ER32 etc collets

[Tony RayParticipant @tonyray65007]

Having decided that ER style collets offer advantages of 5C I’d apreciate some advice on the following

I have a Boxford with 3MT in the headstock but I see a lot of the collect chucks offered with a threaded back plate to go on the spindle nose. It seems to me that using the 3MT is a neater solution – do you agree ?

 

There seems to be quite a price range between fleabay and the known makes for collets / collet sets. Can anyone recommend anything at the cheaper end or are they best avoided ?

 

I currently have no idea about what sort of tolerance I require in terms of run out.

 

Thanks

 

Tony

[Tony RayParticipant @tonyray65007]

[Jim WhetrenParticipant @jimwhetren72358]

Hello Tony, 

I agree, I use ER 32 colets on my lathe, spin indexer and dividing head, also on the mill until I changed to ER 20 here; to gain better access with the smaller closing nut.

 

I also use the ER 20 collets on a Worden grinder, as they cover metric and imperial shank cutters. I have also made a chuck to use them on a Cowells lathe. Making the chucks is quite straight forward, but I cheat and buy the closing nuts which are reasonably priced.

 

My lathe has a 3 MT which will pass 20mm using the ER 32 collet. If you use a 3 MT holder, you will need a drawbar and will loose this advantage.

 

My collets and closing nuts and a backplate fitting chuck for the lathe came from Arc Eurotrade, (usual disclaimer) and are of very good quality and also reasonably priced.

 

Go for it

 

Jim

[John Stevenson 1Participant @johnstevenson1]

One thing to bear in mind with ER collets are they are the only gapless collets that do both work holding and toolholding.

 

On a mill 4 collets will hold all 4 standard Imperial shanked cutters and all 4 standard metric shanked cutters.

 

Seeing as Arc have been mentioned I will say that they are the only people selling a spin indexer that holds both 5C and ER32 so you have the best of both worlds.

 

John S.

[DugsonParticipant @dugson]

Tony,

I have bought several sets of ER collets from CTC tools Hong Kong they have been excellent quality and accuracy. worth a try.

 

[mgjParticipant @mgj]

Chronos do ones by Vertex.

 

I agree about the  ability not to have to use a drawbar.  Being able to pass the rod through the chuck saves a lot of material, and means you can work on longer objects if need be – (and removes the temptation to leave long overhangs forwards, just to get the job into a holder?).

 

Runout – depends on the fitting I suppose. I have one for ER25 collets which screws onto the nose of the Myford. That was a commercial item, and typically the runout is about .001″. I have a chuck for ER32 collets which I put, very carefully onto a D1-4 camlock backplate for the bigger lathe, and the runout is more or less zero. (Normal method – drill clearance holes for the bolts. Soft face hammer till runout was not detectable. Tighten and then 3 dowells + loctite)

 

I suppose that they are fine for close work, but, if it has to be 110% right then you are back to the DTI, and  4 jaw or a Griptru?

[David BlunnParticipant @davidblunn56158]

I also have a set of ER32 from CTC tools in Hong Kong. I also bought a spare closer nut with the MT3 chuck. The mill uses the MT3 chuck and the spare closer (which is a standard item but tricky to make) goes on the lathe. MEW has had at least one article about making a chuck for screw on noses.

Good luck,

Regards Dave 

[Tony RayParticipant @tonyray65007]

Thanks to you all for your input. I hadn’t considered the disadvantages of the draw bar. I think that I’ll go for a backplate mounted chuck & collets from CTC.

 

Regards

 

Tony

[KWILParticipant @kwil]

Further information.

 

My Myford S7 nose clocks 0.000025″, yes a barely discernable movement on a 0.0001″ division.

 

Using an ER25 collet in a commercial screw on adaptor, a 0.500 silver steel sample clocks 0.0008″ > 0.001″ total runout close to collet.

 

Using a Myford 2MT collet, the same test piece clocks 0.0001″.

 

The commercial screw on adaptor is  a lot more sloppy than I would want for precision use, I must make my own sometime to see how much better it would be.

