Spindle speed scaling

I am using a Gs2 VFD to run an 18000 rpm spindle. The drive only lets me enter a maximum rpm of 9999, both in the drive setup and in the Linuxcnc driver setup over a serial connection. Using a lower rpm /hz number “works” but the speed is off. Is there a way to scale the speed?

is it a 100 400 or 800hz vfd you got
there are multiple
and see if the settings is 0-10V or 0-5V as it can also be a case

It’s a 400 hz vfd and the spindle calls for 300hz@18000 rpm. I am connecting over Modbus via the gs2 driver and an rs232 connection.

I was able to scale the speed by scaling the spindle.0.speed-out and connecting the scaled speed to the vfd speed.

I have an issue now with the at-speed pin on the gs2-vfd component. It doesn’t ever turn on. I figured that maybe I’m breaking something by scaling the speed but I also tried it without scaling the speed and it still won’t turn on. Has anyone gotten it to work?

I wonder if at_speed actually works?

The code is here:
github.com/LinuxCNC/linuxcnc/blob/master...comps/gs2_vfd.c#L392

I would have expected a bit of "fuzz" on that comparison.

Do you see the freq-cmd and freq-output HAL pins becoming exactly equal?

It works. Kind of. I think I have a different problem.

At-speed sometimes turns on immediately, sometimes after several seconds, and sometimes not at all.

Sometimes the vfd stops responding altogether and I need to cut the power to make it stop if it’s running. In order to get the communication up again, I need to unplug the usb-rs232 converter and restart Linuxcnc. I’ve tried two different usb-rs232 converters and they both behave the same.

I’m thinking its either an EMI or a ground loop issue. Or maybe something is wrong with the VFD.

You can try lowering the communication speed of the RS232, but it is much more advisable to check the grounding and shielding.

tommylight wrote: You can try lowering the communication speed of the RS232, but it is much more advisable to check the grounding and shielding.

Yeah. I assumed that the negative pin on my usb-rs232 converter was grounded and used that to ground the shield on my cable.

It’s not:


And grounded:


The noise all seems to be coming from the vfd.

andypugh wrote: I wonder if at_speed actually works?

The code is here:
github.com/LinuxCNC/linuxcnc/blob/master...comps/gs2_vfd.c#L392

I would have expected a bit of "fuzz" on that comparison.

Do you see the freq-cmd and freq-output HAL pins becoming exactly equal?

If you would, could you direct me to how I would modify the maximum RPM to something that works for me, instead of using the scaling hack. I could just change that number before I compile it as I am currently running 2.8 compiled from source. How would I then handle future upgrades without breaking my build?

I took apart a cheap USB to Modbus dongle, swapped the USB plug for a pin header so that it would fit directly to a motherboard USB header and also changed the 2-pin screw terminal block to a 3-pin with a connection to the USB ground. (not 0V)