Servo to encoder tuning

I have a 1990 Sabre 1250 with an 850MC with an x-axis servo motor that I put new bearings in. I marked the coupler location for the encoder and put it back together successfully however since I've had this machine i've always notices a high pitch coming from this servo when its just sitting still, the y and z do not do this so this leads me to believe the encoder could be tuned in better. After doing the bearings I turned the machine on and loosened the 4 bolts that hold the encoder on and push the slack one way or the other and I can change the tone of the noise so I'm sure there is a procedure for tuning the 4 potentiometers on the encoder. I've been searching but haven't had any luck finding anything yet. Does anyone here know the procedure or have a document they could send my way. Any help would be great.
Thanks, Joe

Is this a Brushless servo? If so it most likely has the commutation tracks on the encoder and these have to be aligned with the stator windings.
I have posted the method to do this a couple of times here, you need a double beam scope to do it, and a means of powering the encoder on the bench.
Al.

Thanks for the speedy response Al. Yes it is a brushless. I have a double beam scope and I'm sure I can come up with a way to power the encoder I have a BK Precision 1635 variable power supply. I've been searching on here trying to find the procedure, I'll keep looking, if you could otherwise post a link to one of your previous posts with it.
Thanks,
Joe

You need to identify the H1,H2,H3 hall effect equivalent outputs of the encoder, and the three stator windings.
The three stator windings are each connected to a 1k -10k resistor whose other ends are connected together to provide a virtual star point for the scope common.
The probe of one channel is connected to the A phase.
The other scope input is connected to the H1 or HA hall output.
You need to back feed the motor around 200rpm, I use a small DC motor for this.
Then you need to align the two traces so that the H1 straddles the A phase positive going peak.
The PDF shows the alignment for all three tracks.
As long as you have identified the right pair, then it should be OK.
Al.

Al, I have for pots on my encoder board an A-0 B-0 C-0 and Z would u use them to tune each of the hall outputs then I'm assuming not sure what the Z is for.
Thanks,
Joe

I have never known these to be tunable, or even need to on the drives I have used?
I don't know what they would be?
Al.

Here is a picture of the 4 pots, I'm making a coupler now to turn the motor with a dc motor I have to do the alignment like you said, got my 10k resistors and scope ready to go, just have to figure out which wires I need to be looking at to see H1, H2, H3 and I will be set.
Attachment 204852
Attachment 204854

For some reason i took it to be a Siemens 850 control not a Fanuc, this changes the whole picture, they not only are AC sinusoidal powered, they have their own proprietary commutation.
CN1-2-4-8 are the four commutation signals.
These encoders are automatically aligned with the motor by marks on the encoder bracket, there is an alignment check that can be done, it requires a automotive battery to aling the armature and a test jig to monitor the CN signals for all 1's. I made up a little LED powered board to make it simple.
I can see if I can dig it out.
Al.

Here is the method from the servo manual.
I made a simple board using 4 2N7000 FETs and LED's with 330ohm resistors for checking the CN outputs.
Al.

The only problem I see with the procedure you have in the book they show the encoder being on a 3 bolt patter that has slotted holes for the adjustment, Mine has 4 with no slots, there is a tiny bit of play with them just cracked loose but not much probably only .010" or so. I wish I had a manual for this servo motor like the one you posted the pages out of that might have the info I need. Is there a place to obtain such a thing?
Joe

even if you cannot cannot rotate the encoder the bit pattern should be the same when the armature is oriented properly with the battery, you can use a meter to sheck the voltage on the CN-1 to CN-8
Al.

I finally had time to take the motor and hook it up on the bench to test the output of C1-8 I applied the voltage to the u and vw leads and the motor locked into a postition and I checked the c1-8 outputs and indeed they are not correct. I tried 8 times to take the encoder off and move the coupler on the end of the motor shaft but I cannot get the right output of C1-8. C2 is the one I cannot seem to get to get correct with the others. If anyone has a service book or knows where to get one for this type of servo I'd sure appreciate it or has the procedure to do this. Its a Fanuc A06B-0512-B501#7073 with A860-0315-T101 2000p encoder. Al, thanks for all your help in getting me this far, I wish I had the type of encoder that your book has it sure would simplify this I think.
Joe

Did you write out the four bit pattern for all the positions in post#8?
If you cannot obtain this pattern for every 1/4 turn then there looks to be something wrong with C2 detection or output?
The four LED tester makes it easy to check.
Al.

I made the 4 led tester just like you said, yes it makes it easy to see whats going on. Ok so I tried it at all 4 quarters and get the same output, I also left the encoder on, and what happens when I take the voltage away from the servo so I can turn it free again and I start to rotate it just the slightest movement turns the C2 on, I'm just going to note exactly where the 4 pots are and try to adjust them one at a time once, its like the C2 output is just late to turn its output on to me. I guess at this point worst case I will just have to send it off to repair and have them re-tune it.

Turning it free you should be able to clock the 0-F 16 HEX outputs in the sequence shown in the chart.
Al.

Ok so I rotated it and noted the pattern and it is putting out the correct pattern, but when I apply the voltage to the servo in its held position C2 always requires me to let it go and turn it slightly. So I bolted a degree wheel to the other end and bolted a wire to the motor case and put it on zero with voltage applied, then moved the shaft until C2 lit up, and it would seem to be about 1.5 degree's late.

You mean when you apply the battery the four are not one's or the CN8 1 or 0?
Normally there are three positions to secure the outer edge encoder tabs, which gives three options, not sure if yours has this position option?
Al.

With the battery applied C1, C4, and C8 led's light so they are 1's but C2 doesn't light until I take the battery voltage away and turn the shaft like I said about 1.5 degree's. I talked to another guy in town tonight that has shop and he said it sounds like my encoder is just plain junk, he doesn't think the adjustment pots would be able to fix it. The machine does run the way it is though just that you can hear this servo always making a hissing sound, especially when its not moving. I don't have the 3 screw option like I said mine is bolted down with 4 bolts and the mounts are not slotted, the picture in post #7 is of my encoder.

Well I think its fixed. I marked and then messed with the adjustment pots to see what each did, nothing from the A-0F, B-0F or the Z-0, but the one with a CS on it seems to be what I needed it almost seems like its a sensitivity level adjustment, I'm guessing that it works all 4 of the C outputs but I just turned it a bit at a time to see what changed and eventually I got C2 to light when the motor is locked in. I found that if I went to far C1 will stay on through the entire revolution of the motor, so I backed it off until that one worked again. I put it back into the machine and no hissing, and the following and gain error matches the Z and Y axis perfectly now. Thanks for all your help Al!! I still wish I knew what the other adjustments were for now, but I guess thats for another day.