[Hugh GilhespieParticipant @hughgilhespie56163]

Some advice needed please. I have fitted an ER32 collet chuck to my Boxford VSL500. This lathe has a L00 taper spindle fitting and finding suitable backplates isn’t easy. Anyway, I got a backplate from Rotagrip and a collet chuck from MSCJL. I machined the register as best I could but as my machining ability rating is AC*,  needless to say it’s a tiny bit too small. Can’t actually feel any movement when the chuck is on the backplate but the TIR in the collet socket is about 1.5 thou. The runout on the backplate register on it’s own is about 0.2 thou and as the lathe spindle runout is the same I guess the register is nice and concentric, just that little bit too small!

 

So, do I have to live with this or is there anything I can do to improve matters? I have considered cutting the register right back and machining a sleeve but then I will need to make two super accurate cuts, not just one. Another stray thought, could I run some silver solder around the register and then re-machine to size? I like this idea but as I have never actually silver soldered anything and this would involve heating the whole back plate to dull red heat it’s a bit daunting.

 

What else?

 

Regards

 

Hugh

 

* AC = Absolute Crap

[David Clark 13Participant @davidclark13]

Hi There

Can’t you face off the register and have another go?

regards David

[John Stevenson 1Participant @johnstevenson1]

Why not just tap it true and tighten up ?

 

Cylindrical grinders are made with an undersized spigot just so the chuck can be tapped true with a lead block.

 

John S.

[wheeltapperParticipant @wheeltapper]

tapping it true works for me too.

this collet holder I made

fits on the spindle flange.

I made the recess a mite oversize so I can true it up .

 

Admittedly it would be nice to just bolt it on and get it right straight off every time but who am I kidding.

 

Roy

[chris stephensParticipant @chrisstephens63393]

Hi Hugh,

I second John S’s answer and add that you can take out wobble with careful application of cooking foil selectively placed between chuck and backplate . Crude but can be very effective, the voice of experience has spoken.

Some might suggest that you could grind out the bore taper but I would not recommend this as you might end up with  a misalignment with the taper in the closing nut.

chriStephens 

Edit; As Wheeltapper says wouldn’t it be nice if things just fitted perfectly first time every time, but  …..

We call ourselves model ENGINEERS, so, engineering  an answer is the order of the day.

Edited By chris stephens on 28/08/2010 17:32:54

[Hugh GilhespieParticipant @hughgilhespie56163]

Tapping it true sounds like a plan. I feel much more confident using a hammer than a lathe!

 

Just to be sure I understand what’s needed, I assume that I need to do the following:

 

Cut back the register on the back plate so that it misses the recess on the chuck by say 25 thou all round. Then the chuck is free to slide around on the surface of the back plate.

 

Clean everything up, assemble with bolts just nipping.

 

Fit to lathe spindle and set up with DTI reading from the ER32  taper in the chuck.

 

Rotate and tap until minimum TIR then tighten bolts and recheck – repeat as required until minimum TIR with tight bolts.

 

Noe that I have a slight problem with doing this as I have to take the complete assembly off the lathe to tighten the bolts – being lazy, I used the existing M10 tapped holes on the chuck. To use these I had to use socket cap head bolts – not enough clearance for hex head bolts between the spindle nut and the bolt heads – and there isn’t enough room to get an allan key in to tighten. Not a huge problem as I guess I can reduce the head size of three M10 hex head bolts to fit and then I can tighten with a spanner.

 

Sorry to make such a meal of this but it’s all a bit new to me – well completely new to me to be honest.

 

Thanks, Hugh

[John OlsenParticipant @johnolsen79199]

If the register on the backing plate is already too small you should not need to take any more off it. Since it would have been turned while on the spindle, it should be true to the spindle, just a little undersize. So the main problem is to get the body part aligned, then tighten up the screws, which will all have to be done with everything in place. Usually start with the screws just nipped, and as you tap and the TIR comes down, you can gradually tighten them up.

 

So the lack of room is the main problem. Sound like the hex head screws with a reduced head size would be one answer. If there is any room at all you could consider a modified Allen key. I have one (for M3 screws) that was made by silver soldering a very short length of Allan key into a flat shank. (The hole is round, the silver solder takes up the difference.) The shank is about 1/16 thick and about 1/16 of key sticks out. So that one can go into a screw when there is only 1/8 between the screw head and an obstruction above. I think I actually used a longer piece of Allan key, then ground it back flush on one side and to the desired length on the other. For M10 screws you might want to scale up a little, but I think we could make an Allan key to go into a 1/4″ gap OK. You can also shorten the arm on a standard key, but the radius of the bend limits how far you can usefully do this. 

