Fixing a QSC RMX1450a

[timsabre]

So this is a very technical question but I know some of you do electronics repairs. I offered to fix a QSC RMX1450a amp thinking it was something simple but it's got me baffled.

Symptom is clip light for R channel comes on about 2 sec after power up and remains on with no signal. The channel does output sound but very "thin" and distorted.

Amp was completely caked with dirt, it's run 24/7 for 10 years in a school with no maintenance. However after cleaning up the PCB's look OK, nothing obviously failed or overheated and no corrosion or anything.

 

http://sabretechnology.co.uk/downloads/rmx1450.png

 

 

(full schematics here)

 

On poking around the circuit, the +15V rail on the R channel is low (about 10V) with an AC ripple on it. -15V is OK. Main rails are OK. The output of U201 op-amp to the driver transistors has a large spiky 50Hz signal on it which I think is coming from the +15V noise on one end of voltage divider R218/R219.

This is the cause of the clip light but the difficulty is working out what is a cause and what is a symptom. The +15 rail is fed by R235/D212 when the amp is quiescent and R257/D216 when there is audio running. Both those 2 resistors are intact and the correct resistance. The C212 smoothing cap for +15V is OK. It looks like something is loading down the rail but it is difficult to know how to find it.

 

It is a slightly mind boggling circuit because it has all the driver transistor collectors grounded to the heatsinks and it uses a non-grounded power supply so it can basically swing the power supply rails about to generate the output. There is a service manual PDF available here which helpfully explains how the circuit works but the fault finding does not cover this situation. There's a recall for this amp with exactly this symptom, where you have to replace transistor Q205 but I have done this and it does not change the fault.

 

Do any of the electronics experts have any ideas?

 

[KevinE]

make sure the 3-wire ribbon to the fan speed control opto is making good connection as I think on some board revisions it gets its op amp 0V reference through it. I do know it can cause baffling symptoms. Also the power LED is involved in generating one of the references, in certain revisions (not all boards are the same!) so make sure that's lit ok.

In very dirty amps I've had tracking underneath the jumper cable that joins the two boards...under the pcb header itself.

I've had the odd minimelf 4148 go leaky, the odd metal oxide resistor go high and that sort of thing.

What was the Q205 mod, the one where it snaps off at board level? Later ones held the pcb tight in a rebate machined into the heatsink.

 

Sometimes the ribbon that connects the rear DIP switches to the ch1 ® pcb can go open in random places at the transition plug at the back. It can cause weird symptoms.

I've had a faulty transconductance amp which causes mischief for example allowing it to drive one way but not the other.

 

PS if the plating-through fails on the main bus caps (due to stress) it can interrupt one of the power DC lines mid-circuit with interesting results. However I'd have said this was a problem with mobile setups rather than an install.

Edited August 26, 2021 by KevinE

[timsabre]

Thanks Kevin, yes the mod was the one where the legs on the part snap. I didn't think it was likely to be that but it seemed like the first thing to try.

I might try changing some of the 1N4148's next. The resistors all meter out OK. I have the rear DIP ribbon disconnected at the moment as it's easier to work on and it makes no difference to this fault.

 

[timsabre]

Found it - R246 (emitter resistor on the positive driver transistor) was 50K instead of 6.8R. What a carry on to find it though.

[DrV]

But isn't that a great feeling when you finally find the culprit. All the sweeter if it was difficult!