Arc Euro ER16 runout

[Neil WyattModerator @neilwyatt]

Posted by Nicholas Farr on 09/12/2018 13:46:07:I can always remember a saying on the back of a box of matches (can't remember the brand, but there were hundreds of different sayings on the back of their boxes)

England's Glory, Moreland, Gloucester?

[Andy CarlsonParticipant @andycarlson18141]

Thanks all again for the input.

First of all let me say that I have no complaints about the service from Arc Euro.

There are four mechanical components involved here – the backplate collet holder, nut and collet. Three of those came from Arc Euro one I made myself. Then there is a human component (me) of course. The reason for posting here was twofold: firstly to seek some help to narrow down the problem (successfully I think) and secondly to try to gauge what level of accuracy it would be reasonable to expect from the Arc Euro products.

I do really need to make this work – the price of the stuff from Arc Euro is a fairly small consideration compared to the investment of time to make the backplate.

The people telling me to do the nut up more tightly seem to be spot on. Now that I've hunted down a big enough spanner and tightened the thing up the runout has reduced to less than a thou. The runout seems to be slightly less with the 3mm Proxxon cutter shank than it does with the 3.2mm drill. This is a good result – now it is worth my while to chase down the half a thou of runout that I measured inside the collet holder taper.

I do still think that I might want to replace the nut even if it is perfect. The spanner needed to do the thing up is nearly as big as the lathe and hardly convenient to store with the lathe tooling. A nut that uses one of those pegged wrench things seems like a better bet.

For the people who asked… my backplate started out as a piece of 60mm diameter by 20mm long cast iron bar from College Engineering. Most of the machining was done on my Cowells lathe with just the final few ops done on the Unimat. I don't think the Unimat has enough slow speed torque for this job and holding that diameter would also be a challenge. The collet holder is 62mm diameter so I needed to be careful to keep as much of the diameter of the blank bar as possible to provide enough 'meat' for the three M5 fixing screw threads but I think that going up to 70mm bar would have made life tricky on the Cowells.

Finally… I'm no expert machinist but I do have an engineering background… sufficient to know that there is no such thing as perfect. Everything has an error and each part that is added will usually make the cumulative error worse. It would still be useful to see runout measurements from other people with the same (or more high end) collet setups… what is a reasonable expectation for runout from the Arc Euro products and what (if any) improvement is obtainable for more money?

Regards, Andy

[Chris TriceParticipant @christrice43267]

Do Emco do an ER collet chuck for an SL because I'll put money on the return rate for faults being significantly lower, if any, and cocentricity being better than the maximum TIR figures Emco will quote. I know they make an ER25 chuck for the Compact 5 and the one I have is spot on. I made extended tommy bars for mine to make tightening/untightening more hand friendly.

[Andy CarlsonParticipant @andycarlson18141]

I'm pretty sure the answer is no. The SL was replaced by the Unimat 3 a long time ago. There are pre-ER type collet holders for the SL but these usually fetch silly prices as well as requiring an older type of collet.

Regards, Andy

[John HaineParticipant @johnhaine32865]

Indeed, silly prices, even if you can find one.

Andy, I think you'll find that the M type nut that uses the spanner with pegs has a smaller thread – 19mm rather than 22 mm. I made an ER16 holder for my Acute tool grinder, and you have to use the smaller diameter to fit the work head.

[JasonBModerator @jasonb]

You may be able to chase down the 1/2 thou of the chuck by easing off the 3 screws and giving it a nudge in the right direction then tightening up again, a tiny skip off the register may be needed if you can't shift it with a nudge.

that leaves 0.0005" TIR from the collet and I would say at that price point you can't complain, if you want more then you need to start looking a ultra precision collets and pay 5-6 times the ARC price for them. You may get better than that with some from ARC but it would be more by luck than them being made to a gauranteed high spec. A Rego-fix Ultra precision ER 16 is good for 0.0002" TIR and cost about £35 a pop.

By way of comparison this is an ARC 10mm collet giving 0.0004" tir on my mill when fitted in a non ARC (I think) holder that has about 0.0003" tir which has been good enough for me.