 

I agree with the earlier remarks about the Morse taper, when I use my ER16 set on the Myford the holder has the morse taper, and used with the drawbar, limits the length of stock back inside the collet to about 3 inches. This can be annoying. On my Unimat, the holder uses a screw on backplate, and any length of stock can protrude out the back…much more convenient. Making new  holder for the Myford is on the “maybe someday” list.

 

regards

John

[adyParticipant @ady]

Just lap and polish it, don’t micro cut it.

 

Sparey:

“Make haste slowly”

 

There’s a boring job in one of the newnes  books which took at least two days to complete, and that was on the best gear available in the 1950s, when Merlin engines were becoming obsolete.

 

Unless you’ve got a multi-million pound setup you have to take account of the limitations imposed upon your situation.

 

Cutting is great when it works first time, but unless you plan on running hundreds of them off, it’s only an option, especially with a one off collet scenario.

As far as micro runout is concerned, as soon as the bar heats up or you put cutting pressure on the workpiece that 0.001 runout  becomes irrelevant…actually it’s pretty horrible how much that metal actually moves when you poke it gently with a forefinger…

I’m currently cutting a collet chuck for a huge pile of hardinge collets I got from ebay for a tenner…and it’s going to take a while.

 

If all it takes is a collet and a bit of roundbar to make an accurate part then the British car industry should still be thriving!

Edited By ady on 29/08/2010 01:14:30

[John Stevenson 1Participant @johnstevenson1]

Posted by ady on 29/08/2010 00:45:50:

As far as micro runout is concerned, as soon as the bar heats up or you put cutting pressure on the workpiece that 0.001 runout  becomes irrelevant…actually it’s pretty horrible how much that metal actually moves when you poke it gently with a forefinger…

I’m currently cutting a collet chuck for a huge pile of hardinge collets I got from ebay for a tenner…and it’s going to take a while.

 

If all it takes is a collet and a bit of roundbar to make an accurate part then the British car industry should still be thriving!

Edited By ady on 29/08/2010 01:14:30

 .

 

So you are saying we should throw all the lathes away and just grind and lap everything ?

 

Whatever…………..

 

John S.

[KWILParticipant @kwil]

If Ady’s metal moves when he pushes it with his finger, perhaps HE should throw his lathe away??!

[JasonBModerator @jasonb]

And surely a lap will just follow the out of centre bore. Also and expansion due to temp rise will be concentric and you would also want to get your bit of say 1″ steel upto around 100deg C to see 0.001″ movement

 

J

[John Stevenson 1Participant @johnstevenson1]

Posted by ady on 29/08/2010 00:45:50:

 

There’s a boring job in one of the newnes  books which took at least two days to complete, and that was on the best gear available in the 1950s, when Merlin engines were becoming obsolete.

 .

 

Well thank Christ it took less than two days on 1940’s equipment when we had our arses to the wall.

 

Couldn’t have been that hard because we let the yanks build them as well !

 

John S.

[Ian S CParticipant @iansc]

The “yanks” used production line methods to do the Merlin  build, machine ground crankshafts instead of hand lapped etc.Ian S C

[chris stephensParticipant @chrisstephens63393]

Hi Guys,

I think we should ask ourselves why we want to use collets. 

The advantages are;

1/ Repeatable concentricity, but this is negated, if we machine the whole item at one setting.

2/ Better grip, there is no substitute for a firm grip, especially when using a large, coarse threaded die. 

3/ The ability to use ground stock without having to waste material by using the next size up, to get the size we actually want.

I shall stop there and let yu’all add to the list.

chriStephens 

[John OlsenParticipant @johnolsen79199]

On point 1/, it is kind of hard to machine the whole thing at one setting when it needs work on both ends. So collets let us maintain a good level of concentricity when we need to reverse the work.

 

Another point, collets are sometimes good for putting centres accurately into work we wish to turn between centres.

 

Another, Collets don’t tend to mark the work as much as chucks do. 

 

With some types, we have very low overhang of the job from the bearings.

 

regards

John

[chris stephensParticipant @chrisstephens63393]

Yup, granted.

I could have phrased 1/ better, I meant it as follows, ‘the advantage of a collet is negated, as even a cra**y three jawed chuck is good enough if you do all machining at one setting’. 

chriStephens 

[Author]

Posts