8i20 cold start over current error

I am setting up an 8i20 drive connected to a 7i76 and 5i25 for control and am having a problem on cold starts.

When the bus voltage capacitors for the 8i20 are discharged, I get an over current error in the drive and the outputs can't be enabled. When LinuxCNC is started with the bus capacitors charged, the drive starts and operates normally.

LinuxCNC is being used for power sequencing and inrush control so field power is present on the 7i76 and 8i20 before the bus voltage is energized.

Power sequence:
1. Main contactor engaged
2. PC powers up, 12V field voltage present on 7i76 and 8i20
3. LinuxCNC loads
4. After 1 second, Drive power contactor enabled. 8i20 rectifier is connected to power through a PTC inrush current limiter
5. After 2 seconds, the PTC current limiter is switch out of the circuit and the 8i20 rectifier is directly connected to the mains

Any thoughts on what is going wrong? The 8i20 manual does not seem to indicate a power sequencing requirement but I might have missed it.

When is the drive enabled(in LinuxCNC)? It cannot be enabled before the VBUS is greater than BUSUNDERV

PCW wrote:

When is the drive enabled(in LinuxCNC)? It cannot be enabled before the VBUS is greater than BUSUNDERV

The drive enable is driven 1 second after the inrush limiter is switched out of the circuit (after step 5). When I look at the bus voltage it seems to be stable at around 340V when the enable is driven.

What actual error do you get and when do you get it?
Also what is the 8I20 firmware revision?

PCW wrote:

What actual error do you get and when do you get it?
Also what is the 8I20 firmware revision?

I have attached a screen shot of the halscope showing how the error occurs. This is using a test hal file I am using to tune the drive. The machine-is-enabled signal turns on the contactor that connects power to the drive through the soft start PTC device.



It looks like the hm2_5i25.0.8i20.0.1.fault.overcurrent triggers when the bus voltage hits about 16V. The same sequence when the bus voltage is charged does not trigger the overcurrent flag.

The firmware revision as reported by hm2_5i25.0.8i20.0.1.swrevision is 139 (0x8B).

in enable false when you start?

PCW wrote:

in enable false when you start?

I double checked and the enable input on the connector is low when halrun is started and the script is loaded. It is enabled 2 seconds after bus power is connected. I checked both with a multimeter and halscope and the enable input was low until commanded. I also checked for glitches during init and saw nothing on the enable pin as well.

Once I run the "addf hm2_5i25.0.write thread1" command the fault light on the drive turns off and the status led starts blinking showing communication. It seems like the drive faults are cleared at this point and inspecting the fault pins on the 8i20 in halmeter show everything as no fault.

When bus power is enabled, the fault light turns on with overcurrent flag and the status led keeps blinking as expected.

I mean in a normal linuxcnc startup, when is the 8i20 enable set true (and when do you get the fault)
If I get a chance this week I'll check into the firmware rev and see if i can duplicate the issue

What's expected to happen is that enable being false masks any undervoltage or current faults

I see the same behavior during normal linuxcnc startup. The power sequencing is the same as well. I don't see any errors pop up on the screen but the spindle won't start as the drive is faulted. I don't see a way to set a spindle fault like there is for an axis fault.

I would greatly appreciate you taking a look!

I finally got a chance to take a look at this and naturally I cannot duplicate the error
either with the drive enabled or not

Tested device has same software revision (0x8B) and PCB artwork Rev. D (current rev)

Can you list your 8I20 setup parameters (with LinuxCNC running):

halcmd show param | grep 8i20 > 8i20.txt

and post the results here?

Thanks