 

M-Type = Mini so smaller thread.

 

Edited By JasonB on 09/12/2018 16:23:19

[Neil WyattModerator @neilwyatt]

An example, Korloy produce ER20 collets to three levels of accuracy 0.002mm, 0.005mm or 0.010mm run-out.

'Half a thou' is probably a reasonable level of accuracy to expect from out of the box 'hobby' collets of 'ordinary quality'. Note this is the collet alone.

Richard Gordon (MEW 232) found it was quite easy to tap collets into more accurate alignment:

"Conclusions

"I never bothered with the clamp, a few light taps to nut at the indicating high position of the cutter is all that I found was needed to improve the concentricity of the ER collet system. I hope this is of use to others, or perhaps you knew this already!? By the way, I don't think this will correct for any significant errors in the spindle, nut or collet, it just re-positions the nut back to the correct concentric position after tightening. If you are still finding it difficult to achieve 0.01mm T.I.R or less, then check for cleanliness and try rotating the collet 180 deg (easier said than done, I know). If still no improvement with multiple collets, try another ER nut – I prefer the hex ones rather than the C spanner where available."

The full article is well worth looking up as though short it contains much of interest, I'm surprised it didn't generate more comment.

Neil

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Andy Carlson on 09/12/2018 14:52:20:

I'm pretty sure the answer is no. The SL was replaced by the Unimat 3 a long time ago. There are pre-ER type collet holders for the SL but these usually fetch silly prices as well as requiring an older type of collet.

Regards, Andy

.

If I understand correctly … The older type of collet, E16, differs only slightly from ER16

This page is worth a look : **LINK**

http://paulthehalfwit.blogspot.com/2014/01/a-collet-chuck-for-unimat.html

MichaelG.

[jimmy bParticipant @jimmyb]

Its common practise in industry to clock reamers in ER collets, if they are out, slacken SLIGHTLY and tap true and re-tighten.

You can pay a staggering amount for precision collets….

Jim

[Neil LickfoldParticipant @neillickfold44316]

I have spent a considerable amount of time looking into errors with the ER collet system. MY conclusions are , if the collet itself is not concentric everywhere, then the errors seem to multiply and get worse. The ultra precision collets are worth it in the end as the tooling does work better as it is held more concentric. ER collets should only be used with parts or cutters that allow the entire length of the collet to be engaged in the collet as well.

Another issue is the concentricity of the thread on the outer body to the inner 8 deg taper, and then the concentricity of the inner thread of the nut to the 30 deg taper. Errors in these, will prevent the collet from correctly locating and being concentric.

I have trued the inner 30 deg taper of the nut, by locking the nut onto the outer with a sleeve. Then made a chisel looking boring bar to just clean up the inner 30 deg taper. In my case a screw cut the M50 thread and then turned the shoulder in the same setup. I had made the sleeve 1st. This has corrected the cheap nuts that I brought. The other nuts with the rotating inner are junk and I found no simple way of correcting these. They will only be to about 0.02mm or so for concentricity . Yes they can be tapped to knock the item into alignment. The RegoFix 1 piece nuts with their low friction coating is concentric to better than 0.005mm when tested on my correction fixture. Did not use a better dti at the time. The cheap 1 piece nuts were out by about 0.03mm aprox. After correction, and using the good collets, a part or cutter can be held and be better than 0.01mm for concentricity all the time.

Neil

[Andy CarlsonParticipant @andycarlson18141]

This evening I worked on it it some more and made things worse.

I found half a thou of runout on the register diameter so I gave it a really light shave. This of course lost the close fit between the collet holder and the register but I hoped that I could nudge things into place before fully tightening the screws.

It didn't work. I think the reason is that the counterbores for the fixing screws seem to be somewhat dished so the collet holder moves when the screws are tightened.

So I think that getting the register right is the only answer and it's time (another day!) to face off the register and make a new one. There is room for at least one more attempt or possibly more since I think it can be shallower than my first effort without losing any of the effective contact area.

