Spindle Runout Issues on PM-728VT

[higgite]

I have the same mill. I have measured my spindle taper (assuming reference is to lowest part of spindle), and mine looks to be less than .0001" runout.

Sorry, I should have been clearer. I meant has anyone measured the spindle taper runout on the OP's machine that is in question.

Let me rephrase that.

Netmagi, if it was covered earlier, I apologize for overlooking it, but have you measured the runout of the spindle taper itself? (See step 3 in pontiac428’s post #11) Or is that a moot point at this stage?

Tom

 

[rodm1]

I think, could be proven wrong, that Starrett is now made by Mitutoyo....

Starrett Factory Tour!

Come along for a one hour tour of the Starrett factory in Athol, MA where they make everything from calipers to micrometers to levels and more! A big thanks to Starrett for allowing us to film and to our wonderful machinist tour guide and host Bob! ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Links for this...

yewtu.be

 

[netmagi]

I care a lot, trust me, I want everything to be perfect, every time.

So we will assume that the spindle top bore of the R8 is at the Max tolerance, .9506"
And your measurement of .9486", max possible difference of .002"
Which assume they are centered, .001 on each side at the top from runout.

So if its pushed to one side up at the top, if you measure the runout of the test bar at equal length from the spindle face, about 4" down, it should be equal to what its pushed off center up top. (If that is what is happening)

If that was the case, no matter what the taper, or even if there was no taper, with that top bore holding it still, it should not be able to move more than the .001" off center.

I will ask what the taper spec is on those. I've seen a lot of different info. That is a good point if it only looks like its contacting at the end.

Thats assuming everything else is dead on, which it seems to be. But you need a long indicator to reach up inside of that spindle to get that internal bore.

Also - When they said to sand inside of that bore, that is to remove any burrs left if there are any, no to enlarge it. If they said to enlarge, thats not correct, but on some of the Chinese mills over the years we have had burrs left on the bottom edge of that top bore which the R8 fits fine at first, but then later on the burr sort of folds up from that ledge and tooling gets stuck or tight.

Some more things to check, take out the set screw key in the R8 so you can put the test bar in any way you want. Put it in where the key is first and record what you get, mark it, then flip it around 180 degrees and see if that repeats or if it changes.

You don't have to do any of that, we will have a spindle assy coming for you after its assembled and checked out so its an easy swap, but I am just curious. Once that gets to you and its all settled, I will also check it out here later on

RE: measure of upper bore in spindle (that the R8 shank fits in):

It is indeed difficult to measure in that bore, but I was able to reach it with a telescoping gauge, and it's 24.126 (.9498). Set screw is not installed. Didn't want to take any chances on anything keeping the tapers from seating.​

From my understanding of how an R8 taper/shank works:

• The taper is responsible for aligning the tooling/collet with the rotating axis of the spindle, and providing the friction necessary to turn the tooling.
• The drawbar is responsible for both pulling the tooling/collet into taper seat, as well as keeping it there, but the friction to drive/center the tool still happens at the taper.
• The upper bore where the shank of the tooling/collet slip into limit how far the tooling/collet can 'lean' in the event that cutting forces (crash?) are soo high that the drawbar stretches (we're talking tenths here), allowing the taper to unseat. Ideally the upper bore is just "slightly larger" than the shank to allow easy changes, but to limit that movement if things get wild.

Imagine a scenario where the upper bore was an inch larger than the shank. The drawbar would still seat the tooling/collet and everything would run true and probably be fine for 90% of work, especially in a mill of this size. But now imagine that we crash a 3/4 HSS endmill into the side of a vise. If the drawbar stretches and allows the taper to unseat, that collet could lean way over at the top, potentially bending the drawbar where the threads meet the collet/tooling, making it impossible to even remove the drawbar/tooling.

As long as there is a small amount of clearance 360 degrees between the upper bore and shank, the taper is what's aligning the tooling/collet to the spindle axis.

 

[netmagi]

What you're seeing is simply a pivot point with a single point of contact. The fact that you tried other things like collet chucks and none of them fit differently than the test bar suggest the taper of the spindle is incorrect.

This thread has been fascinating; I am sorry it came at the cost of time delays and frustration. I have learned a very important lesson about R8 tooling in general. I have R8 shank tooling from many origins. I have made the naive assumption that they were either close enough or, being a taper, were tolerant of variance. I intend to measure and blue my tooling to assess the fit and open my third eye to all of the inputs of variance.

