G320x - Fault over jogging only

[MecAut]

Hi guys.

I tuned my servos using this great site (cnccookbook.com).
The motors now are quieter BUT something is happening and I don't know the root cause,

It can change directions with nice acceleration, so I can keep reversing directions the whole day without any fault (based on 40mm travel each direction)
BUT
When I try to jog the axis for long distance (>100mm) the motor stops with fault condition. Is that noise disturbing everything? Or can be a CPR + computer hardware + wrong software parameters. Even using rapid speed about 1000 mm/min I still had this error.

Below, my machine parameters and Mach3:


Update1: I will check my ESS setup, I may have 256kHz - and based on the following link that will create some problems:
https://groups.yahoo.com/neo/groups/.../topics/140122

Similar Threads:

• VMC15 fadal, 88HS error#9, spindle driver fault, servo amplifier fault
• Fault/Reset circuit for G320X drives
• Problem- G320X Fault Source
• Problem- 139 x motor z fault and occasional ground fault
• Inverter fault / Drive line fault ?

[MecAut]

Hi, after posting here I spent a lot of time trying to understand what's going on here.

1) I noticed the problem persists in all axis.
2) To isolate the problem I decided to turn off the driver and the encoder power supply for axis Y and Z.
3) I decided to tune using the Gecko procedure (here) in PID trimming, the X motor is quiet and it is not bouncing.
4) This is a close-up video, that "bump" in the motor operation happens independently of axis and I cant rid of it!



Questions:
a) I'm not sure if that problem is related to noise or ground loop....any ideas or suggestions please?
b) Should I turn my attention to any dodgy signal from my Computer+ESS? Is that really possible to have problems over the Ethernet interface?
c) The Gecko error limit switch is 1/2 of my encoder line count (500*4=2000, so I'm using 1024) is that correct?

[phomann]

Hi,

Does the bump occur when running G-code?

At first glance it does look like a power supply spike. Are you able to monitor the signals with an oscilloscope?

I have not seen a problem before on an ESS but it would be wise to try a different computer and/or the parallel port to eliminate them.


Cheers,

Peter

[MecAut]

Hi Peter,
Yes, but I wrote a simple code using the "teaching" mode on Mach3 (similar code that can be found in Mach3 installation manual, page 5-25, to see how reproducible the positioning is)

Ok, I appreciate the two suggestions, I will try to monitor my power supply (it is a KL-7220) and eliminate the ESS.

I'm not a pro using scopes, any suggestion about sample frequency?

Regards,
Rafael.

[phomann]

The time base you chose on the scope may depend on the scope you have.

If you have an analog scope, set it up to a period so that you can see the step pulse stream, 20-30 pulses across the screen for the jog speed you are using.
Once the axis ramps up to the jog speed you should see a steady waveform. When the bump occurs see if you can see in in the stram of step pulse.

Cheers,

Peter

[MecAut]

Hi Peter,
No sure if that could help, but I noticed when the bump happened, some LED"s blinked (~16 seconds is the first event) in the breakout board, please check the following video:
For this video, I ran just G0 X1000 (2000 mm/min)


I could not use my scope, because the input connector says MAX 35V....and my power supply has 72V (it is a Hantek 6022BE PC Based USB Digital Storage Oscilloscope 20Mhz Bandwidth)

[MecAut]

Hi,

Does the bump occur when running G-code?

At first glance it does look like a power supply spike. Are you able to monitor the signals with an oscilloscope?

I have not seen a problem before on an ESS but it would be wise to try a different computer and/or the parallel port to eliminate them.


Cheers,

Peter

Hi Peter,

Finally I could test the whole thing using a different computer + ESS (did not solve the problem), so I removed the computer from my black list.
Then I connected the BOB using a parallel cable per your suggestion.

The result was:

1) Switched on the power supply (5VDC) for my logic (BOB)
2) Pressed RESET on Mach3 - the software was ready and the board status was according to the photo 1.jpg - The e-switch worked fine.
3) Switched on the motor/driver power supply (72VDC)
4) Pressed the driver reset switch, the Gecko driver was READY and the motor is quiet and without crazy dittering BUT
5) All LED's were lit in the BOB (please refer to the attached picture)
6) Mach3 was in Emergency Mode all the time - I could not reset it - even pressing the reset button, the Mach3 went back to error.
7) Turned off the G320's power supply
8) I could reset the Mach3 and the whole system went back to the step number 2 above.

