Problem with Z-axis direction

[acondit]

I recently finished assembling my new router (from my old router's parts).

I also just upgraded to Ubuntu 10.04 and EMC2 2.4.5. I was running 2.3.? on 8.04 and experienced the same problem, so it may be that it is not EMC2 related but I am not sure yet.

The problem is that the z-axis does not seem to be responding consistently to the DIR signal. I have an LED on the DIR lines (BOB inputs) and when I try to jog the z-axis, if I press the minus button in Axis sometimes it will go down and sometimes it will go up. The direction that it moves corresponds to whether the LED is lit or not.

It was working fine before I took the old machine apart but now it is giving me fits. I don't know whether it is a computer problem or a software problem. The X and Y axis are not exhibiting the same problem.

Any suggestions??? Maybe I should move the z-axis connector to the x-axis and see if the problem follows the connector?

Alan

[lumberjack_jeff]

definitely switching axes can help with diagnostics.

Is it possible that the new z is too heavy and you're losing steps?

[CarveOne]

I had something like this happen when my Dumpster CNC delrin coupler started slipping.

As a test of whether the Z axis is too heavy, or is binding, loosen your coupler and see if the motor consistently reverses direction.

If you have a connectorized cable, make sure all of the pins are fully engaged to the same depth. DB9 connectors from some vendors can get pushed back into the connector body and will become intermittent before not working at all.

CarveOne

[acondit]

Originally Posted by lumberjack_jeff definitely switching axes can help with diagnostics. Is it possible that the new z is too heavy and you're losing steps?

I am not even trying to run code yet. It is just in manual jog that I am experiencing the problem, so I don't think losing steps is an issue yet??

I will try switching cables and see if the problem follows the port or the axis.

Alan

[acondit]

OK, I incorrectly described the problem.

The dir LED goes on when I press "+" button and goes off when I press "-" button. So, the control box seems to be receiving the correct signal from EMC2. However, regardless which button I press, sometimes it goes up and sometimes it goes down.

If I press the "+" button and it is going up (correct action), then I have to press the "-" button to get it to go down. The problem is that sometimes it continues going up and I have to press the "-" button multiple times to get it to change direction.

If I press the "+" button and it is going down (incorrect action), then I have to push the "+" button multiple times to get it to go up. Same basic situation in reverse for the "-" button.

I have flats ground on both the screw shaft and the motor shaft for the setscrews of the Lovejoy connector. I am using 4 pin power connectors (look like audio mike connectors but slightly higher amperage ratings).

I would suspect a loose direction wire on the z-axis drive but when I switched the Y and Z connectors the Y axis behaved fine with the z drive and the Z axis behaved the same way with the Y drive. So that would seem to indicate to me some kind of mechanical problem. I just don't understand what.

The z-axis is no heavier than when it was working correctly, in fact it is lighter right now because I don't have the router mounted. I am using the same leadscrew and nut but the rails are mounted slightly different. If it is mechanical, the possibilities seem to me to be: 1. a binding on the rails or 2. a binding of the nut against the screw. I tend to rule out no. 1 because the carriage seemed to slide freely on the rails before installing the screw.
That leaves the very real possibility that the screw is binding in the nut. I guess I could take the motor off and borrow a sensitive torque wrench to see how much force it takes to turn the screw. Thoughts?

I would have thought that if things were too tight it just wouldn't move not move in the wrong direction.

Alan

[lumberjack_jeff]

It seems to me that you've verified that everything between emc and the driver is okay. I would look carefully at the wiring between the driver and the z stepper. Be sure the wire hasn't failed/shorted and that it's not picking up any noise. If there's enough wire slack, I'd plug your 4-pin z connector into another stepper.

Your connectors should make diagnostics relatively easy. The goal is to try everything to get a different axis to display the z behavior.

I would also swap two stepper motors. Alternatively, I'd remove/uncouple the lovejoy coupler and see if the stepper's behavior changes.

I'd recommend ruling out any electrical problems prior to chasing mechanical ones.

[acondit]

Originally Posted by lumberjack_jeff It seems to me that you've verified that everything between emc and the driver is okay. I would look carefully at the wiring between the driver and the z stepper. Be sure the wire hasn't failed/shorted and that it's not picking up any noise.

OK, I'll check the wiring.

If there's enough wire slack, I'd plug your 4-pin z connector into another stepper.

I thought that was what I did? I plugged the z into y and y into z. Is that what you mean?

Your connectors should make diagnostics relatively easy. The goal is to try everything to get a different axis to display the z behavior

Yeah, but if it is mechanical, its not going to happen.

I would also swap two stepper motors. Alternatively, I'd remove/uncouple the lovejoy coupler and see if the stepper's behavior changes.

I will think about this one. I see the logic, I just don't like the work that is involved. It may come down to this to prove that the problem isn't the z stepper though.

I'd recommend ruling out any electrical problems prior to chasing mechanical ones.

I thought that I did that when the z-axis driver drove the Y-axis fine. Forgive me if I seem snippy, it is not my intention. I'm just frustrated that I tore down a working machine to build a better one (that now doesn't work right).

Alan

[lumberjack_jeff]

Originally Posted by acondit OK, I'll check the wiring. I thought that was what I did? I plugged the z into y and y into z. Is that what you mean?

Maybe I misunderstood. I inferred that you have plugs on both ends of your cables and that you unplugged cables from the drivers. I'm suggesting you rearrange the ends at the steppers.

Yeah, but if it is mechanical, its not going to happen.

In my experience it's easier to diagnose electronics. The mechanical part is too subjective ("How much friction is too much? Is this binding?). YMMV.

[acondit]

No, I just unplugged the z and y cables from the controller box and switched them. The cables are fed through cable chains and different lengths so it would be difficult to actually switch the y and z cables.

Alan

[acondit]

Jeff and CarveOne,

I didn't want it to be the z-motor but it was. I replaced the 200oz/in stepper with a 471oz/in one (the only other Nema34 that I had that was even close to running on that power supply). It runs like a top. Instead of jogging at 45ipm it runs 120ipm without breaking a sweat. It didn't even try any faster for the moment.

Alan

[lumberjack_jeff]

I'm glad you have made progress. I've never personally had a stepper motor fail on me, they're pretty durable. Have you tried swapping it onto another axis to verify that it's bad and not just underpowered for the mechanical conditions?

It seems to me that you still have two possible explanations; bad motor or mechanical friction.

[acondit]

Originally Posted by lumberjack_jeff I'm glad you have made progress. I've never personally had a stepper motor fail on me, they're pretty durable. Have you tried swapping it onto another axis to verify that it's bad and not just underpowered for the mechanical conditions? It seems to me that you still have two possible explanations; bad motor or mechanical friction.

Not yet, but I suppose I ought to do it just to know for sure.

Alan