Hi Charles,
Interesting. I think that indicates sig GND on the drives is a source of noise. You might look at with your scope relative to Kanalog GND.
Hi Tom,
Back on it again. Had to take a break. Tried the low pass filter. Didn't seem to have much effect. Didn't find an oscilloscope, but found and bought a Perytech DSO-U2200 USB Oscilloscope for use with a PC. Now I have to try it out. Never used one much, so I have to read up and refresh.
I just had some success with different connections on the axis command cables. First tried to minimize the exposed unshielded wire on the command cables. My cables were originally shield drained at the drives. I tried switching the shield drain to the Kanalog board.
Turns out, as long as the axis shield drains are both disconnected at the drives, jogging does not affect the spindle.
It doesn't seem to matter if the shields at the Kanalog end of the command cables are connected to each other or to Kanalog ground. As long as they are not connected to the drives, the spindle is not affected by axis jogging.
I had them connected at pin 2, marked Signal Gnd, on the AMC 30A8 drives.
Connecting even just one shield drain at an axis drive causes spindle jerking when jogging X or Z.
As of now, axis and spindle command cable shield drains are connected to Kanalog ground and not connected to drives.
Spindle still vibrates a little when enabled. It may be that tuning would help with that.
Sorry to take so much of your time before I tried this.
Thoughts?
Hi Charles,
Interesting. I think that indicates sig GND on the drives is a source of noise. You might look at with your scope relative to Kanalog GND.
Regards
TK http://dynomotion.com
Hi Tom,
As you suggested, I checked the AMC drive sig GND with the scope. Trying to remember the little I knew about using a scope. Looks like a lot of noise on the drive sig GND without moving the axis. See first picture, checking X drive. Noise is much worse when moving either axis. Second picture is when moving axis. Spindle still trembles a little when enabled.
Hi Charles,
That's huge more than 15V? You might do a sanity check by checking something like Kanalog +5V. It should be 5V with small noise < 0.1V.
The period seems to be ~50us or 20KHz likely the switching frequency of the amp.
Regards
TK http://dynomotion.com
Hi Tom,
I may have something wrong in the scope setup. Using 10X on the probe.
Checked the 5V on Kanalog to Kanalog ground. See first image.
Checked the AMC drive Signal Ground to Kanalog ground. AMC block diagram notes that Signal ground and Power Ground are internally connected for "Non-ISO" configuration. See second image.
Hi Charles,
+5V test shows scaling is correct. 20V differences between GNDs seems crazy. I suppose one or the other is floating.
Regards
TK http://dynomotion.com
Hi Tom,
Will look at scope readings again.
New problem: When I try to Save/Compile/Download/Run a C program, I get a message:
DSP_KFLOP.out Date Stamp Doesn't match KFLOP Firmware
It says the Firmware is 4.35f and the .out file is 4.34.
It tries to load from the 4.34_back folder. I tried to force it to load from the 4.35f folder, but it comes back after loading and restarting with the same mismatch error.
Thanks
See attached
Make sure you are running the KMotion435f\KMotion\Release\Kmotion.exe. Check Help About.
Regards
TK http://dynomotion.com
That did it. I had the shortcut going to KMotion434.
Private message sent.
Thanks
Other priorities took me away from the lathe for a bit.
Now trying to figure out why some moves in Z axis stop along the way, with DRO still counting, then after a bit, Console shows "Following error disabled axis:1". This will happen when I Home the machine, Z axis only, or sometimes when feeding Z axis. Never while jogging or doing Step Response moves, even with 2" step response moves. Never X axis. Position error is about 60 counts max. Max following error is set to 100000.
Hi Charles,
Well that would seem to indicate the Amplifier or the Supply switched off, motion stopped, then after 100000counts of further motion a following error occurs. What is required to recover?
Regards
TK http://dynomotion.com
Hi Tom,
Thanks for your response.
When I hit Home, X axis homes ok. When Z axis moves toward the home switch, it will randomly stop, with the DRO still counting. Shortly the DRO will stop and turn yellow. Console screen shows Following error.
To recover, I hit Init, but it waits a couple of seconds, then a box shows up with "Timeout waiting on Action to Complete".
I hit OK and then Init will work.
This Z axis stoppage also occurs randomly when running Gcode, but not when manually jogging.
The home C program uses Jog commands.
Hi Charles,
So again that would indicate KFLOP is commanding the Amplifier to move and it isn't moving for some reason. You might monitor the DAC voltage with a voltmeter to verify KFLOP is commanding the motor to move but it is not. You might also set the Following error to infinity so it remains in that mode forever so you have time to track down what is wrong.
Regards
TK http://dynomotion.com