Parport to 7i92

Shut down everything!
Find a power cable for the PC WITH and ground pin, this is a must.
-move spindle cables as far from 7i92 as possible
-use very short cabling on the 7i92 headers
-do not use a parallel cable from 7i92 to BOB if you have the M version, just plug the 7i92M to the BOB
-use a grounded cable for the spindle drive and the spindle, also not negotiable !
The 9V peak-to-peak on the 5V supply is bad, very bad.

Yes as soon as I saw the scope I turned off everything.The VFD and the 40V power supply are grounded.The PC is powered with a laptop 19V adaptor + a PicoPSU, so I'm not shure how to do this.
But I ordered a those filters for the 5V to the 7i92 and for the VFD.

Hello
I installed the 2 above filters and some ferrite as describe here
Now it's much better.
Thanks for your help

The 5v was better, but my probe could sometime trip without input.
So I ordered a Huanyang HY02D223B (2.2kw) and removed the yl620a. I switched to hy_vfd driver.

First tests with front panel commands are fine, 5V is completely clean.

Then switched to RS845 and set :

• PD001 = 2 (set communication port as source of run commands)
• PD002 = 2 (set communication port as source of operating frequency)
• PD163 = 1 (slave address 1)
• PD164 = 1 (baud rate 9600 bps)
• PD165 = 3 (8N1 for RTU mode)

In LinuxCnc with debug enabled, I have errors on all messages. see attached  "debug_22kw.txt". It seems the received message is missing 1 byte

As I couldn't find the problem, I unplunged the HY02D223B (2.2kw) and switch to a spare HY01D523B (1.5kw).
Without any changes on LinuCnc config the comm is fine. see attached "debug_15kw.txt"

I reconnected the 2.2KW VFD and same problem again.
So the problem seems to be linked to the VFD
Factory reset for parameters and reprogramming didn't solve anything.
Both unit have the same software version : ASr 1.0  and  Vr 1.00

I downloaded and installed SpindleTalker2 on windows... ans I CAN control the VFD !!!
So the RS485 VFD is not broken

The 2.2Kw is receiving commands, but responds it sends are gardbage

Check the spindle talker how many data bits, stop bits and parity bits, then set the same on the Linux machine.

AS you can see in the log files:
spindle-vfd: device='/dev/ttyS0', baud=9600, bits=8, parity='none', stopbits=1, address=1, debug=1, PID=6030
and in the VFD:
PD164 = 1 (baud rate 9600 bps)
PD165 = 3 (8N1 for RTU mode)

The commands are understood and executed : I can make it run with by Spindle Taker 2. This validates the comms parameters on the VFD side.

But hy_vfd loops on his first read query as it gets an invalid response: only sending
hycomm query = [01][01][03][05][00][00][2C][4F]
to get the ACV from the inverter

I saw people adding 120 ohms termination resistors on the RS485, between A and B.
Any hint on this ?

Adding de resistor did not change anything