Quite what went wrong with my first attempt at finish turning the register is not 100% clear to me. The sound of the cut was and the resulting finish were not completely even around the diameter which may be down to a lack of uniformity in the casting. Next time I will do the last few cuts with a HSS tool instead of a carbide insert, use a slower spindle speed and rather than doing the whole job in one 'hit' I will leave the thing to fully cool down before starting on the last few cuts.

Regards Andy

[John OlsenParticipant @johnolsen79199]

Emco did make an ER16 chuck and colletts for the Unimat 3, I know because I have one, bought new back in the early eighties. The back plate for it was turned on the unimat itself with no problems. Someone above questioned the ability of the Unimat to turn steel. The main problem is (or was) the abysmal system using o-rings as drive belts. Since those days mine has acquired a toothed belt drive which transforms the machine.

John

[Andy CarlsonParticipant @andycarlson18141]

The first job I did on my SL was to make a couple of new tommy bars by turning down the ends of some 8mm steel bar salvaged from an old hoover. I had no problems at all. It also managed to finish turn and face the cast iron backplate for the collet holder.

I don't use O rings – my drive belts are orange things made by jointing some general purpose drive belt material. I don't plan to change the Unimat drive because I can put more difficult work on the Cowells. The Unimat on the other hand has ball bearings on the spindle and can spin small diameter work at a more suitable speed.

Regards, Andy

[blowlampParticipant @blowlamp]

Posted by Andy Carlson on 09/12/2018 20:48:07:

This evening I worked on it it some more and made things worse.

I found half a thou of runout on the register diameter so I gave it a really light shave. This of course lost the close fit between the collet holder and the register but I hoped that I could nudge things into place before fully tightening the screws.

It didn't work. I think the reason is that the counterbores for the fixing screws seem to be somewhat dished so the collet holder moves when the screws are tightened.

So I think that getting the register right is the only answer and it's time (another day!) to face off the register and make a new one. There is room for at least one more attempt or possibly more since I think it can be shallower than my first effort without losing any of the effective contact area.

Quite what went wrong with my first attempt at finish turning the register is not 100% clear to me. The sound of the cut was and the resulting finish were not completely even around the diameter which may be down to a lack of uniformity in the casting. Next time I will do the last few cuts with a HSS tool instead of a carbide insert, use a slower spindle speed and rather than doing the whole job in one 'hit' I will leave the thing to fully cool down before starting on the last few cuts.

Regards Andy

Do you have a counterbore or endmill that you could use to flatten off the faces?

Martin.

[Andy CarlsonParticipant @andycarlson18141]

Posted by blowlamp on 09/12/2018 22:41:01:

Do you have a counterbore or endmill that you could use to flatten off the faces?

No, my Proxxon miller only goes up to 3mm. In time I may get some bigger cutters for use with the vertical slide on the Cowells but at the moment I have nothing that large.

Regards Andy

[Michael GilliganParticipant @michaelgilligan61133]

Posted by John Olsen on 09/12/2018 21:48:13:

Emco did make an ER16 chuck and colletts for the Unimat 3 …

.

**LINK**

http://www.nielsmachines.com/en/emco-unimat-3-collet-holder-er16-and-5-rego-fix-c.html

.

But listed here as using ESX 16 collets : http://www.rejon.co.uk/Unimat_2016.pdf

The plot thickens !

MichaelG.

.

Update: Quoting from another forum:

https://www.practicalmachinist.com/vb/general/noob-questions-about-collets-especially-er-collets-269209/

Sounds like you figured it out already… but ESX is indeed exactly the same in all design respects to ER. ESX are only termed such, though, by Schaublin. There is a historical reason for this, but it is really not worth getting into here. The Schaublin ESX collets are high quality, Swiss made collets equivalent in precision to Rego Fix (also Swiss manufactured). The reason I bring this up is to emphasize the quality of the set you already own.

Edited By Michael Gilligan on 09/12/2018 23:07:46

[Brian GParticipant @briang]

Hi Andy

This is on my "to do" list should I be able to use my SL (I must first print a set of guards so I don't fall onto the pulleys or chuck, which I suspect would hurt) although I would probably go for ER11 as I already have the collets.

