iMac A1311 PSU repair questions

[Crystaleyes]

Hi all.

I have an ongoing issue with a 2010 A1311 21.5" iMac. It starts up, chimes, lights up and boots to the OS, however the screen starts to lightly flicker, steadily getting worse before switching off altogether. Measuring the voltages reveals that the 12v rail drops to around 10.5v once the Mac is switched on.

Until yesterday, I wasn't sure if the problem lay with the PSU, the backlight inverter or the logic board.
Yesterday, however, I borrowed a good PSU and the iMac works fine.

Hence my opening a new thread within this troublshooting section.

Searching online, I could find only one other person with this problem, on the iFixit site. He then managed to trace the fault back to a shorted Mosfet on the output of his PSU and resolved the problem.

This is where I am looking for some advice.

It seems reasonable to assume considering the identical symptoms and model iMacs that the problem should be similar, or at least related.

It has become clear to me that I don't have particularly good diagnosis skils when it comes to good or bad mosfets. Mosfets are not something I've much experience with but this is how we learn.
The fella in iFixit had an easily identifiable short. The PSU here doesn't.

I have pulled three of the five fets on the board. All three read a mosfets with a Chinese T7 component tester. Two of them switch on and off fine, when using the DMM in diode mode. One doesn't (although it does show as N channel Fet with the T7).

So my first question is, "which test is more reliable? Whether the Fet turns on and off with a DMM, or the Chinese tester?"

The Fet in question is a P15N60C3 and Drain/Source either shows 'open circuit' or the reading steadily climbs from just above zero (maybe 20) rising up to the 'OL'
Datasheet is here;
https://pdf1.alldatasheet.com/datash...PI15N60C3.html

[Crystaleyes]

The thing that made me realise how little I understand Fet testing was that yesterday, whilst testing a brand new 20N60 Fet, that it wouldn't turn on or off at all with DMM's, whilst appearing to test fine on the T7.

Therefore, another questions is, "why do some Fets turn on and off with a DMM in diode, whilst others don't, and does that not make the DMM test essentially useless?"

[PeteS in CA]

DMM diode checkers source around 2V, which is too low to turn on a MOSFET. Depending on the type, MOSFETs need ~3V (for logic level types) to ~5V, Gate-to-Source, to start to turn on, and even then not fully on.

Bipolar transistors turn on and are controlled by Base-Emitter current. MOSFETs turn on and are controlled by Gate-Source voltage. To a DMM a Base-Emitter junction looks like (and is) a diode junction; a Gate-Source junction looks like an open circuit.

[Crystaleyes]

Originally Posted by PeteS in CA

DMM diode checkers source around 2V, which is too low to turn on a MOSFET. Depending on the type, MOSFETs need ~3V (for logic level types) to ~5V, Gate-to-Source, to start to turn on, and even then not fully on.

Bipolar transistors turn on and are controlled by Base-Emitter current. MOSFETs turn on and are controlled by Gate-Source voltage. To a DMM a Base-Emitter junction looks like (and is) a diode junction; a Gate-Source junction looks like an open circuit.

Thank you.

I fully get it regarding the BJT being current controlled, versus the Fet being voltage controlled, just hadn't considered the DMM having a diode-test voltage too low to trigger the Fet.

Every day's a school day, as they say...

I was actully just looking at the schematic of a 5v DIY rig to properly test Fets.


Back to the PSU, if the Fet has not failed, yet the symptoms are identical to a failed mosfet, can anyone suggest where next to look?

Or shall I just buy another PSU?

[Crystaleyes]

Originally Posted by Crystaleyes

Thank you.


Or shall I just buy another PSU?

As a final word, I threw in the towel, bought a used PSU for a tenner, and now the iMac lives again!