Problem- This is NOT a normal noise

[Dancy]

Everything was fine. I had just done a test run of a new program with the spindle at 2000 rpm, things looked fine but when the program ended and the spindle stopped it sounded harsh. I placed the stock and started the program again and my spindle motor sounded horrible! Bad bearing, rotor scraping horrible! Emergency stop! Pulled the spindle motor loose from the tranny and tried 500rpm, bad noise! E stop again. Took the motor down and dissasembled it, polished the slight rust from the rotor and replaced the bearings even though them motor was smooth when turned by hand. All back together a week later and still nasty noise, sounded like somebody put a hand full of cannon grape in it. Took the motor apart again looking for signs of contact or scraping. Nothing. So I got the hair brained idea to try a very low rpm. It sounds like crickets in there. I have attached an audio recording of about 90 rpm motor (10 rpm) spindle. The inverter is having a fit and the really weird thing! No alarm codes.

Now the spec's A06B-6055-H106 drive and A06B-1006-B908 motor.

Now the questions: have you ever heard this noise before? Can you tell if the problem is in the inverter or the motor. How do I fix this (anybody have schematics/diagnostics?).

I have two theorys: One of the windings in the motor bad, or IGBT or its drive. Sounds to me that one of the three phases is mucked up. I will get my O scope out and look at each phase but I don't know what a proper wave form looks like (I hope I can figure it out) and then my ohm meter and check the windings.

Dan
Way out in Summerville, Oregon

[mc-motorsports]

If your checking the spindle motor, a scope won't help you. You can use a VOM and check phase to phase, should be low ohm, then phase to ground. You REALLY need a surge tester to check the motor. Then you can do a high voltage phase to ground and surge the phases and compare the wave forms. Doesn't sound like a motor to me, my experience is if there is something wrong with the motor, it will get hot or just won't run.

Sounds more like a bearing or something broke off and is rubbing inside? Can't really tell by listening to the .wmv

Spin it by hand as fast as you possibly can, use an allen wrench in the side of a tool holder or something. You should be able to spin it 300 RPM no problem with proper technique. That will tell you A LOT! Sometimes spining it real slow doesn't give away clues.

[guhl]

Hi
Try to take out the collet and feeding tube, leave the spindle alone and turn it on.

[Dancy]

The noise in the recording is from the windings in the motor, a magnetic field noise as the windings are energized from the drive. It is a fairly normal noise to be expected from a chopper type drive. What is not normal is that it is not uniform. The next question is what is causing the unevenness. I am hoping someone familiar with this motor/controller combination can give me some insight tracking down the source. Something along the lines of "Oh yea, those drives! Here is a list of the common failures we see in those. Yours looks like blah blah"

Musing:
1. Could a short in one of the motor field windings upset the drive enough to cause this? For instance, suppose each winding is 100 turns and winding 2 has a short between wind 62 and 70 creating a lower impedance and resistance, hence the drive sees an uneven load and as it tries to maintain rpm sends uneven phase signals.
2. Could be something wrong with the drive. Perhaps a filter capacitor is shot and the raw switched wave from is reaching the motor. Perhaps one of the IGBT drive circuits is not turning on the gate fully. Perhaps the feedback circuit is not sending the same information to each of the IGBT drivers.

Help? Help?
Anyone? Anyone?


Thank you!
Dan

[fanuservicetech]

One quick test you can do. On the circuit board A20B-1001-0120 (on the spindle unit A06B-6055-H106), locate a gray connector at the bottom of the circuit board labeled CN2. This connector is usually secured to the circuit board by 2 small screws on either side of the connector. Loosen the screws but do not disconnect the connector yet. Start put the control and command the spindle at about 1500 to 2000 rpm. Go back the the spindle board and disconnect CN2 connector from the board (yes while the spindle is running). Machine will go into alarm state and spindle will freewheel.
If you CAN still hear that noise, it is mechnical (motor bearing etc.).
If you CANNOT not hear the noise anymore, it is electrical i.e. drive unit misfiring or motor windings or feedback sensor problem. This test will not pin point you to a particular cause BUT will point you to whether it is mechnical or electrical issue.
Good luck.
IMPORTANT: DO NOT CONNECT the CN2 back in with control power turned ON. Power down the machine and then reconnect the CN2.

[Dancy]

Good to know fanuservicetech! I will give that a try this afternoon.

mc-motorsports, I looked up the theory of a surge testor, will build one this weekend and give it a "shot"!

[mc-motorsports]

Originally Posted by Dancy mc-motorsports, I looked up the theory of a surge testor, will build one this weekend and give it a "shot"!

Wow, I didn't realize people were building thier own! The one I used to use to check spindle motors cost about 20k, I think a spindle or motor repair shop charges between $25 and $75 to test a motor.

And yes, a "surge tester" will tell you if you have a short in pole #2 between windings 62 and 70. The surge tester will put in a 2000 volt pulse and it will "echo" and create a wave-form. You compare 1&2, 2&3, 1&3, if one is bad, the wave-form will look totally diffrent. I'm not an electrical engineer, but a surge tester will verify the motor and there's no science behind it, just compairing the "echo" (surge test) between the phases, then a 2000v phase to ground and low ohm phase to phase.

[Dancy]

Well MC, the pile of junk I will toss together will not be as polished and easy to use as a professional version but the theory is strait forward. I have a digital storage ocilliscope, several HV capacitors and a power supply I can charge them with so we shall see! I did check the wave form from the drive last night and all three phases are the same. It looks like a pretty good chopped sine wave where sine freq=rpm and chop freq=power. If I had a problem with the drive I would expect one of them to be diffrent that the other two and I did not see that so my attention is back to the motor.

Dan

[mc-motorsports]

That should work! It always confused me why the damn things are so expensive when they seem fairly straigh forward and simple, it wasn't very fancy, but well built, the shop had a couple of them, made by Baker.

If one of your phases are bad on the motor, the wave form will be WAY diffrent then the other two. If all three are diffrent??? Either means your doing somethng wrong or 2 phases are burned.

[Dancy]

Well the home made "surge testor" worked fine and indicated a bad winding. Just got the stator back from the rewind shop and it works $425 better

Problem solved!
Dan

[mc-motorsports]

Originally Posted by Dancy Well the home made "surge testor" worked fine and indicated a bad winding. Just got the stator back from the rewind shop and it works $425 better Problem solved! Dan

Excellent! So how did you make your own surge tester?

MC

[Dancy]

470uf electrolitic cap + old computer power supply + digital storage scope.

cap from scrap box
power supply from junk yard
storage scope from ebay for $125

hooked the scope across the winding in question
then I just charged the cap with the 12v supply
and discharged it through the windings
captured the wave form stored it for a ref
then compared to the other two windings. (the difference was obvious)

granted my pile of junk can not increase the voltage across the cap until the windings fail (true surge testing). First off, I do not want to do that to one of MY motors! I know from my experience that once you create a path for an arc it is much more likely to arc again and at a much lower voltage (in other words, you probably ruined it). Second I am not interested in the high voltage failure point of my motor, I just want to see if there is a short in one of my windings (different impedance).