Knowlege needed

I have a hurco bmc 20 1992 (and im very green).
Here is my problem.
If i put an indicator in the head and zero it on an edge in Y. Then jog the Y axis with the wheel in tenths say in the positive direction, the indicator will read what the screen says. Now when i change direction (turn the wheel the other way) the indicator jumps .002" and the readout does not. as you continue turning both the indicator and the readout show the same amount of movement. Now you stop, as you just turn the wheel in the oposite direction it will do it again. then continue on reading the same distance change as the screen. (you can feel the table jerk if you rest your hand on it when you change the wheel direction) Of course the total distance moved is not close because of the jump at the direction change. The X axis does not do it, and the Z does not either. The problem just started today. Rebooted and no luck the problem is here for good. Has anyone seen this before, or have an idea of what direction i should be looking at for this issue?
thanks
Darren

There’s a couple of things it could be. First a crappie indicator can cause that easily. I see it all the time. So you have to start ruling things out, and I’d start there. Use another indicator… Then if using another indicator you get the same readings you’ll have two more possibilities. One being backlash in your table which would mean you’ll have to tighten up your gibs at the adjustment screws. But not to tight, you want to leave about .0005 measured on an indicator worth of movement in the table both axis’s. Use a mag base and isolate the two moving sections meaning mag base stuck on one and set your indicator to read the other. Then bump the table or push the table and watch your indicator. Chances are if this problem didn’t exist before this isn't the problem now. But this will rule out any gib slop issues just to cover all the bases. Then there's the most likely candidate,,, power… As you may or may not know your drive cards are extremely sensitive. The voltages through your drive cards cannot very more then about +or- .4 of a volt. I wouldn’t recommend you taking this task on unless you are savvy with these electrical devises!! You can screw-up your machine in a hurry beyond all believe! But there is a way to work around this problem if you own a good maintenance manual!! If you do, read very carefully any sections pertaining to the adjustments of voltages and the balances on your drive cards. I cant stress this enough,,, if you don’t know what your doing stay away from your cabinet. Get help. good luck

BIG DADDY!
Thanks for the great advice! Its not the indicator, and unfortunaltly its not the backlash. Its measures .0005 as you say. So its time to call in a repair shop. Can you tell me specificly what to tell him to expect to see when he travels to my shop in CNC lingo? Ive called the company before and because i know so little its painfull to get them to understand what im asking.

Thanks again!

Well you can say that your machine is acting erratically when you jog your table back and forth. Then they’ll ask what do you mean? Then you tell them about the indicated error that you found with your indicator. Also let them know that you have ruled out a bad indicator plus you have checked for slop in all your gibs and everything was fine! They’ll probably right away after that suggest a service call to your place.

Wait, let me give you a really good option; contact these guys at www.solverstechline.com
John & Paul are some of the best in the industry and for a nominal fee they can really help save you a lot of money. Chance are they'll walk you through this step be step and track down the problem or problems. I got a pretty good idea what it might be but I'm not there so I won't say. Let them help you. good luck.

Thats great thanks so much!

Darren.

Because you can feel the jerk, I suspect there is some backlash compensation turned on that needn't be.

It works like this.
Say the backlash compensation is set to 0.002", then when you change direction, the screw will reverse (by the 0.002") instantly, just as you are measuring.

This is setup in the CNC machine setup. I don't know this unit, but it looks like the experts are on hand.

Neil.:cheers:

I would tend to think of too much backlash compensation in a setting as well, except that such a problem would not 'just happen today' unless you were 'fiddling with parameters just today' :D

I would carefully check any couplings between motor and encoder or resolver for slackened screws. It could be that a little bit of lost motion in such a coupling could cause the controller to jerk the motor a little bit when it attempts a reversal but does not see the expected reponse in the position feedback loop.

I already thought of all of this. 1) if there was a servo coupling issue I doubt very seriously this is the problem he’s facing. Why? Because all you would have in this situation is “LOST MOTION”!! 2) Backlash comp: What does that mean??? Same as number one!! LOST MOTION!!
What he has is a jump!! This is power related!! Either DC invertor (outside chance) or most likely a voltage balance problem. In this case hopefully a simple adjustment to the balance potentiometer will solve his problem.. But I did also give this a lot of consideration last night and anything is possible I geuss. What the heck,, it's a Sunday! No bodies open! It's worth a shot!! Try and tighten up the X & Yservo couplings! And take a look at the backlash as well.
May as well check to see if the servo itself needs to be tightened up while you’re at it! But I'll also say this, if it was anything mechanical this would have been a gradual occurrence. Over the course of time it would be gradually getting worse and worse in most cases.

The servo coupling is different than the encoder/resolver coupling and has a different effect. Some of these tiny couplings are rather fragile and could have broken components after many years of usage.

Here’s a little story; Back many years ago I use too work in a shop where to owner was somewhat of a tyrant. He demanded production. So we had no choice but to push his Fadal’s hard. The consequences were every six months we were tightening up servo couplings. But I’ve really haven’t seen this happen outside of his shop. Most people don’t push their mills this hard. But with that being said I guess over time with an older machine that’s seen a lot of hours something like this would probably happen. And it would be just a good idea just to check all couplings anyway. Great maintenance tip Mr. Moderator!

Hu is of course referring to the coupling between the feedback device and the motor .(usually on the rear of the motor, sometimes a coupler, sometimes just a set screw on the shaft)

If the coupler is cracked or the screw comes loose when the servo is commanded in the opposite direction the motor begins to move but the encoder does not. Since the feedback loops sees nothing the integrator continues to build until the motor (and screw) move far enough to move the encoder. Seeing motion the loop now stops the motor. This has nothing to do with how hard the machine has been used but will cause exactly the action described. Usually it will get even worse over time.

Check that the feedback device is rigidly attached.
Bob