[lobojk]

So imagine you are cutting an array of squares, 4 inches on each side, over and over and over... just as a test of the reliability and repeatability of the machine. You want to see every square turn out the same no matter where on the table it is cut.

And you want every square to be the same no matter how many times you run the cutout pattern, day after day.

With a good quality machine I'd expect this simple task to produce many identical squares.

Now imaging watching the machine cutting square after square successfully, day after day, and then, out of the blue, as it is cutting down the Y axis, only moving along the Y axis according to the g-code, you observer the machine's X axis leap into motion and move the X axis in the negative direction by 2 inches...

And then, it goes on cutting squares just like it was before... with no problem.

What can cause this?

The g-code has been run over and over. The square, in this position in the array of squares, has been cut correctly many times.

How do you diagnose or troubleshoot the random "Let's just move the X axis here for 2 inches then pretend it didn't happen"

Of course when the X movement randomly appears, it ruins the previous square because when the machine returns to normal, it continues cutting the square, but now it is in the wrong place. (these are stepper motors so no location feedback)

Where do you begin? What is the diagnostic procedure?

(I am in contact with the motion controller people who built my system, but I'm curious about what you would do if this was a DIY machine)

Thanks for your thoughts!

[Dr.Fiero]

I would start by looking for mechanical binding. If a bit of swarf got caught in the nut, and randomly dropped into the ball path it would wreak havoc.

A sticky spot on your ways.

Next would be some EMI sneaking into a cable from somewhere, loose cable, that sort of thing.

That's a start anyhow.

[H500]

Rule out the easy stuff first, binding slides, sticking ball screws, loose couplings. Check the electrical connectors for looseness. Remove and reinstall each wire, WITH THE POWER OFF.

[lobojk]

Let me re-iterate, let's dis-ambiguate this

It is perhaps more common to see a problem where an axis DOES NOT move when it should move.

The problem I am having is NOT that problem.

The problem I am observing is that the X axis moves WHEN IT SHOULD NOT MOVE.

The problem I am NOT having is the one where things stick and I should check my ways.

But tell me how a piece of dirt, even a pound of dust can make an axis move when it should not move.

Again, I'm not having the "it doesn't move when it should move" problem.

This problem is "X axis moves when it should not move"

I appreciate the help. Just looking for something that is more on topic.

[lobojk]

I've gone through every connection and they are all good and firm. Remember the machine is working correctly 99.999% of the time

48 hours straight, pretty much, of cutting the same pattern over and over, and in that period there were 2 failures where the x axis moved dramatically, each abnormal move lasted less than 2 seconds.

I think that works out to 99.999% correct but check my math, please!

[lobojk]

oops... it's been checked and I was wrong... it represents 99.9977% of the time

[lucas]

My guess:

Your X axis drive is picking up noise from a spurious disturbance source, is there anything special at the moment of failure?
Fluoresent lights switched on?
Wife hoovering the carpets?
Don't forget the neighbours......
Is your system cabled, schielded and grounded as it should be?

There could also be a program running on the PC in the background wich causes spurious pulses from time to time.

[lobojk]

Lucas,
are you suggesting a power conditioning box be introduced between the circuit box and the cnc, something with a battery that will maintain volts/amps at a constant level? Would that help?

Is there a power conditioning box for 220 that includes a log of input and output voltage/amps that can be checked for how clean the power is?

[lucas]

Real power conditioning isn't needed, maybe an input filter.

But you didn't respond to the question:
"Is your system cabled, shielded and grounded as it should be?"
This is the first thing to check and the most important for problems like you have.

Is this a USB or parallel port system?
How do you connect from PC to the drives? Are you using a BOB?
Is everything on the same outlet?

A picture of the electronics setup would help.

[lobojk]

The computer is on one circuit. The VFD and the drives and motors on on a different (220) circuit. There is a BOB. The machine is still under warranty and the company is replacing the x driver.

The full answer to "is everything cabled, shielded, and grounded" requires itemizing each wire and testing it. Since I bought the system and it worked for north of 50 hours from an established, experienced company I tend to assign to them the capability and responsibility for the wiring, particularly the "shielded" part and the grounded part.

Can you provide a name and a part number for a 220v input filter?