7i78 Joint 1 and 2 no longer working

tommylight wrote:

Michael wrote:

Joint 0 step and dir pins both swung from -4.7 to 4.2 when jogging

Joint 1 step pin swung from -5.5 to -6.2 when jogging (LCNC off was -4.65)

Joint 2 step pin swung from -5.5 to -6.1 when jogging (LCNC off was -4.45)

I would venture a guess that the drive 1 and 2 have no ground connected.
I am scared to think of other possibilities, none of them are good.

Grounds are all connected and checked. There is stray voltage coming out of the drives step pins with nothing else connected to the drives besides mains power.

All 3 seem to have this voltage to a certain extent but Joint 2 (Z) has the higher voltage. All of this was checked with just servo drives powered up. Everything else was disconnected. Is this just noise? And if so why am I picking it up across the two legs of what is basically an LED?

I checked the pins versus earth ground and they had 0v. I checked the connector housing on the drive vs earth ground and it had a similar fluctuation.

Here is what the servo internal circuit diagram looks like, I am testing pins 11 and 23:

I did some deep diving on the card and drives this morning. So the card is obviously fried with Joint 0 working fine. Joint 1 and 2 Step not working. Joint 2 Dir is also not going to -5.

Checking the drives completely disconnected both the Step and Dir pins have some trace voltage with Dir not being as bad as Step (see below for numbers). I honestly think this is just noise from the drive being picked up. It didn't matter much if they were enabled or disabled. But when I connected the DB25 to the drive the Dir pins no longer showed any voltage but the Step pins still did. Looking at how I wired this it was all in a 8 wire cable that has 4 pairs of individually shielded wires. Each individual shield was run to earth ground. I was a bit short on these individual wires and soldered them together into a single larger gauge wire to make it the remaining way to the ground bar. Could this be causing noise to stay on the step line and not the dir? Should I not run step and dir in the same cable and just use individual sets of shielded wire all the way to the grounding bar? Obviously something is working on the dir line to remove the noise and yes I did test that each shield is attached to ground.

Testing the 7i78 with nothing connected to it except the cable from the 7i92

Powered up but no LCNC
Joint 0: Step -5.03 Dir -5
Joint 1: Step 0 Dir -5.04
Joint 2: Step 0 Dir 0

Powered up with LCNC on but not enabled
Joint 0: Step 5.01 Dir 5.02
Joint 1: Step 0 Dir 5.02
Joint 2: Step 0 Dir 5.02

Powered up with LCNC on, enabled and jogging
Joint 0: Step -5 to 5 Dir -5 to 5
Joint 1: Step No response Dir -5 to 5
Joint 2: Step No response Dir 0 to 5


Testing the drives while not connected to the 7i78

Drives with no DB25 connected, testing pins only
Joint 0: Step .3 to .6 Dir .25ish
Joint 1: Step .3 to .6 Dir .25ish
Joint 2: Step .3 to .6 Dir .25ish

Drives with DB25 connected both step and dir are shielded with 7i78 end to earth ground
Joint 0: Step .3 Dir 0
Joint 1: Step .6 Dir 0
Joint 2: Step .3 Dir 0

Are these measured between STEP+ and STEP- (or DIR+ and DIR-)

In that case a 0 value is impossible unless the 7I78 driver chip(s)
are damaged

Yes measured between the different polarities on each. And joint 1 and 2 don't move. I will be sending you the card for repair probably on Wednesday.

In the meantime I am going to see what I can to reduce the noise on the lines coming from the drives.

Sent the board out for repair Thursday morning and received it Saturday afternoon. Superb! Does help to live in the same metro area. Installed and working.

As for the noise on the lines I was able to determine that the drive was producing the noise and it was being caused by the drives db25 shell not being grounded or shielded. Oddly the DMM technician I spoke with specifically told me not to tie the db25 shell/connector into the shielding for the wires. He said the same thing for the db9 encoder output off the drive as well. The manuals wring diagram specifically shows both tied to the shield and thus ground. The solution for now at least was to run a separate earth grounds to the db25 shell.

All is working fine now with no transient volts on the step/dir lines.