monkeyhead
Member
Anyone ever seen anything like this?
We have a Mitsubishi FX2n-64MT-ESS/UL that uses the ABSD, absolute drum, function to control input and output timing based on the position of a 13 year old machine. A 360PPR Quadrature encoder, with the A,B and Z channels wired to inputs X0, X1, and X2 respectively, is used to determine machine position. The high speed input C252 is used to monitor the position of the encoder. Every scan of the PLC, the value of C252 is copied to C2, and this is in turn used by the ABSD function to set and reset 16 bits based on the position of the machine.
The machine also uses the SPD function, which monitors input X0. The FX Programming Manual makes mention of this function and cautions using it with other high speed functions, but as far as I understand our machine should be okay because the encoder ends up outputing about 300Hz, which is well below the 5kHz rating of the high speed inputs.
If everything was working as expected, each one of the 16 bits should set and reset 1 time with every full revolution of the encoder. What we are seeing is that sometimes one of the bits doesn't turn on as expected.
For troubleshooting purposes, I increment a variable for each bit on the rising edge. On average, the counts start to get off from each other after 500 cycles. By 2500 cycles, we've seen the counts being of as far as 6 or 7 counts.
The PLC scan time is around 30 msecs, and at minimum the bits should turn on for at least 5 scans. The PLC should have plenty of time to set and reset these bits.
So far we've tried:
-replacing the PLC (with an equally old, but less used unit)
-replacing the encoder
-disconnected and megged out all the encoder wiring to check for insulation weakness (checked against ground and braided shield in cable).
-checked to ensure the shield is properly grounded
-verified that nothing mechanically is causing the encoder to bounce or get jolted
-pulled the encoder wiring out the wireway it was in (shared with wires between VFD and motor)
-Disconnected the SCADA and HMI
-Verified all voltage sources are stable and not fluctuating
Nothing so far has worked. I wish I had a data trending oscilloscope, because I'd love to see what the encoder output looks like when this occurs.
I've thus far put about 10 hours of troubleshooting time on this, and am starting to get really frustrated by the issue. The machine is 13 years old and we haven't made any changes to it or any of the surrounding equipment for years, so I can't explain why this just randomly started happening. The machine also is fed from a step down transformer that doesn't feed any other equipment. An identical machine runs next to it and does not have this problem.
If anyone has some advice or tricks up their sleeve for an issue like this, I'd certainly appreciate it.
We have a Mitsubishi FX2n-64MT-ESS/UL that uses the ABSD, absolute drum, function to control input and output timing based on the position of a 13 year old machine. A 360PPR Quadrature encoder, with the A,B and Z channels wired to inputs X0, X1, and X2 respectively, is used to determine machine position. The high speed input C252 is used to monitor the position of the encoder. Every scan of the PLC, the value of C252 is copied to C2, and this is in turn used by the ABSD function to set and reset 16 bits based on the position of the machine.
The machine also uses the SPD function, which monitors input X0. The FX Programming Manual makes mention of this function and cautions using it with other high speed functions, but as far as I understand our machine should be okay because the encoder ends up outputing about 300Hz, which is well below the 5kHz rating of the high speed inputs.
If everything was working as expected, each one of the 16 bits should set and reset 1 time with every full revolution of the encoder. What we are seeing is that sometimes one of the bits doesn't turn on as expected.
For troubleshooting purposes, I increment a variable for each bit on the rising edge. On average, the counts start to get off from each other after 500 cycles. By 2500 cycles, we've seen the counts being of as far as 6 or 7 counts.
The PLC scan time is around 30 msecs, and at minimum the bits should turn on for at least 5 scans. The PLC should have plenty of time to set and reset these bits.
So far we've tried:
-replacing the PLC (with an equally old, but less used unit)
-replacing the encoder
-disconnected and megged out all the encoder wiring to check for insulation weakness (checked against ground and braided shield in cable).
-checked to ensure the shield is properly grounded
-verified that nothing mechanically is causing the encoder to bounce or get jolted
-pulled the encoder wiring out the wireway it was in (shared with wires between VFD and motor)
-Disconnected the SCADA and HMI
-Verified all voltage sources are stable and not fluctuating
Nothing so far has worked. I wish I had a data trending oscilloscope, because I'd love to see what the encoder output looks like when this occurs.
I've thus far put about 10 hours of troubleshooting time on this, and am starting to get really frustrated by the issue. The machine is 13 years old and we haven't made any changes to it or any of the surrounding equipment for years, so I can't explain why this just randomly started happening. The machine also is fed from a step down transformer that doesn't feed any other equipment. An identical machine runs next to it and does not have this problem.
If anyone has some advice or tricks up their sleeve for an issue like this, I'd certainly appreciate it.
