PSU probs-when is a ground, not a ground.

[Diabolical Artificer]

I'm working on this old Shure mixer PSU and it's playing up. Firstly the +/-15v rails are out of whack. If I bring it up on a varaic, the rails track each other (are equal-ish) until they get to 13v approx, then the neg rail drops, the pos rail increases past +15v to +20v.

I check the 78** reg IC's, all kushti, as well as the diodes, caps etc, then started seeing what these rails are connected to. Apart from the LM339 comparators, there's a potential divider joining the two rails, that checks out. At a loss to find anything obvious I checked voltages which led me to an odd finding, IE PCB ground is at +8v approx.

If you look at the schematic you'll see a gnd on the LHS @ P501 B, this connects to main chassis gnd, however PCB gnd isn't connected to this, it's floating. It could be that this PCB ground isn't supposed to be at 0v but it connects to two power amps called non signal gnd.

With this non sig gnd (2nd schematic, P601 B) connected to main chassis gnd, the power amps work ok, but if not connecting a load on the OP causes LF oscillation and massive current draw fro the main PSU.

So, is the ground not being at 0v a problem, if so is it indicative of a fault & where to look?

TFL, Andy.

PS attached PDF's as pics not the best.

[Terry_VK5TM]

If you look at the signal/connector chart for connector P503, you will see that chassis ground and nonsignal ground is connected together.

Assuming this power supply is supplied by a transformer, there must be another connection from the centre of the winding that supplies the +&- 15v regulators that connects to ground.

[McMurdo]

is it that the various grounds are only connected together when it's all assembled? Maybe to avoid a hum loop of some sort?

[Diabolical Artificer]

"If you look at the signal/connector chart for connector P503, you will see that chassis ground and nonsignal ground is connected together." missed that Terry. As these PCB's/modules are no longer used in the mixer, all connections aren't as standard. It's making sense now, if PCB gnd wasn't connected to main chassis gnd through P503 (B), then it has no reference, the gnd will float and the +/- rails won't know where they are, hence readings given in post #1.

"is it that the various grounds are only connected together when it's all assembled? Maybe to avoid a hum loop of some sort? " Some are hard wired, some are connected by multipin connectors, overall if other PCB's aren't present then some gnd paths are OC.

I hope that's the answer, will check later when it's light, just hope the WS roof hasn't blown off.

Thanks lads, Andy.

[Diabolical Artificer]

Problem solved- ish, connected chassis gnd to PCB gnd. Rails now mostly equal though the negative rail is a tad low and a few voltages as per the schematic are off.

A.