I love my Tormach, but...

[philbur]

Is the motor not selected to be run with the VFD? Where did you find the design criteria for the motor maximum rpm?

Phil

PS: the rated rpm is not the design rpm.

It is a 1750 motor and at the design rpm it works fine, its the VFD that puts the rpm higher than its designed for.
mike sr

[SCzEngrgGroup]

It is a 1750 motor and at the design rpm it works fine, its the VFD that puts the rpm higher than its designed for.

You have no factual basis for making that statement, and I'm quite certain Tormach, being a company that does very thorough engineering on their products, would NEVER operate the motor outside its manufacturers maximum ratings. To do otherwise would expose them to a HUGE potential liability were one to fail at high speed. 1750 RPM is undoubtedly the *base* motor RPM, measured at a 60Hz frequency, which tells you absolutely NOTHING about its safe maximum RPM capability at higher frequencies. Inverter-rated motors are *designed* to be operated well above their base RPM, typically by 2-3X or even more. Were this not the case, there would be no reason for every inverter sold to be capable of running at least 120-200Hz or more.

Regards,
Ray L.

[riabma]

We are talking about the 1100, but what about the 770? Anybody has issues with unbalanced pulleys or similar?

I got this picture of the 770 in the Tormach site. The pulleys look like they are machined in aluminum instead cast iron like the 1100.


Ricardo

[mike sr1]

I took off the post in question as I sure dont want to start a war over it.

I will say that the Tormach does make a fine machine for the price period, and I love mine!

[philbur]

With 10,000 rpm max I think the 770 is a different kettle of fish. I believe specific balancing requirements where needed. Have a read in the 770 technical info, I think even the TTS had to be balanced for 10,000 rpm.

Phil

We are talking about the 1100, but what about the 770? Anybody has issues with unbalanced pulleys or similar?

I got this picture of the 770 in the Tormach site. The pulleys look like they are machined in aluminum instead cast iron like the 1100.


Ricardo

[TurboStep]

Spinnetti
If you just got your machine set up then you definitely don’t want to start worrying about problems that you may not even have, and which may not even bother you anyway.

Philbur
First of all, I’m not having fun at all. The finish I’m getting is, in my opinion below the real capabilities of the machine. Maybe I just have a Monday machine but other posters appear to be telling similar stories.
Am I going OTT? I’m just reporting what I’ve discovered and the results of my work. If someone else can benefit from this, then that’s what the forum is all about.

Riabma
Greg wrote “We balance the 770 which needs to run at 10,000 RPM, but we’ve never had a balance problem on the 1100 machine.” Which I guess answers your question.

The first results are promising but I still have some further testing. I’ll post the final results at the weekend – regardless of whether they turn out good or bad. Then I’ll just be a sitting duck and you can open fire

Step

[David Bord]

Some great work on this thread! I love all the analysis and testing.

The motor fan looks like a pretty simple piece. Couldn’t you just machine a new one out of alum that could be designed so that it should require very little if no balancing?

Maybe to simplistic, but another thought is balancing the parts like you would RC helicopter blades or a motorcycle tire/rim with having the pulley sit on horizontal bar with low friction bearings? Spin and watch where the pull stops and add and remove weight till it doesn’t stop in same location.

How about chucking the pulley in a lathe and just trimming a tiny amount off the casting where the belt doesn’t ride, so the thickness is consistent. Have no idea if this would impact its strength though.

David

[btu44]

Turbostep,
Great idea for the poor mans piezo sensor. I think I'll give it a try when time allows.

Here is a short vid I did to find the RPMs to avoid during a finish cut.



This is a vid of the deflection of the Z axis way. Granted this is a pretty heavy cut that would exaggerate deflection, but it still may have an effect on the surface finish.

http://www.youtube.com/watch?v=TeFR0ac0QoM&feature=plcp]PCNC1100 Vertical Colum Defelection During Heavy Cut - YouTube

[philbur]

I get the same pattern at the top end spindle rpm. The numbers are:

500 mm/minute
2 lines/mm (exactly 20 artifacts counted over 10mm)
5,294 rpm (measured)
9.525 mm 3 flute cutter (3/8")

based on cutter diameter the depth of the artifacts is of the order of 0.01mm (Measured on a CAD drawing, to lazy to calculate).

Converting to imperial and using your formula:
50.8 lines/inch
X 19.7 inches/min
x 5.4
= 5,404

Which is close enough to be a match for the measured rpm of 5,294.

So the numbers do appear to suport your theory.