Any ideas / suggestions?

[phomann]

The jumpers on your board have it set up for active low. That means that the resting state of the signal is 5V.

Photo 3 is indicating the the breakout board is not enabled as all the LEDs are on (5V) When you press the activate the drives with the reset switch, you are disabling the breakout board. You can check this by looking at the status led on the breakout board. If it is doing a slow double flash, it means that the breakout board is disabled.

Presumably this Drive reset switch is being fed into Mach3. My guess is that when Mach3 see this Reset signal, it is acting on it, putting Mach3 into reset which in turn disables the breakout board turning on all the LEDS

In Mach3->config->general there is a tickbox for "Chargepump in Estop". This needs to be checked. You want the charge pump to be present whenever Mach3 is running, even in EStop.

Also, put the Charge pump jumper back in.It will remove one more possible problem why sorting out the problem.

Cheers,

Peter.

[MecAut]

Thanks Peter.
I will check everything.

Why the reset switch thing happens only when the bob is connected using a parallel cable? Because I have not changed anything from the ESS to parallel. Because this setup was stable (but with the weird bump).

[MecAut]

I think I've found something
Please watch the video:


A) The pin#10 led was flickering at the same pattern/frequency as the servo motor dither! Bugger!

B) the pin#10 led intensity increases when I try to spin the servo motor pulley with my hands (0:26). As soon the motor start dithering the led started against to flicker.

C) the loud "click" on my video is the moment I turned off the 72vdc power supply - check how the led turned off after the PS discharging (1:01).

D) I could reset Mach3 without the constant e-switch request.

Question:
1) Can I say this is a noise issue coming from the power supply and the e-switch wiring is putting this back to the system?

2) Peter do you know why at 16 seconds & 24 seconds all breakout board LED's blinked? I assume that is connected to that weird bump I have already referred to.

Thanks
Rafael.

[phomann]

So, what is pin 10 connected to, and what function is it performing in Mach3?

The reason that all the LEDs flash at 16s is that the board is being disabled momentarily causing the LEDs to turn on.

I notice you didn't do what I suggested in my last email.

Peter

[MecAut]

Hi Peter,

The pin 10 is connected to the e-switch following the board instructions (image-01.jpg and image-02) - Schematic here

In Mach3, the pin 10 is setup as e-stop (image-03.jpg)

I am really sorry, but I thought I addressedd all your suggestions, 1-Tick the Mach3 option about charge pump (image-04) and put back the jumper in the bob to bypass the charge pump (image-05) - seriuslsy what did I miss? sorry again.

Rafael.

[phomann]

This is what I think is going on.

The EStop signal is being fed into pin 10 of the breakout board and also into the Breakout boards 'Lo' enable input. The noise on this line is triggering the 'Lo' enable input, momentarily disabling the breakout board, causing all the output LEDs to momentarily turn on. This input is more sensitive than the buffer circuitry on the pin 10 input so it does not trigger the Mach3 Estop.

You can test the above by inserting the enable override jumper on the 'Lo' enable input. That should stop the thumping.

As to the noise on the signal you need to fix that. If it is a long run, or is the same loom as the servo power cables, the noise may be pickup up that way. In any case you should use screened cable to reduce the pickup of noise.

Also the jumpers on the board in the photo are showing that the pullup resistor for the imputs is pulling the input to 0V. So with that setup I cannot see how the estop for Mach3 on pin 10 will work. With the estop switch in the off position (contacts closed), pin 10 will be at 0V. When the switch is pressed and the contacts are opened, the pullup resistor will pull the input to 0V, so the signal will not activate Mach3.

My schematic shows that the jumper should be pulling the input up to 5V when the switch is pressed.

Peter

[MecAut]

Hi Peter.

First thing, I attacked the problem inserting the enable override jumper to "Lo". That killed the problem - So I could concentrate to track the noise at e-switch level.

As I had the e-switch and the reset driver switch at the same location in my machine I used the same cable (4 wires with shield and drain, picture attached)...I've never thought that two circuits were that sensitive! - Lessons learnt!

About the noise, yes I need to revise the whole thing (even using screened cables for motors and encoders).

Now the board is "stable" and without any error but I noticed the LEDs #2 to #9 have the same behavior that the LED #10 had (some flickering when some torque is induced to the servo motor pulley) and when the led 2~9 blinks I have again a Mach3 error...so this shows me there is some noise coming from somewhere!....

Cheers.