Surprising manufacturing error in Thurlby Thandar PL330QMD

[cbutlera]

I've just been repairing my TTi PL330 Quad Mode Dual power supply, which has recently failed after many years of reliable service.  I was surprised to find a major manufacturing fault.  The centre tapped winding that provides the auxiliary power to the control board was wired incorrectly at PJ6 on the driver board connector.  Pins 1 and 3 were swapped over, see bottom centre of the schematic below, and the connector circled in red in the photo.  The photo was taken after my repair, so shows the wiring in the correct position.  When I found it, the black wire from the centre tap of the winding was swapped with the orange wire at the other end of PJ6.   

This meant that what was supposed to be a full wave dual rail bridge rectifier, was acting as a half wave dual rail rectifier, supplying nominally twice the intended voltage!  This caused all three of the auxiliary supply regulators to run hot, particularly R59, which together with D11 forms a shunt regulator.  This poor 1/4 Watt resistor had been dissipating over 1 Watt, and had burnt off all of its coating, but it was still working.  The failure was in the soldered joints for R59, which became dry and went open circuit, it also badly damaged the pads.  The resultant loss of the -5 V rail caused the main series regulator to switch fully on with no current limit, blowing up the device I was testing at the time.

I replaced R59 - circled in blue in the photo - before I discovered the root cause, so I fitted a 2 Watt resistor, which I stood off the board.  It turns out that this wasn't necessary, but I'll leave it as it is.  For reference, I placed the original resistor at the foot of the replacement in the photo.  I also replaced the nearby electrolytics C22, C23 and C24 which had suffered from the heat and developed a high ESR, in addition, C22 had been running beyond its rated voltage for a couple of decades.

The PL330QMD contains two almost identical single rail power supplies, and the same fault was present in both of them.  I assume that the wires and connectors were fitted to the transformers before they were installed in the chassis.  Which makes me think that my power supply is probably not the only one with this fault, as it seems unlikely that that my unit got the only two incorrectly wired transformers.  I'm not sure of the exact date of manufacture, but the transformers are dated 05.09.0 and 26.09.0.  If anyone has a similar PL series unit, it could be worth checking.  No need to take the cover off, you can peer in through the slots to check PJ6 - on both sides if it is a dual supply.  The black wire should be as shown circled in red in the photo.

I recalibrated the power supply, and it is now working perfectly.  The idling power has also dropped from 31 to 26 Watts.

[bd139]

Ooooh nasty. Thanks for reporting.

I've got a PL154 and PL310 here from around the same dates. Guess what I'm doing tomorrow!

Edit: found a picture of inside the PL154 which appears to be wired correctly. Looks like driver board is different revision as well.

[cbutlera]

Interestingly, the way connector was originally wired, matched the embossed numbering on the connector.  However, to make pin 1 on the connector align with the white triangle on the PCB, which I assume is the pin 1 marker, the connector would have to be oriented so that the wires exited on the inboard side of the connector.  I could have fixed it just by rotating the connector through 180 degrees, as there is no key to prevent it, but I though that looked untidy, so swapped the pins instead.

[bd139]

I would have swapped pins too.

I reckon you got a “Friday afternoon job” there

[blackbird]

Just checked my PL330QMD, 1999 vintage based on transformer date. PJ6 is correctly wired on both boards. 

On the backside of the master pcb is a power resistor mounted. A very quick scan on the schematics could not reveal if this resistor should be here or if it is a result of a repair job.

[PTR_1275]

That looks like the current shunt. What value is it??

Edit... looks to be a R10 which is 100mR

[thm_w]

Non polarized non-locking pins too, wouldn't be hard for someone to screw up connecting one of those backwards or off by a pin, etc.