It makes much sense to look for problems suspecting the cheaper part (the shank) and to trust the expensive part, at least initially. It follows the principle of parsimony that the correct solution is usually the simple one.

Yep, I just assumed the taper angle must be correct and wasted a lot of time chasing other causes. Even early on when I saw a poor result bluing, I thought it was a concentricity issue with the upper bore.

 

[netmagi]

Sorry, I should have been clearer. I meant has anyone measured the spindle taper runout on the OP's machine that is in question.

Let me rephrase that.

Netmagi, if it was covered earlier, I apologize for overlooking it, but have you measured the runout of the spindle taper itself? (See step 3 in pontiac428’s post #11) Or is that a moot point at this stage?

Tom

Yes, this is the first thing I measured on each. It's excellent! First spindle, at the bottom of the taper is less than half a tenth.

2nd spindle is about 3/4's of a tenth (which is still really good).

This measurement alone though doesn't tell us much. You're always measuring a combination of axial and radial runout. You need to measure at multiple distances from the axis of rotation to see how much you have of each.

For example, let's say you measure 1 tenth of TIR at spindle nose, then you again measure 1 tenth of TIR 5" down from nose at the same point in rotation. In that scenario, you'd have ZERO axial runout, and 1 tenth of radial runout. It's never that simple though. Axial runout will get worse the further you are from the axis of rotation, and radial will still the same.

I'm having axial runout which gets worse the further down I go, likely due to the taper only touching at the bottom and the tooling/collet leaning out of concentricity from the spindle.

 

[netmagi]

Yes. It measured less than .01mm right at the spindle, and was .02mm (.0008) 150mm down from the spindle (Max there is .03mm). That is within spec.

This is huge, and I don't want it to get "lost in the thread". Earlier on I was complaining that the "max inaccuracy" on the 728's spindle wasn't defined, and we were discussing what good should look like for a mill in this class, but it's now defined:

.03mm / 0.0012" TIR @ 150mm / 5.9" from the spindle nose.​

​

Matt, in my opinion for a mill in this "class", that's excellent, and I will be very happy when mine is running within that spec. Thank you for sharing that. It's no wonder the 728 is so highly regarded, and I'm guessing the tapers on the two spindles I have were just ground at the wrong angle.

I noticed all the other numbers on the drawing you shared were removed. What do they define as the "target" angle for the taper grind? I really hope they didn't intentionally grind the taper wrong to accommodate the $25 poorly made collet sets and $30 chucks China floods us with like the design change they made to the upper bore. At least their change to the upper bore shouldn't really effect much, but the taper will.

 

[verbotenwhisky]

Starrett Factory Tour!

Come along for a one hour tour of the Starrett factory in Athol, MA where they make everything from calipers to micrometers to levels and more! A big thanks to Starrett for allowing us to film and to our wonderful machinist tour guide and host Bob! ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Links for this...

yewtu.be

I did say I could be proven wrong... LOL

 

[qualitymachinetools]

All I meant was that if the upper bore is concentric with the taper, the taper does not matter, if there was no taper and it was a straight bore at the bottom with that same arrangement, it would be the same. That's all. I'll measure it when it gets back here too. The only way I've found to do that is with a probe type setup thats good to under .0001. I am waiting to hear what the taper angle spec is made to, for my own curiosity.

 

[G-ManBart]

I keep coming back to this thread and reading pretty much all the posts again. I just looked more carefully at the schematic showing the taper and noted the lines where the two tapers should be parallel don't really match up well at all, and the only point of contact matches what we see with the OP's test bar. Maybe that's a coincidence, maybe not.

Just out of curiosity I did a quick check on my BP last night. It had an imported R8 collet with a 3/4" end mill mounted, so I measured the top of the end mill just below the collet. From all the way up, to all the way down I was getting from around .0004 to just under .001 which seems reasonable considering it isn't a test bar. I'm going to measure the spindle itself next, but I wanted sort of a worst case scenario for starters.

 

[qualitymachinetools]

Does not matter as in this simple test - It has to be tapered and that matters, I just mean for the concentricity that I am talking about.

But if you are getting .9498 in that top bore, and .9486 at the test bar, thats pretty close. Definitely something going on, I just do not know exactly what yet.

What he says about the drawbar does matter too, but you would feel it when tightening. If it was off enough to make it runout, it would be tight going in.