Phil

If you think you are seeing this effect, estimate the number of waves per unit (e.g. 65 per inch) and multiply this value with the feed rate (e.g. 12 in/min). Multiply this result by 5.4. If my theory is correct then this result should be very close to your spindle speed (preferably measured with a tacho).
Step

[TurboStep]

As promised, the second part of my “balancing act”.
First of all, thanks to Phil for his measurements. David suggested machining a new fan out of Alu. I thought of this as well, but only if the end results weren’t good enough. I can’t take credit for loudspeaker idea because this came from a paper written by Miklos Koncz http://www.zmne.hu/aarms/docs/Volume...pdf/12konc.pdf. His mechanical solution is much more refined and he also wrote software to evaluate the results using the parallel and game ports of a PC. While his system looks really interesting I just borrowed the sensor idea. One thing I didn’t mention is that the speaker cone is like a reference surface. It must remain stationary while the remainder of the speaker vibrates. I simply rested two fingers lightly on the cone, this turned out to be more than sufficient.

I had suspected the spindle pulley to be the largest contributor to the vibration due to the fact that I had measured a significant deviation in the thickness of the outer section and because it simply has much more mass at a larger radius than the motor pulley. To get a better idea of what was going on I decided to strip the paint. As I mentioned in an earlier post my pulley has a dent on one edge and it appears to have been painted twice: maybe once before the dent and again after?
Now, without the paint, the thickness deviation around the outer edge is 0.10mm. A hole had already been bored on the underside at approximately the same position as the thickest part of the wheel (first photo – the rust came after the paint was removed). I can’t say how accurate the resulting balance was, or whether it was balanced with, or without a key in the keyslot. The dent can be seen quite easily without the paint and with the edges beginning to rust (second photo). This pulley isn’t going back on my mill!
As I had never used the slow pulley position I initially contemplated removing the entire outer flange, just leaving the smaller section for the high-speed setting, but the first step was to try to balance it in the same manner as the motor.
I also had a new spindle pulley that was delivered with a total of 4 balance holes, each smaller than the hole in the original pulley, but all of the same size and depth. 3 were on one side and the fourth was approximately diametrically opposite. This means 2 holes would roughly cancel out. Again, I drilled and tapped 3 pairs of holes at 120° in this new pulley.
With the whole machine re-assembled I tried to improve the balance by adding grub screws as before. That wasn’t enough so I drilled out my threaded holes on the heavier side to twice the original depth. This wasn’t enough either! Then I noticed the hole in the spindle flange. This is the big “nut” supplied as part of the power drawbar kit that screws down onto the pulley. The hole for the set screw is quite deep and the set screw itself is short and hollow with a small brass plug at the end. I made a longer, solid set screw as described in my earlier post, which re-filled the hole (third photo – by the way, that’s a Fenner belt) and I was then able to achieve a balance.

It’s difficult to tell but I would say that the motor/spindle is less loud now (by no means quiet). I had intended to make before and after noise measurements but I forgot – and I’m not going back! Yesterday I was running a finishing cut at 5000RPM and realized that I could hear the stepper motors running. I don’t think I’ve heard them at these speeds since the Series 3 upgrade.

This whole thing is about finish quality and I can say that the improvement is significant. The 4th photo shows a before/after comparison at 5000 RPM (measured), 0.1mm depth of cut, same stock material and using the same 4 flute carbide end mill still mounted in the same tool holder. The wave effect is gone (the difference is more evident with the naked eye). Obviously there are tool marks but that’s not unexpected. Some have reported that the finish is acceptable at 4000 RPM or below. The 5th photo shows the same test but performed at 4000 RPM (measured). Although the wave effect isn’t very visible in the “before” sample the “after” result still better.
Probably the poorest finish that I have ever produced (on any mill) was on the chamfer shown in photo 6. This is a 2x2mm chamfer made in one pass with a ¼” Maritool Mill/Drill. Other than the fact that the first chamfer was on a 3 3/8” boss and the second (photo 7) was a straight edge they were reproduced under identical conditions. I would have liked to have shown the tool marks on the second image but they were so fine that I couldn’t photograph them. You will also notice the tool marks on the vertical face of the second photograph made by a 2 flute end mill. These marks correspond to the spindle frequency and indicate that I had some runout and was cutting with only one flute. I couldn’t expect the finish to be better than that!

I believe that face milling results are also better but I had previously removed my column shims to make sure that they weren’t affecting the rigidity of the machine, and after restoring them to the original values I still have a (very) small tramming problem. I’m not sure whether the tramming is very slightly worse than before or whether I wasn’t able to see it because of the vibrations. I would need to re-adjust the mill before I can deliver conclusive results but that’s not currently on the top of my priority list.

