeit a while ago in the past, I could barely get anything to register on the needle. At this point a 0.01 reading clock makes it hard to say what it truly is running off at. But i'm satisfied it's fairly accurate.

The crux of the matter seems to be that i'm unsure whether I can trust the collets themselves. I'm willing to do all that Ade has done in order to correct it, even to the point of reskimming the face and taper bore of the chuck in situ. In some ways It doesn't surprise me that this is necessary because we're talking about fitting a chuck to a machine it wasn't made with.

A 0.05 or even a bit more would be fine if we we're talking about a 4 jaw self centre or independent, it isn't that i'm too fussy at all, at least I think anyway.

An interesting note, is that while the inside jaws of my 3 jaw can hold stock very well the outside jaws don't fair nearly as well and often go around 0.1-0.12. That might be a lot to do with the fact that there isn't a lot to grip with.

The other thing could be that my datum bar stock itself might be off, the only way to correct that would be to turn a slug then do the reading, something I notice ade didn't try.

I will also try marking the position of the rotation the chuck has been fitted on. I didn't think of doing that. Old machines used to come with a scribed mark for this but i'll have to find a way of doing the same on my machine.

His channel is quite interesting even if it is a bit unknown.

A better test might be to see if I can skim the bore taper of one of the morse taper chucks, using a carbide tip and seeing what result I get? It doesn't matter if I mess it up because I would've written it off anyway, as this suffers largely from the same problems as the flanged chuck.

Michael W

[Howard LewisParticipant @howardlewis46836]

Probably, the best that you will do for accuracy would be to take a blank mandrel on a MT taper, and turn the requisiite size x 1.5mm pitch thread on it, and a 16 degree, included, internal taper.

The MT taper and the seat for the ER collet should be about as accurate as your machine will produce.

Any eccentricity should then be down to the collets used.

The downside is that, unlike a collet chuck on a backplate, located on the spindle nose, you cannot have work protruding back into the bore of the spindle.

Howard

[richardandtracyParticipant @richardandtracy]

I have an MT4/ER32 collet chuck. I can see the needle on my dial gauge move, but it's less than 0.01mm. Not bad for a £22 inc postage collet chuck direct from China.

Could I buy the material for much less than that? Don't think so.

The question has to be: Do you enjoy making tools, or making stuff with tools? For me making the stuff with tools is what I want to do, and if the time making the tool is excessive (saving less than £2/hr) I'll buy not make. A collet chuck may be quick for some, but not me. And the accuracy of my cheap chuck is at least as good as I could make.

Regards

Richard.

[Neil WyattModerator @neilwyatt]

We ran an article by Richard Gordon a while ago where he virtually eliminated any residual error by tapping the closing nut.

Bear in mind that buying cheap is a lottery, you may do well as Richard has, but beware that it is far from unknown for goods that fail QC to disappear out the back of the factory and reappear on 'well known auction sites'. I've bought speed controllers that clearly have failed machine soldering and been restored to working order by a gorilla with a pair of gloves and an oversize soldering iron.

Not just a Chinese practice, I understand that Sinclair electronics used to buy reject transistors and pick out the better ones in its early days…

[John RuddParticipant @johnrudd16576]

Posted by Neil Wyatt on 14/04/2018 22:26:37:

.

Not just a Chinese practice, I understand that Sinclair electronics used to buy reject transistors and pick out the better ones in its early days…

I think it also included the Sinclair Super IC12 which was based on the Texas Instruments SN76013N audio chip….

[MuzzerParticipant @muzzer]

I doubt many of us tighten ER collets anywhere near the recommended torque, have the required tooling to do so or even know what the recommended torque is. Difficult to correlate experiences meaningfully in that context. Good talking point, though.

Murray

[not done it yetParticipant @notdoneityet]

Muzzer, so right. Two 150mm spanners are likely not long enough for many….

[Clive Brown 1Participant @clivebrown1]

Not just a Chinese practice, I understand that Sinclair electronics used to buy reject transistors and pick out the better ones in its early days…

That was my understanding, but I also remember being told that manufactures would produce a batch of transistors and then type-label them after testing their characteristics. That was in the days of germanium, has that now disappeared?

[richardandtracyParticipant @richardandtracy]

The process is certainly done with Micro SD cards, the smaller size ones being big ones formatted not to use an area that fails.

Regards

Richard.

[HasBeanParticipant @hasbean]

I have three backplate mounted ER collet holders, all have a TIR of less than 0.005mm as best as I can measure, can't vouch for the accuracy of the collets themselves though.

I bought a MT2 ER25 chuck from a well known site for not a lot of beer vouchers in the hope I might be lucky and get a good one, I was wrong! The morse taper had several areas where there was a 'dip' in the surface, the drawbar hole was off centre, the face where the nut makes contact hadn't been finished at all and has deep workholding marks on it and finally the TIR was 0.3mm +.

My other MT ER chucks are from more reputable sources and are all 0.01mm or better.

Paul

[Neil WyattModerator @neilwyatt]

Posted by Clive Brown 1 on 15/04/2018 08:52:06:

Not just a Chinese practice, I understand that Sinclair electronics used to buy reject transistors and pick out the better ones in its early days…

That was my understanding, but I also remember being told that manufactures would produce a batch of transistors and then type-label them after testing their characteristics. That was in the days of germanium, has that now disappeared?

In ye olde days it was standard to classify transistors into gain groups after manufacture.

These days production is less hit and miss – look at how ICs with tens of thousands of transistors on a single chip can be made.

Neil

[Author]

Posts