Juki KE-750 T & Z servo failures| 30 August, 2019
I will try outline Theta and Z-axis servo issues, we were facing last time. Story behind. Our 750 is used seldom and last time after switching it on initialization procedure ends with T axis of right head: Servo alarm. Opening front cover of 750 two drivers has indicated ALM4 failure. This failure was disappearing after several switch on/off cycles and machine was running again.
I was searching if somebody has similar problem and found post from Oriol (https://smtnet.com/Forums/index.cfm?fuseaction=view_thread&CFApp=1&Thread_ID=21410&#Message80999), this helps where to look in the beginning. Anyway I have removed and measured all e-CAPS (through as well as SMD), Have to say all through hole caps were OK, capacity within the range as well as serial resistance (below 1 Ohm depending capacity value). Usually lower capacitance CPAs have shorter lifetime that was found on 10uF/16V SMD caps that resulted in around 2 uF and Rs > 10 Ohm. I think this will be not an issue. Taking into account construction of e-CAP, looks electrolyte was leaking and affecting neighbor circuitry (corroding pads and creating leaks on the surface of the PCB).
We have found just replacing CPAs is not sufficient once the problem was on leaks on the board caused by electrolytes that escaped from the SMD caps (C11). In order to remove leaks I have used acetone and teeth brush and clean board on places where SMD CAPs were. Evaporated electrolyte damaged pads under CPAs and it was impossible to save them. I have used 10uF/25V tantals and 47uF/25V through hole caps from Panasonic (105°C 5K hours). C24, C25 were ok, need justy scribe PCB and make new connection as per images, rest of caps (C11, C19, C1) alternative connection. In some cases C1 I have attached by Cu wire once there were no more cooper on the board just vias. Such replaced caps I have fixed mechanicaly by using UV glue and finally passivated. This fix resolved ALM4 issues. Once tis issue will affect all drivers (servo amplifiers) soon or later I have refurbish all drivers including spares just for case some will fail, however they are pretty robust except of such cap issue, but who will except these machines will be still used.
Maybe one more addition, have also inspected Power-unit-PU0-S1010R-00 (checking all e-CPAs), only one C21 reported 20uF (instead 33uF) and Rs=8,3 Ohm, this cap is located next to power resistor, usual problem , increased temperature will exponentially decrease its life, therefore ofor fast identification look first on hottest places on the board(s).