Okuma MCV-A intermittent all axes oscillation

[TheRick]

We've got a 1985 Okuma MCV-A with OSP5000M control. For some time we've had an intermittent problem with the axes. It appears as though X, Y and Z are all oscillating! Enough that the milling finish will be unacceptable.

OK, so what do they have in common that could cause this? CNC power supplies are correct and no ripple. Servo 300V power supply is correct, the old one had some ripple on it (but under .120VAC) and we replaced it today, anyway, with no change.

What's left? Maybe the axis board in the CNC computer. In the mean time, has anyone seen and solved this? What did you find? Maybe a bad ground somewhere?

 

[jz79]

1. What sort of encoders are on those axis?
2. You say that all 3 axis oscillate, but do they do this all at the same time or randomly one at a time?

regarding 1., some encoders need periodic maintenance and/or repair, brushed ones need cleaning, inspect for stuck brushes, replacement of worn brushes and collector may need refurbishing, accumulated carbon from the brushes wearing down can cause all sorts of problems, light source may degrade on optical ones, will need replacement and so on

regarding 2., if all 3 always act up at the same time, this may be control related, if any one at a time, can be servo amp or motor/encoder related, but it might be also possible that one axis leads to all 3 oscillating, pay attention if one of those motors always starts to oscillate first and then it progresses to 3, or all 3 at the same time

 

[TheRick]

1. What sort of encoders are on those axis?
2. You say that all 3 axis oscillate, but do they do this all at the same time or randomly one at a time?

regarding 1., some encoders need periodic maintenance and/or repair, brushed ones need cleaning, inspect for stuck brushes, replacement of worn brushes and collector may need refurbishing, accumulated carbon from the brushes wearing down can cause all sorts of problems, light source may degrade on optical ones, will need replacement and so on

regarding 2., if all 3 always act up at the same time, this may be control related, if any one at a time, can be servo amp or motor/encoder related, but it might be also possible that one axis leads to all 3 oscillating, pay attention if one of those motors always starts to oscillate first and then it progresses to 3, or all 3 at the same time

Thanks for your input. Regarding 1, I've been repairing CNC machines since 1978 and have had my own business since 1981 and have to say that I have never seen an encoder with brushes. Resolvers? Yes, but never an encoder. I'm not sure what the purpose would be. The light source shines through the etching or slot onto a photocell and through some other circuits and out the wiring. Nothing rotating requires any electrical connection. That said, feedback systems clearly can cause issues similar to this, but not all 3 axes at the same time and in a scenario like you're suggesting might be going on, I should be losing some positioning. Zero never changes. The axes just vibrate.

Regarding 2, when the issue happens, it seems to affect all axes but in varying degrees. Always worse on Z. Bad enough to affect finish on a mill cutter or simply make a rattling noise you can hear. I'm not the first one to look at this so I can't say if the other company started making adjustments and made things worse or it's always been like this. Right now I have to assume no human involvement was initially involved.

In my opinion, I've either got a faulty servo board in the controller or I've got a grounding problem. For the record, I don't usually need to ask questions but in this case it's an odd problem. I've seen axes oscillate before but intermittently on all axes yet consistently because if it's going to happen they're all going to do it, is illogical. An intermittent consistency? Power supplies have all been checked for voltage and ripple. The "other" guy who worked on it replaced the computer power supply. I thought I saw something on the drive power supply so I put in a new one earlier this week. Monday I'll head back and double check all the ground connections and see if maybe the servo board just needs to be re-seated in it's slot.

 

[ViktorS]

He probably meant tachometers, not encoders.

Either way, if all axis starts and stops at the same time, the reasonable thing is to start at the interface between the controller and servodrives. If it's an analog +-10V signal then it sure sounds like an internal voltage stabilizer or the like is on the way out on that interface card.

 

[jz79]

yes, I used the word 'encoder' as a blanket word to cover all possible feedback sources for the servo drive, I mentioned optical ones since older ones (including Heidenhain glass scales, and is a known problem for those older scales) use small light bulb as light source, those get cloudy and dark over time which in turns makes reading signal on the photo resistor weak, the amplification circuit that then conditions this signal can cope with it for a while, but once the signal is too weak, the amplification is becomes too noisy and problems start to occur

if Z axis is always first, is there a brake on it? and perhaps there is a stick/slip issue that the servo is trying to deal with, might be something mechanical causing the problem, that might be why trying tuning it didn't help at all, and I'd check backlash on the screw thrust bearings and the nut itself

and the supply voltage ripple may not show up when the machine is operating normally, but once it gets into oscillation - that means very high and peaky currents pulsing on the drives - that can cause overall power supply become dirty and just feed into the failing feedback loop

 

[TheRick]

He probably meant tachometers, not encoders.

