[Bob La Londe]

Question 1:
What EXACTLY is "supposed" to happen with the G320X when an encoder fails?

Question 2:
Is unplugging the encoder a good simulation of a failure?

No, I don't have a bad encoder, but I have a reason for asking.

Questions 3:
Does GeckoDrive still monitor this forum?

I can send them an email and get a response usually within a week, but I prefer to ask here so the answers can help other people.

[Al_The_Man]

On most controls when the drive or controller would activate a move and not 'see' an encoder response, the drive would shut down or error.
The difference between anticipated position and actual is known as Following Error.
There will always be some in a servo move, but if out of the pre-determined range, the error will occur at this point.
Disconnecting the encoder is one form of simulated encoder failure.
I have no exact knowledge of what Gecko do in this case, but I would expect something along these lines?
Al.

[Mariss Freimanis]

The G320X monitors the Channel A and Channel B signal voltages. The monitor circuit faults the drive if the encoder signals are between 0.8V and 2.8V and an unconnected encoder input is biased at 1.8V.

Any broken encoder signal wire triggers a fault, a logic '1' less than 2.8V and a logic '0' greater than 0.8V also triggers a fault.

Mariss

[Bob La Londe]

The G320X monitors the Channel A and Channel B signal voltages. The monitor circuit faults the drive if the encoder signals are between 0.8V and 2.8V and an unconnected encoder input is biased at 1.8V.

Any broken encoder signal wire triggers a fault, a logic '1' less than 2.8V and a logic '0' greater than 0.8V also triggers a fault.

Mariss

That is what I figured would happen roughly.

When I unpluged an encoder with the drive idling it suddenly took off and drove the table to its limit. Fortunately I was standing next to the main blade switch for the machine and was able to cut power before it crashed into the mechanical limits. There are no external resistors or caps on the encoder or the drive.

I did not realize that the G320X monitored the encoder for a failure until I started looking at the document which discusses adding a cap and a pair of resisters for noise filtration.

[Mariss Freimanis]

You probably had DIP switch 6 set to "HEDS" (ON). The G320X loses the ability to monitor the encoder signal quality when DIP switch 6 is on; unplugging the encoder then results in a servomotor runaway. With switch 6 OFF, disconnecting the encoder results in the motor immediately braking to a stop and the drive into FAULT.

Mariss

[Bob La Londe]

Thanks. I will check that.

[Bob La Londe]

I did not have dip switch 6 turned on.

For more informaiton I turned it on this morning and did some experiments. Here is what happened. (didn't test the unplug of course)

"I have been having headaches getting decent speed and acceleration out of my KMB1 conversion. The drives would fault inconsistently and intermittently during test. I am running a Smooth Stepper into a C23 breakout board which feeds GeckoDrive G320X servo drivers. My encoders are all US Digital E6 1000 lines. (4000 PPR) I was having problems at an acceleration setting in Mach 3 of 10 and speeds as low as 100 IPM. This morning I tried something new. I switched the drivers to HEDS mode and I was able to run with no problems at all. At 200ipm and 20 acceleration I had some issues. At 200 and 10 I had no issues. Any ideas? Does this thing really need to accelerate that slowly? I know the table is heavy, but I have seen bigger mills slinging the table around so fast that you can watch it crawl across the floor until its bolted down."

Does it need to be in HEDS mode for these encoders or should I do something else like the cap and resister mod?

Should I try a higher gain setting?