The TTS Set Screw holders have a balance dimple for speeds above 8000 RPM. When I compare the size of the screw on the fan and the hole in the drawbar spindle flange with the size if these dimples then I’m not surprised that I was getting measurable vibration at 5000 RPM. I was able to improve the total imbalance of the motor/pulley and the spindle/flange/pulley combinations by adding/removing weights to pulleys but it’s hard to say whether the pulleys themselves were really out of balance (they are balanced to some degree as supplied). I also only compared the finishes at the very end and I don’t have a comparison corresponding to just a balanced fan and spindle flange. This might be an interesting test for anyone who doesn’t want to go OTT like me!
At the moment I’m happy with the finish that my machine is now giving me so I’m not intending to make any further tests but I may conclude with a surface machining update after I’ve re-aligned the mill – that’s going to take me a while.

If anyone does attempt to replace the grub screw in the spindle flange then don’t do it the way I did! As I added screw-lock and tightened it down for the last time I could feel the spindle thread getting crunched. I guess I won’t be taking the flange off again in a hurry but a little piece of brass in there sure would have helped.

OK, fire away, I’m a sitting duck!

Step

[tbaker2500]

Excellent results! Thanks for posting such a detailed report. I'm glad to hear that this did help the finish. It'll probably help lots of other things last longer, too.

[mike sr1]

Turbostep,
you have a pm, I am not sure it went thru OK

[Praymond1209]

I just got my used 1100 and feel that the vibration in the spindle is not good so I took the Motor rotor with the pulley and fan attached and balanced it in our Schenk balancer at work.
It was not out much but I will report if it made things better.
I balanced to 4W/N

[mike sr1]

I just got my used 1100 and feel that the vibration in the spindle is not good so I took the Motor rotor with the pulley and fan attached and balanced it in our Schenk balancer at work.
It was not out much but I will report if it made things better.
I balanced to 4W/N

I am very interested in your results!
I finally got a reply from the motor shop, its 82.50 per hour to balance an armature, fan and pulley assembly. he said appproximately 2 hours to balance it.

He also said some motors say inverter duty on the nameplate. He also said that basically means the insulation is better.

[TurboStep]

I'm by no means an expert but would image that a large part of the 2 hours is required to set up the motor in the equipment and to determine how much a known change affects the balance. The time required for a second and following motors is likely to be much shorter (but then most of us only have one mill). I spent some time with intertial navigation equipment and if I remember correctly we were able to balance the rotating gyro wheel assembly in about 8 minutes but the equipment was specifically designed for these wheels.
Step

[mike sr1]

I dont know how these will show up, smallest resolution the camera has.

20 IPM
.005 DOC
max RPM 5100
4 flute .480 carbide, a reground .5
It has horizontal marks from the cutter flutes, the vertical marks are almost nonexistent, smooth enough to show fingerprints and the finish is good enough for me.

[philbur]

Whats the material?

Phil

I dont know how these will show up, smallest resolution the camera has.

20 IPM
.005 DOC
max RPM 5100
4 flute .480 carbide, a reground .5
It has horizontal marks from the cutter flutes, the vertical marks are almost nonexistent, smooth enough to show fingerprints and the finish is good enough for me.

[mike sr1]

Whats the material?

Phil

Its a piece of scrap aluminum, cuts like 6061 t6

no coolant, climb milling.

I did one at 50ipm and one at 80 ipm, both look good, the vertical cutter marks are showing a bit at 50 more so at 80.

[Praymond1209]

I did balance my rotor assembly, FYI, setup was10 min, balance was 15 min. As I stated earlier it was not out much and I balanced it to a very tight standard, 4W/N.
Results:
The top speed vibration was improved, I still can hear and feel a "tick" at low RPM's only in forward, it runs smooth in Rev ???
I am going to run for awhile and see how finish is.
One more thing, when I tighten the SHCP and washer to hold the pulley on it bottoms out on the shaft prior to holding the pulley tight against the spacer and shoulder on the shaft. I will make a washer relieved to clear the shaft and see if that helps.
Thanks, Pat

[philbur]

What have you changed in order to achieve this improvement?

Phil

I dont know how these will show up, smallest resolution the camera has.

20 IPM
.005 DOC
max RPM 5100
4 flute .480 carbide, a reground .5
It has horizontal marks from the cutter flutes, the vertical marks are almost nonexistent, smooth enough to show fingerprints and the finish is good enough for me.