slipperygrey
Well-known member
I was able to get the TDK manufactured power supply for a recently acquired Performa 476 working and reliably powering the computer after a fair amount of repair work, but both the +5V and +12V rails are producing somewhat above-spec voltages. The +5V rail stabilizes at 5.2V, and the 12V rail at 12.3V when powering the system. My main question for the experts on this forum is, would this level of mild overvoltage be potentially harmful to the system in the long run?
Additionally, in order to achieve this level of voltage I had to adjust the potentiometer on the low voltage side of the PSU all the way counter-clockwise (as far as it physically goes). Before this adjustment with the potentiometer in the neutral (12 o'clock) position, I had +5.4V and +12.7V respectively, which seemed too high for peace of mind. Another question for y'all is if the additional voltage regulation would cause problems in the PSU in the long run, e.g. due to additional heat generated? Am I just masking some other underlying problem by doing this?
One interesting finding is that the PSU in these machines are definitely designed to drive a SCSI HDD as well as a relatively power hungry fan (both on the 12V rail). At one point during testing I had both disconnected, which cased voltages to climb above 14V which in turn caused the PSU to start resetting rapidly (I assume due to internal overvoltage protection.) A SCSI2SD device draws way less power than a HDD, not surprisingly, which can also add to the overvoltage problems. Something to keep in mind if you are fitting such a device in your pizza box Mac...
A final tip that I wanted to leave here, for anyone else attempting to repair a TDK power supply and running into extreme overvoltage, is to test / refit / replace diodes on the low voltage side. If it was heat or cap electrolytes (or both) I don't know for sure, but I had bad traces and/or cracked solder joints on all four diodes. Simply desoldering, cleaning the legs up, and refitting them made a huge improvement on my specimen. One bodge wire was required to fix a trace. Hope this helps someone!
Additionally, in order to achieve this level of voltage I had to adjust the potentiometer on the low voltage side of the PSU all the way counter-clockwise (as far as it physically goes). Before this adjustment with the potentiometer in the neutral (12 o'clock) position, I had +5.4V and +12.7V respectively, which seemed too high for peace of mind. Another question for y'all is if the additional voltage regulation would cause problems in the PSU in the long run, e.g. due to additional heat generated? Am I just masking some other underlying problem by doing this?
One interesting finding is that the PSU in these machines are definitely designed to drive a SCSI HDD as well as a relatively power hungry fan (both on the 12V rail). At one point during testing I had both disconnected, which cased voltages to climb above 14V which in turn caused the PSU to start resetting rapidly (I assume due to internal overvoltage protection.) A SCSI2SD device draws way less power than a HDD, not surprisingly, which can also add to the overvoltage problems. Something to keep in mind if you are fitting such a device in your pizza box Mac...
A final tip that I wanted to leave here, for anyone else attempting to repair a TDK power supply and running into extreme overvoltage, is to test / refit / replace diodes on the low voltage side. If it was heat or cap electrolytes (or both) I don't know for sure, but I had bad traces and/or cracked solder joints on all four diodes. Simply desoldering, cleaning the legs up, and refitting them made a huge improvement on my specimen. One bodge wire was required to fix a trace. Hope this helps someone!