What's new
What's new

ER-25 collet chuck, 2MT runout question

bikemutt

Aluminum
Joined
Mar 5, 2022
I installed an ER-25 collet chuck into a 2MT spindle, and chucked a 3/8" gage pin to check runout at the business end. The collet chuck claims 0.0002" concentricity.

The spindle, on it's outside, measures about 0.0002" TIR.

The gage pin, installed an a different ER-32 chuck also measures about 0.0002" or better.

I was disappointed to observe a TIR of about 0.0025" midpoint on the gage pin's stick out of 1" or so.

Thinking maybe I have a crappy ER-25 3/8" collet, I tried a few others (with different gage pins); same thing. Tried a different collet nut; same thing.

Then I tried something I didn't think would make much difference but, in the interest of beating myself up thoroughly, I marked the spindle and the chuck with a sharpie and proceeded to install the chuck at 90, 180 and 270 degrees.

At 90 degrees, TIR moved up to mid three thou :(. At 180 degrees, TIR moved to just under a thou :). At 270 degrees, TIR moved to 2 ten thou :D.

I don't have a test indicator presently so I can't check the inside of the spindle's concentricity. Regardless, can a chuck's insertion point in a tapered spindle affect terminal runout as I described? If so, is there a faster, better way to converge on the ideal insertion point? I only tried 4 points, hate to have to try 360 of them, lol.
 
Sounds a lot like a "price point" issue.

Quality endures long after the price is forgotten.
 
Your 2MT spindle inner taper will have some degree of runout from when it was made. If there is a burr or embedded chip or dirt or rust inside the taper, runout could be much worse than when it was made.

Then there is the chuck, which was probably not perfectly made. Ditto the collets.

So first clean the inside of your spindle and try again. Then inspect the spindle for burs or embedded chips or rust. A reamer can help if that is part of the problem.

Decades ago, I bought some German-made Morse taper socket cleaners. They are tapered wood with thin leather strips glued to them and a handle on the big end. They do a great job of catching dirt and chips.

Another possible source of runout is a mismatch in the male and female Morse tapers. You will get random changes in runout as you insert and remove the chuck in the spindle. Or the runout will change as you do a radial bump against the chuck nut in different directions.

Larry
 
Last edited:
I'm willing to bet that you have an accumulation of errors leading up to that .0025" runout.
The collets could be perfect but if your morse tapers aren't perfectly aligned, and your chuck perfectly aligned then you're going to see more than the .0002"
 
If your first chuck runs out and you switch to another and it's good with the same pin, spindle etc. it has to be that chuck, no?

Get a DTI and indicate inside the taper on the first collet chuck, that'll show the runout for sure.

What brand is it? What kind of shape is it in?

My little Clausing 8520 has an MT2 spindle and I find using MT2 collets to hold end mills works the best. I also buy MT2 drill chucks, boring heads etc. rather than holding a 1/2" shank in a collet, helps bigtime with rigidity and runout.
 
Moden manufacturing out of a chucker it is not uncommon to have a wobble in a device ( ER-25 collet chuck) made that way. You may be obligated to mark your collet chuck and the spindle and then use it at the line-up.
Yes, could be an error on the machine spindle and in the collet holder causing the total error.

Having a lathe with dead centers you might use 3/4 X 5" piece steel and make your own #2 taper gauge. That would be an hour of work well spent. Kept in a nice wood box even one of mild steel would last for years.

That is why the likes of a Royal live center is so much higher priced..To get high accuracy with no wobble the manufacturing method for holders, live centers, bearainged shafts, most everything is very different than what is common nowadays practice.

Some companies quote .0002 (or What ) the way they hold it, and then properly checked it may be more.
 
I installed an ER-25 collet chuck into a 2MT spindle, and chucked a 3/8" gage pin to check runout at the business end. The collet chuck claims 0.0002" concentricity.

The spindle, on it's outside, measures about 0.0002" TIR.

The gage pin, installed an a different ER-32 chuck also measures about 0.0002" or better.

I was disappointed to observe a TIR of about 0.0025" midpoint on the gage pin's stick out of 1" or so.

Thinking maybe I have a crappy ER-25 3/8" collet, I tried a few others (with different gage pins); same thing. Tried a different collet nut; same thing.

Then I tried something I didn't think would make much difference but, in the interest of beating myself up thoroughly, I marked the spindle and the chuck with a sharpie and proceeded to install the chuck at 90, 180 and 270 degrees.

At 90 degrees, TIR moved up to mid three thou :(. At 180 degrees, TIR moved to just under a thou :). At 270 degrees, TIR moved to 2 ten thou :D.

I don't have a test indicator presently so I can't check the inside of the spindle's concentricity. Regardless, can a chuck's insertion point in a tapered spindle affect terminal runout as I described? If so, is there a faster, better way to converge on the ideal insertion point? I only tried 4 points, hate to have to try 360 of them, lol.

this is perfectly normal. dont expect too much from collets. most are pretty lousy. clocking in is the way to go for sensible work. you dont have to check 360 points. after three points you see in what direction tir is moving and after another two or three you have the the minimum most of the time
 
....
I was disappointed to observe a TIR of about 0.0025" midpoint on the gage pin's stick out of 1" or so.

When fiddly farting and trying to correct such problems do not check the pin in the middle.
Check as far in and as far out as you can. The difference between these checks help you find out what is going on.
One first thing if unknown collet is to mark it, remove and install at 90, 180, 270 and then back to the start orient. How did the runout change?

It is an assembly and all count. Single circle checks on any and are not great feedback. I so hate seeing this done that I could just scream.
In most cases your cutting tool is a cylinder not a circle in space.
Indicating the tip of your endmill means jack shit unless you are only making .005 deep cuts.

On super picky stuff we will put the holder in, put the collet in, and ID grind with attachment spindle to true as part of the job setup.
Take this holder off or change the collet/nut position and all is very bad. Basically the collet is throwaway with one job or maybe two uses.
Bob
 
At 90 degrees, TIR moved up to mid three thou . At 180 degrees, TIR moved to just under a thou . At 270 degrees, TIR moved to 2 ten thou .

So you need to line up your marks to the good ruinning spot always every time you use it, not all that uncommon.
 
On super picky stuff we will put the holder in, put the collet in, and ID grind with attachment spindle to true as part of the job setup.
Take this holder off or change the collet/nut position and all is very bad. Basically the collet is throwaway with one job or maybe two uses.
Bob

How do you compress the collet to simulate clamping forces? Shims in the slots?
 
At 90 degrees, TIR moved up to mid three thou . At 180 degrees, TIR moved to just under a thou . At 270 degrees, TIR moved to 2 ten thou .

So you need to line up your marks to the good ruinning spot always every time you use it, not all that uncommon.

Thank you all very much, what michiganbuck suggests is what I'm doing. I can spend lots of time in my head trying to figure out why, or just roll with what the tool likes; it's marked, that's where it goes.
 








 
Back
Top