ER collet chuck runout

I'm sure this is a very common issue but out of sheer interest i measured the runout of various aspects of my lathe with a baty dial indicator.

I measured the run-out of the spindle, i could barely get the dial to twitch within the smallest division being 0.01mm, so it pretty much hovered near zero, so far so good.

I put a 1/2inch bar of ground steel in the 3 jaw self centre and measure the runout from 30mm away from the chuck, it measured 0.02-0.03, not bad for a self centre i thought.

I then try my ER40 MT4 collet chuck, check the internal 8 degree taper, it runs at 0.05-0.06, slightly disconcerting.

I then move onto the same test as the 3 jaw, using a bar of ground steel measured at 30mm away and, again, the run out of the taper reflected in the bar held, 0.05-0.06mm, i'm slightly saddened, thinking i should be able to get lower runout from a ER chuck than a 3 jaw.

I can't see how i could fix this with my equipment so would anyone be willing to regrind the body of the ER chuck to get closer to 0 for a nominal sum?

The only other option i can see would be to use threaded collets directly to the spindle in future.

Michael W

Edited By Michael Walters on 26/04/2016 14:01:54

Michael,

It's not by any means the full answer to your woes ... but I suggest you read this current thread. Raymond has done us all a great favour by getting 'benchmark' specifications from some of the better suppliers.

MichaelG.

Edited By Michael Gilligan on 26/04/2016 14:34:02

Hi Michael, Various folks on here [me included ] have been having an interesting confab regarding collets under the title [excentric collet] The runout you are seeing at the taper is indeed well out. if your reading were microns then fine, but not hundredth's. If you are getting the same reading from the taper, AND with a piece held in the collet at least the axis of the collet and holder bore are in line ie, not angled. or you would be getting a different reading.

From that it would appear the the axis of the collet holder bore is not true to the axis of the MT4 shank. so yes, it would need to be reground [best if it were the MTshank not the bore] I could well be wrong but I suspect that the process would not be easy for a home workshop in view of the accuracy needed. others no doubt will have their own views. Yes the runout figures for your 3j self centering are good and your collet should be at least half as good again.

Hi Michael , Is your collet nut fastened down UNDER the nut , ie the off center ring inside the nut .

Sorry Michael, I should add that you would have to find out what axis it is that is "off " in relation to the lathe spindle.and correct that area. Still not easy to do with an accuracy to get the best out of ER collets.

Michael W - There are lots of us with only 3 and 4 jaw chucks and old worn out lathes that can still produce good work. If you need really accurate turning concentric to an existing diameter put it in a four jaw and set it dead true with a DTI then turn the new diameters.

No need for ER whatever collets and chequebook engineering in search of the last .001 mm on your equipment.

Think about what you can do with what you have rather than only thinking what you can buy to do a job. You think a model engineer in the 1930's or 1940's making a few shillings a week could afford the fanciest and best collets etc? not likely, and yet some of the finest best made models of all time came out of their shops. Have a look at the work of guys like Peter Dupen for example. If they could have afforded it, they might have bought some fancier equipment, sure. But usually they just used what was available, and a lot of thought, and got the job done well in the end. JD

I did have the collet "snapped in" position so to speak, just a little dissapointed with the runout, maybe it can be solved one day but like jeff said, i'm gonna use whats available and make a small drive dog and turn between centres, ensuring absolute concentricity of all diameters.

Michael W

Further to JD's post ...

'Accurate' collets are a great convenience, but inaccurate, or unreliable, collets are probably no better [or worse] than a 3-Jaw chuck when it comes to your 'machining workflow'.

Remember that the very finest horological tuning is still done between 'dead centres' ... there is no workholding method that can better this for precision.

MichaelG.

Provided your two dead ctrs are true to the lathe axis otherwise you may get precision tapers

Provided your two dead ctrs are true to the lathe axis otherwise you may get precision tapers

.

Accepted, Jason ... Although 'the very finest horological turning' is done with a graver, on a handrest.

The big advantage [for this very small work] is that the piece can be removed for inspection/measurement, and replaced for further work, without worrying about chuck or spindle eccentricity.

MichaelG.

Works for big stuff too, certainly have crankshafts in and out the lathe a few times while they are being made.

Out of curiosity is there a point where between ctr work becomes inpractical as the diameter increases and the length decreases for example you would not want to make say a cylinder cover between ctrs.

J

PS I only tend to use a hand graver for (intentionally) curved work

It's all a matter of 'proportion', of course, Jason ... but the watchmakers are using dead centres and a bow ... just like a Pole Lathe, in principle. The point of 'encouragement' I was trying to make to Michael W, is that the finest work can be done with 'unsophisticated' tools.

MichaelG.

Too late to add this into my previous post, but it's worth reading:

[making a balance staff]

**LINK**

MichaelG.

One answer to er collets is to regrind the taper in the holder on the machine. It wont need much removing just careful setting up. Given a decent stone a Dremel could even do it but it seems rather difficult to get decent stones for those.

The collets themselves often seem to be pretty good but can only grip completely round a circle at one specific size. Not that this means it will increase run out, just decrease holding power. The biggest problem with them really from a work holding point of view is the grip length especially at larger sizes such as ER40.

John

-

.

John,

If you haven't done so already ... it would be worth reading the recent posts on the Excentric [sic] thread.

MichaelG.

Michael W, I've been down the same rabbit-hole several times myself, and while the axis of a chuck and the Morse Taper shank are generally not precisely coaxial, or co-linear, or even co-planar, I have found that the biggest contributor is often an error in the Morse Taper itself (I should note here that I'm talking about far eastern products, the government pension won't subsidize anything better). Several times I have found the taper angle to be minutely greater than spec, with the result that while it can be tightened into the socket, there is the possibility of a very slight misalignment and all the consequential errors. Furthermore, the dependence of the mating being solely reliant on a contact area near the nose of the socket leads to risk of the slightest 'ding' having a disproportionate effect.

It would be well worth doing some bluing tests.

Roger

Hi Michael,

What model of lathe, and what type of ER collet chuck mounting are you using.

Is the chuck taper mounted or face mounted, if taper mounted, either re-machine the taper or get some good quality tough steel and make a new one true to the axis of your lathe spindle.

If face mounted, machine the register down, so you can tap-adjust on the mounting bolts/screws.

These are all practices I've ended up having to do to get that last 1/4 thou on ground bars, such as gudgeon pins or MT straight sockets, cold rolled bright bars are not round, even silver steel can be deceptive.

Regards

Bernard

Hi Michael,

I would seriously think about dumping the MT4 mounted ER40 chuck and replace it with a backplate mounted version. This will actually allow you to passup to 30mm diameter stock into your spindle. It will also give you the controls needed to set the accuracy of just how true your backplate mounted holder will be.

I haven't found any lemons yet Michael. I have a set of Vertex ER16 and a set of ER25's I bought from the USA because at the time there were no precision collets on ebay in the UK. There is a need to be careful. They sometimes spec run out on the bore of the collet which is totally meaningless in terms of true accuracy - just as it more or less is when people check the run out of the bore in a holder that way.

Reading the other thread my understanding of the correct way to measure run out of the collet involves measuring it a varying distances from the mouth of the collet. The distance being inversely related to diameter.

John

-

Apologies for not noticing your chuck is MT4, it does however as MalcB stated, limit your bar length/size, my comment on re-machining the taper refers to the ER taper, or as suggested make your own.

Bernard.