drive issues

Hello all!

Preface; I have been a tool and die maker for 30 years, some cnc experience. I bought a bench top clone ZX7045 mill completed retrofit. It never ran code at g-code feed rate. Change the program feed and it would still run at the same rate. Running a sound logic BOB and Gecko G320 drives with reliance electric E712 servos. That system failed this spring and I replaced with a C76 UC300 from CNC4PC and switched to UCCNC control software.

Fast fwd to where the machine stands now.

I can get the UCCNC to run the axis calibrate program, but only one axis at a time. X calibrated with power to all three drives just fine. Y wouldn't calibrate at all. It would start to move and then jump erratically causing all drives to fault.

I then disconnected power to x and z and the Y axis would then calibrate and I was able to adjust the dithering to a minimal amount.

I then disconnected Y power and calibrated the z axis. This again has minimal dithering.

After the previous statements I connected power to all three drives and the thing dithers heavily and jumps in all three axis! ( this is with the C34G320 hard enabled ) with the c34 soft enabled the fault light is on and there is no axis movement when typing a MDI line of code. UCCNC shows DRO movement, but not the machine.


any input IS GREATLY APPRECIATED!

If anyone is in Michigan with experience with this system, I would be open to have you look at it with me.

Thanks for reading! Pic attached.

Without knowing your setup, my guess is you need to tune the servo loop. The tuning is not in the UCCNC. The loop is between your G320 and the servo. Once this is stable, then setup / calibrate in UCCNC.

Quote:

---

Originally Posted by Jmhx2

Without knowing your setup, my guess is you need to tune the servo loop. The tuning is not in the UCCNC. The loop is between your G320 and the servo. Once this is stable, then setup / calibrate in UCCNC.

---

I was able to tune most all the jitters out of the drives one at a time. There is still a small amount of movement ( not entirely sure what to call it when the motor is confused on which count it wants to stop on), very minimal amount of motor pulley oscillation. No hum or whine noise when a single axis has power. When I connect power to all three, things go crazy.

I will attach a pic or two. i'm not sure how to identify/name pics for this post....but I will try. The old system will be much more clean looking, as I was just trying to get the thing to move properly before I cleaned up the wiring in the wire tracks.

The old system uses contactors to power the axis motors and a toroid transformer. Both of those are still in the system, however, I am not sure now what the toroid is powering except for the fan. The toroid and capacitor could probably be removed and use the new BOB to power the fan. Again just wanted to get the thing to move properly with the least amount of rewiring.

When the machine failed, the contactor for the spindle was jumping in and out which caused a breaker to trip. this happened two times then the machine would not even move an axis. it was completely dead, even though it seemed I had correct voltages at the proper places. Talking with Arturo at CNC4PC I just decided to upgrade.

fast fwd to today.
I have a C76 UC300 BOB. this is powered by a 24 volt P.S. in the lower right corner of the panel

Gecko G320 drives which get their power from a contactor relay ( closest to the center of the panel, for pic reference). My thoughts are telling me that I have dirty power coming from the contactor which powers my drives. I think this because of the fact of the humming and jumping in and out when things failed.

My encoders require a external P.S., so that is coming from the 5v 500ma port off the BOB. The drive motors would not function with the external power supplies I had purchased and connected, which tells me somewhere in the logic there is a switch allowing the C34G320 Ethernet adapter card to function properly. ( Connected to the Gecko)

I have, at the encoder, a C72 differential line driver connected to RJ45 running to the electrical panel connected to a C46 differential to single ended converter. this then plugs into the C34G320 board which is connected to the Gecko G320 in place of the encoder wires.

talking with Arturo at CNC4PC, he suggested that I shouldn't have to re-tune my drives/motors.

the only things that should be connected to the new system, that was in the old, is the drive motor power supply ( or contactor ). I'm suspecting corrosion or a cracked housing which s allowing contamination between output lugs. ( L1, L2, L3 etc) this is 120v in 57v dc out) this is the only thing at this point which makes sense. My plan is to get a power supply for each drive motor. The machine in the old configuration was very quiet and stable. no motor oscillation at all and only a very slight bit of whine or hum from one motor.

