gregormarwick
Diamond
- Joined
- Feb 7, 2007
- Location
- Aberdeen, UK
Just got done repairing one of these in one of our old workhorses, and it led me on a wild goose chase in the process on the back of some faulty assumptions and diagnostic oversights on my behalf. Decided to post about it here in case anyone is interested, or might be helped by it.
Over an extended period of time, the machine has been intermittently but increasingly faulting out with a z axis servo alarm when enabling the servos (cold startup or releasing e-stop etc.). Reset and try again - perseverance would make it go, but required more and more of that (this is a big clue in hindsight that I overlooked). In the winter, we would point a heater at the control cabinet for a few minutes and usually that would make it go. The servo amps are in a different cabinet on the other side of the machine.
Servo amp shows no alarm - all fine, ready to go. So, I check the cable, all fine. I check all the way back to the pin in the connector on the Z axis board, signal high, all good. So, I assume that there is a fault on the board. I'm thinking that the heater pointed at the cabinet aspect suggests a dry solder joint or something on the board. Trace it back, first thing is a resistor array. Checks out fine, but I replace it anyway. No go. Next is a ILQ74 opto isolator and a latch. I think I've hit the jackpot, replace those, but again to my disappointment, no change.
After that the routing branches off in about 10 different directions, so I decide not to continue. I have a spare board, that is slightly modified for a different configuration. I re-modify it to match the problem board, swap over the proms, set the jumpers, switches, etc. and plug it in. Same result. This seems unlikely, and I'm about to throw in the towel, but it's already late and I'm not in any rush to go home, so I decide to take a closer look at the fault signal coming into the board.
I soldered a wire onto the back of the board at the connector pin to attach my scope to. For some reason I decided to try it at this point, and it fired right up! Cold solder joint at the connector? While I'm kicking myself for not checking that first, I hit estop and try it again. No dice, servo fault. Ok, just a shitty coincidence to mess with me.
I hook up the scope, and sure enough the servo fault signal is going low for a few ms and coming back again. I do the same check right at the drive, and yes, it's the same there.
Externally, it still appears that the amplifier is working fine. All the correct leds are lit, no error on the display. I dismantle the drive and look for anything obvious on the boards. Everything looks fine. All three servos on this machine are identical, so I swap the controller board between the Y and Z axis. Still faults out on the Z axis, but I notice now that the fault signal is dropping low for a much longer period, and actually the error display is blanking out for a second when I try and enable the servos. I suspect the power supplies, so I swap the power supply boards between the Y and Z axis, sure enough the fault moves to the Y axis amp.
The power supply board has lots of electrolytic caps. I look for swelling and leaks but find no evidence of any. I pull a few of them and check them out of circuit. Under two 25v 470μF caps there is actually some evidence of leakage, and one of them measures significantly low on capacitance. I happen to have a bag of these on hand from a previous unrelated repair, so I put them in, reassemble, and immediately the machine works perfectly!
A long route to a simple repair. Have ordered complete replacement cap sets for all three servos...
I guess in the wintertime with the heater, enough warm air was working it's way into the servo amp cabinet to eventually make a difference. But that led me down the wrong path for sure.
Over an extended period of time, the machine has been intermittently but increasingly faulting out with a z axis servo alarm when enabling the servos (cold startup or releasing e-stop etc.). Reset and try again - perseverance would make it go, but required more and more of that (this is a big clue in hindsight that I overlooked). In the winter, we would point a heater at the control cabinet for a few minutes and usually that would make it go. The servo amps are in a different cabinet on the other side of the machine.
Servo amp shows no alarm - all fine, ready to go. So, I check the cable, all fine. I check all the way back to the pin in the connector on the Z axis board, signal high, all good. So, I assume that there is a fault on the board. I'm thinking that the heater pointed at the cabinet aspect suggests a dry solder joint or something on the board. Trace it back, first thing is a resistor array. Checks out fine, but I replace it anyway. No go. Next is a ILQ74 opto isolator and a latch. I think I've hit the jackpot, replace those, but again to my disappointment, no change.
After that the routing branches off in about 10 different directions, so I decide not to continue. I have a spare board, that is slightly modified for a different configuration. I re-modify it to match the problem board, swap over the proms, set the jumpers, switches, etc. and plug it in. Same result. This seems unlikely, and I'm about to throw in the towel, but it's already late and I'm not in any rush to go home, so I decide to take a closer look at the fault signal coming into the board.
I soldered a wire onto the back of the board at the connector pin to attach my scope to. For some reason I decided to try it at this point, and it fired right up! Cold solder joint at the connector? While I'm kicking myself for not checking that first, I hit estop and try it again. No dice, servo fault. Ok, just a shitty coincidence to mess with me.
I hook up the scope, and sure enough the servo fault signal is going low for a few ms and coming back again. I do the same check right at the drive, and yes, it's the same there.
Externally, it still appears that the amplifier is working fine. All the correct leds are lit, no error on the display. I dismantle the drive and look for anything obvious on the boards. Everything looks fine. All three servos on this machine are identical, so I swap the controller board between the Y and Z axis. Still faults out on the Z axis, but I notice now that the fault signal is dropping low for a much longer period, and actually the error display is blanking out for a second when I try and enable the servos. I suspect the power supplies, so I swap the power supply boards between the Y and Z axis, sure enough the fault moves to the Y axis amp.
The power supply board has lots of electrolytic caps. I look for swelling and leaks but find no evidence of any. I pull a few of them and check them out of circuit. Under two 25v 470μF caps there is actually some evidence of leakage, and one of them measures significantly low on capacitance. I happen to have a bag of these on hand from a previous unrelated repair, so I put them in, reassemble, and immediately the machine works perfectly!
A long route to a simple repair. Have ordered complete replacement cap sets for all three servos...
I guess in the wintertime with the heater, enough warm air was working it's way into the servo amp cabinet to eventually make a difference. But that led me down the wrong path for sure.