[MITSUBISHI SERVO] Home Position keeps shifting after the machine is restarted.

Hardware

PLC: Mitsubishi Q06UDV

Servo Module: Mitsubishi QD77MS2

Servo Amplifier: Mitsubishi MR-JE Series Amplifier

Customer has restarted their production after 1.5 months of lock-down.

Yesterday, I got a call that a servo is not returning to its home position. This servo uses 'Data Set' method as home position return method. The servo is moved to the home position (indicated by a proximity sensor) and a 'Set as Home Pos.' button is pressed on the HMI screen. The current position of the servo is set as 0.

Yesterday, during an auto cycle, the servo had completed its positioning, but had stopped 20 mm below the home position. When the servo was moved to the home position, the position was -21.60 mm. After teaching the home position, we ran a 5 cycles, there were no issues.

Today, when the machine was switched ON, the home position had shifted from 0 to -81.3 mm.

In addition, the servo has a bunch of hardware problems: heating, noise, jerk, and motor current is 40% during idle and > 100% during operation.

I'm not sure what the exact issue is?

Edit: Talked to the programmer who had commissioned this machine. He mentioned that there was some issue with the encoder cable, which was causing the motor to heat up.

I have had a couple of issues on servos, there are some things to look at.
1. We had a problem where it would on occasion not get back to home position. This turned out to be a curly air supply that had been shortened and it pulled the servo back off the 0 position, however, this showed up as being 20mm or so from 0 not showing 0 but actually reading the correct distance (does not sound like your problem.
2. Mechanical issues i.e. warn bearings or loose encoder etc. on the drive, this can cause the encoder to jump about a bit. have had this many times so a 6 monthly PPM was issued to replace certain bearings/couplings & check encoder was secure.
I suggest that these are the probable issues, encoders can go bad but generally they either fail completely or are worn, loose or other mechanical parts like bearings. It is unlikely to be software, however, I did once have one where the variable for the zero position (this was an offset from actual) was resetting to 0 when power was dropped off the PLC, It turned out to be a flat battery as the program was held in Eprom but relied on the battery for retentive memory.

One of the maintenance engineer at the site mentioned that the coupling might be loose. He said he would get it fixed, not sure if it was carried out.

Also, I talked to the programmer who had commissioned this machine. He mentioned that there was some issue with the encoder cable, which was causing the motor to heat up. This might have something to do with the position data.

Yes a cable with a bit of a break in it can cause issues, not sure how this would make the motor heat up unless you are talking about the motor supply cable and not the encoder one.

If encoder is not mounted exactly at predefined position, you will have higher heating, current and noise issues. I have seen that issue several times. If you are lucky, you will have markings on shaft and on encoder. Mitsubishi is setting encoders on servo at 150 degrees.