I was told that the C76 or C34 card handles the ERR/RES function, which Gecko states to jumper together, and it can be left if you don't need to monitor the fault( I think I stated that correctly )

Also, Gecko states that there should be a 470 ohm resister connected on terminals 6 & 7 if a external power supply is used. I don't have that connected yet. I just found out a week ago. I have them, I just need to get it implemented. I would love to think it was as simple as that, but, nothing has been simple in this whole conversion. ( probably my fault. trying to do it while working lots of hours and being tired. )


Servo motors are Reliance electric E712, which I cannot find a data sheet for to check on requirements. With that said.....the axis functioned perfectly before catastrophe hit, just not at g-code rate. they were quiet , smooth, and locked in place when stopped.

In the upper right of the panel is a card for connecting a MPG. Currently not connected.

All products are from CNC4PC.

I do agree I have a issue with hardware. I was hoping to find someone who is using this setup and had some struggles with cures.


The contactor on the right side of the panel is what used to power the old BOB.



THANK YOU FOR YOUR TIME

You aren't going to like this advice... but I suggest removing all wiring from the enclosure and starting over.

Make a list (spreadsheet) of all components, what they do, what the input/output voltages are, how many amps they draw, etc.

Once you have that list you should be able to explain to yourself - in English - how everything functions.

Then draw a wiring diagram before you connect anything. I like TinyCAD - it's free - but there are others. Even pen and paper is fine.

Only then should you start wiring things up.

Example:

(AC power)
1. Mains/line power enters enclosure. L1/N/GND terminate at Main Disconnect Switch.
2. L1 continues to Fuse 1 (or Breaker 1). N terminates at neural bus bar. GND terminates at star ground.
3. L1 goes from Fuse 1 to AC Distribution Block.
4. L1 goes from AC Distribution Block to Contactor 1 and Power Supply 1 (24VDC out).
5. L1 goes from Contactor 1 to Power Supply 2 (48VAC toroid),
6. 48VAC from PS2 goes to Rectifier 1.

(Drive power)
7. 57VDC from Rectifier 1 goes to Drives X/Y/Z (Geckos)

(Control power)
8. 24VDC from PS1 goes to 24VDC Distribution Block.
9. 24VDC from DC Distribution Block goes to:
- C76 board, pin xxx
- xyz components (limit switches, whatever)

(5VDC power)
10. 5VDC from C76, pin xxx goes to 5VDC distribution block
11. 5VDC from 5VDC block goes to encoders, pin xxx

You get the idea. Do the same thing for all main power (including spindle), control, and signal components/wires.

I would try turning down the gain on the servo amps. It sounds like resonance. When all three drives are powered, the resonance of one loop is throwing the others out. Mechanical slop can intensify this. (tightening the gibs can help)

Quote:

---

Originally Posted by Jmhx2

I would try turning down the gain on the servo amps. It sounds like resonance. When all three drives are powered, the resonance of one loop is throwing the others out. Mechanical slop can intensify this. (tightening the gibs can help)

---

Thanks for the reply.

I am using the C76 BOB with the UC300 Ethernet system from CNC4PC. The Gecko website states that there should be a 470K ohm resister soldered across the encoder + & - pins ( 6 & 7 I think ). Soldering this resistor to the C34G320 board ( Ethernet cable adapter board for the encoder ) has calmed all drives to a VERY usable condition.

However, I cannot jog at more than 10% feedrate.

The gain is around 1/4 from bottom ( CCW)

I have the current limit pot at maximum.

I am curious to know what differences there are between BOBs. My old board was a sound logic, and was powered by 120v AC. My new system is 24 v DC. Can this affect the power driving the motor or the drive? Seems like the drive should provide power to the motor, so , this should be a power to the drive issue.



The other point I am curious about ...could this be an acceleration or speed conflict in UCCNC. ( currently all set at default )


Thoughts?

Thank you for your time