**LINK**

Each of the stages in making an ER16 collet chuck on the Unimat SL is illustrated, and each photo has a text description. The screwcutting is done by an ingenious method that takes advantage of the headstock quill movement, which removes the need for the rare and expensive Unimat thread chaser and masters.

Brian

[John HaineParticipant @johnhaine32865]

Yes, I'd forgotten that the older (pre Unimat 3) models had the ability to rotate the headstock around a vertical axis for taper turning, so one can make the collet taper without a topslide. My 3 of course doesn't have that feature.

[Andy CarlsonParticipant @andycarlson18141]

I really like that way of cutting a thread on the SL. It's a bit of a roundabout route and it took me a few double takes to figure out what was actually happening but you have to admire the ingenuity.

At the moment I'm persevering with the backplate. The front end has had 3mm taken off the register and the fixing flange (using the Cowells) and the finish turning on the SL is now down to 0.12mm oversize. I'm taking a break now.

Some runout reappeared during the finish turning. I'm pretty sure this is down to unevenness in the casting. I've now resharpened the HSS tool and tightened every Allen screw down. I think I'll also seek out a wedge to add some extra support under the nearer of the two bars. Anything to help the rigidity!

[Nicholas FarrParticipant @nicholasfarr14254]

Posted by Neil Wyatt on 09/12/2018 14:11:33:

Posted by Nicholas Farr on 09/12/2018 13:46:07:I can always remember a saying on the back of a box of matches (can't remember the brand, but there were hundreds of different sayings on the back of their boxes)

England's Glory, Moreland, Gloucester?

Hi Neil, yes they were the one's. They were the type that could be struck on just about anything, downside of course, they were not a safety match.

Regards Nick.

[Andy CarlsonParticipant @andycarlson18141]

The register is now finish turned to fit the chuck again. I'm pretty sure there is no measurable (by me) runout on the backplate register now. The fit is good again but fit-wise I think I did a bit better at the first attempt.

After trying all three possible orientations of the chuck I got a runout between 0.6 and 1 thou inside the chuck taper.

The runout was consistently towards the same side of the chuck so I think this is coming from the chuck.

With some judicious tapping before finally tightening the screws I was able to get the runout down to about 1/4 thou, with a similar reading on with the 3mm Proxxon cutter in the collet.

I think I will stop messing with it now.

In other news, I found at least one place (Cutwel) that offers ER16 full sized nuts in a choice of hex and pegged wrench varieties (plus they do the mini one too). They also do a 25mm A/F spanner that looks rather more sensibly proportioned for this job.

I do think that I will try making an ER16 chuck for the Cowells myself (on the Cowells)… but not for a day or two.

Thanks again for all of your input.

Regards, Andy

[John HaineParticipant @johnhaine32865]

Following Joseph's observations I thought that a good substitute for a double-sided spanner would be a socket. I didn't have a 25mm socket in my set, and anyway since one will be tightening the collet with either a tool or material in the collet a long-series socket is needed to give it room. It seems to be near impossible to get a long series 25mm conventional socket, but I managed to locate one designed for impact driving on Amazon from KS Tools. Conveniently it has transverse holes at the drive end that with a bit of easing allowed a 25 mm tommy bar through. This allows me to tighten the nut hard without applying much or any transverse force.

I also needed a way to lock the spindle – the collet end has a pair of flats which are 20 mm a/f bit only about 6 mm wide. Believe it of not buying a thin 20 mm open-end spanner seems to be near impossible, they are all about 10mm thick at the business end. Eventually I gritted my teeth, got out my trusty 8 mm TC slot drill, and reduced the a/f dimension to just under 19 mm so I can use the very nice "Williams Superslim" spanner that came with my VMB. This was a job made very easy as I could hold the spindle in a JS ER40 collet block.

[Andy CarlsonParticipant @andycarlson18141]

I've ordered a 25mm spanner from RDG. It hasn't arrived yet but it seems (from the picture) to be more sensibly proportioned for use with the collet nut. I still need to make a proper tommy bar to fit the holes in the chuck because (inevitably) they are a different size from my other chucks and my die holder.

[Author]

Posts