Problem when tying drain wires together

[mmcleod]

Hello everyone,

I’m a long time lurker and have learned a great deal from this forum. I wish it has been available 22 years ago when I built my router. I’m hoping someone can explain why the following happened:

I had replaced my stepper cables with new shielded cables that had a drain wire. The old cables were not shielded and had no drain. I tied the drains on the new cables together and grounded them to the controller chassis.

After I powered up the motors I was hearing lots of clicks and squealing from the motors. Worse, my X axis would occasionally stutter and lose steps. I suspect noise was getting into the system but could not find any logical source.

When I disconnected the drain lug from the chassis I still had the problem, however; when I untied the drains from the common lug the problem went away.

While I can run this without the drains grounded I do wonder what caused the problem. The cables are all well separated. They are 22 gauge and attached to old chopper drivers that are set for 2 Amps/phase. The cables are about 12’ long.

Thanks,
Mark

[Al_The_Man]

With the three drain wires disconnected from the common lug, check for continuity from drain to ground also check continuity between the drain and the other conductors in the cable.
If OK, re-mount the ground and common lug and connect one drain at a time and then try each of the combination of pairs.
Al.

[mmcleod]

Al, Thank you for your ideas. Here are my results:

- There was no continuity between the drain wires and any of the conductors.
- With the drains disconnected from the ground lug – there was no continuity between the ground and the drain wires.
- There is no continuity between the drains and the motors or the router frame.
- I can ground any individual drains separately and everything seems okay.
- I tried all the combinations. Whenever I connect 2 or more drains, whether grounded or not then I have problems with the corresponding motors that are connected. Clicking and squealing as well as stuttering.
- I plugged a circuit tester into the 120V side of the controller and I have a good electrical ground. The only other item on the A/C circuit is a refrigerator. There is no other electrical equipment nearby except the computer. I moved the controller and cables about 3 or 4 feet away from the monitor and computer but there was no improvement.

I don’t have a scope to see what the interference looks like. Do you see any potential for problems down the road that would necessitate tracking this down?

Thanks,
Mark

[Al_The_Man]

I guess technically you could run the stepper motors unshielded, the principle object is to prevent radiated EMI, steppers are typically low impedance devices so in theory they should not suffer from received EMI as would be servo encoders for e.g. that you want to prevent receiving motor EMI.
I am surprised that they appear to be affected this way.
The only other thing I could suggest if you really want to try using the drain in order to shield the stepper conductors is to try grounding the stepper supply common, as long as this is electrically feasible.
Al.

[mmcleod]

Well, I'm out of ideas. I'll just clip the drain wires and run it that way. It is a mystery and it bugs me that I can't figure it out.

[mactec54]

Hi mmcleod

It sounds like you have both ends of the sheild connected, you can only have the input end connected to ground, clip the motor end flush with the end of the cable & ground the other end

[Al_The_Man]

I believe he already performed that test that I suggested earlier.
Although that brings up another possibility, are all the motor frames bonded to earth ground, if any are above ground it could be one root cause of the problem.
Al.

[RomanLini]

Don't connect the shields directly to ground, connect them through a 100 ohm resistor.

That should fix your problem with the high frequency ground loops and still provide a fair level of RFI suppression.

If you haven't already done so, it's also a good idea to add a large capacitor (1000uF or 2200uF) at each stepper motor driver directly to it's PWR+ and Gnd pins, mounted as close to the driver as possible. That will help to keep the chopper recirculating current local to that driver and its motor and not looping back through driver-motor-main psu.

[Al_The_Man]

The recommended approach to ground/bonding is to take a separate ground conductor from ALL motor frames back to the common earth ground point.
Another possibility is to earth ground the power supply common to the star point also.
Al.

[mmcleod]

Well the router frame isn't grounded, it sits on a wood table and the router motor is attached to the frame using non metallic (wood) brackets. I meant to ask about that earlier.

I'll run a test to a grounding rod and let you know what happens. Do I really need to ground each stepper motor separately or would I get most of the benefit by just grounding the frame?

Thanks for your help,
Mark

[Al_The_Man]

Strictly speaking the common bonding technique is to take a ground conductor back to the central earth ground point from each motor and any metallic parts of the machine, gantry etc.
There is no real need for a ground rod, the service supply ground should be sufficient.
The concept behind bonding is to make all metallic parts of the machine one potential, this attempts to rule out the possibility of a ground loop, this can only occur if the grounded metallic/shield/supply common parts of the machine have a different potential between them.
Don't forget the router motor frame, either.
http://www.cnczone.com/forums/showthread.php?t=71923
Al.