Either way, if all axis starts and stops at the same time, the reasonable thing is to start at the interface between the controller and servodrives. If it's an analog +-10V signal then it sure sounds like an internal voltage stabilizer or the like is on the way out on that interface card.

His response was that he was using "encoders" as a generic term for any feedback device. I'm guessing he either is young and inexperienced or ESL. If you want to help someone, you need to know how to express yourself and use the correct term for what you mean. Clearly he did not. He also seems to not be reading, or understanding, what I've written. Maybe that's my fault, but where now did he interpret what I've said as though if always starts in Z first? Yes, Z has a brake (in case he reads this), but that's the source of the rattling noise when Z starts vibrating because the brake disks usually have some looseness between so they can move.

Analog drives. 1985 machine. Power supplies all check out. Looks like someone else put in a new control power supply and I thought I saw something odd on the servo power supply so I got a new one and replaced it last week. It really comes down, now, to the interface card. I only posted this question hoping someone had seen this issue before and might have some insight.

 

[jz79]

jesus christ how touchy we are, if you had the experience, you'd understand what I meant with "encoder", for the servo drive logic it doesn't matter if it is optical, magnetic or mechanical - it gets converted to voltage that is used then by opamp comparators, if that signal is intermittently lost or weak, the drive logic will think the speed is too slow, it will increases motor power output to make it go faster, the next moment the signal comes back, the motor speed is too great, the drive overcorrrects trying to slow it down

this intermittent loss of signal causes jerky movement, if the control is trying to stand still or move slowly, it will result in oscillation, this oscillation in turn will make the intermittent signal feedback situation even worse

and the cause can not only be the "encoder", but also mechanical issue with the axis, hence me mentioning the brake or worn bearings or worn nut

if I had 40+ years of experience troubleshooting these, I wouldn't need to ask these sort of questions on internet in the first place, I'd fix it and perhaps make the solution known the world on some internet forum instead of asking about simple things and then complaining that the suggestion wasn't worded correctly

good luck fixing your machine

 

[TheRick]

jesus christ how touchy we are, if you had the experience, you'd understand what I meant with "encoder", for the servo drive logic it doesn't matter if it is optical, magnetic or mechanical - it gets converted to voltage that is used then by opamp comparators, if that signal is intermittently lost or weak, the drive logic will think the speed is too slow, it will increases motor power output to make it go faster, the next moment the signal comes back, the motor speed is too great, the drive overcorrrects trying to slow it down

this intermittent loss of signal causes jerky movement, if the control is trying to stand still or move slowly, it will result in oscillation, this oscillation in turn will make the intermittent signal feedback situation even worse

and the cause can not only be the "encoder", but also mechanical issue with the axis, hence me mentioning the brake or worn bearings or worn nut

if I had 40+ years of experience troubleshooting these, I wouldn't need to ask these sort of questions on internet in the first place, I'd fix it and perhaps make the solution known the world on some internet forum instead of asking about simple things and then complaining that the suggestion wasn't worded correctly

good luck fixing your machine

If you had the knowledge, you'd know that an encoder in a servo system and a tachometer in a servo system are two different devices. When you're trying to give someone ideas about a possible solution to a problem, you need to use the correct word for what you mean. Encoder loss is going to result in position loss. That was not the issue. If you read, and understood, my post, you'd also understand I am following someone else who has already worked on the machine and I have no clue what they've done. The simple things, like what you've suggested, have been checked. I don't come here looking for simple stuff. Other than a bad chip on a board (most likely the servo control board in the computer) I was hoping to find someone who had seen something "out of the box". I have not found that person here.

IMO you are not old enough to have worked on anything but AC drives. That explains your error in the use of the word encoder. Tachometers are used on DC drives, primarily, and generate a DC voltage themselves. It does NOT, as you suggest, get converted to voltage as it already IS a voltage rather than pulses. A dirty tachometer could cause a similar problem to what we see, but it would not be intermittent. It would be there all the time until you clean or replace the tachometer. This machine will run for days with no problem then they come in one day and it starts acting up. Sometimes all day, sometimes not. It can be fine in the morning and mess up in the afternoon or it can mess up in the morning and be fine in the afternoon. A tachometer is a likely issue but that's only suggested by the fact that we don't have positioning problems. Encoders give positioning data and with no loss of position accuracy, an encoder can not be the problem. And remember, it happens at the same time on all axes. It, therefore, can't be a device on a single axis. It's something common to all. That also eliminates all the mechanical issues you suggest. Stay on point. Observe the problems, understand how a machine works and some day you'll make a good serviceman. You've got the basics down, but I don't think you fully understand how they work, regardless of your remarks.

 

[jz79]

oh dear god, did you just call tachogenerator a tachometer?

and chips? potatoes or integrated circuits, be clear, please