SS REF 400 issues

[azvrt]

I have an original (the ´94/´95 non-s/sx model) Soundstream Reference 500.

I was told this type of damage (see pic) was common when people tried to squeeze even more power out of them by keeping the amplifier in High Power mode at an impedance where it should have been in High current mode.

I was told to replace the parts indicated in red and so we did. I did not include the identical parts on the SG3524 side of the capacitor bank but we replaced them as well of course.

We carefully powered on the amp with a 5 amp fuse. No speakers and RCA, just power wires.
I noticed something was wrong immediately. Idle current was way too high, approx 4.1A
This model should idle at approx one fifth of that I'm guessing based on many, many other SS amps I've had on the test bench.
I only have had one Ref 500 in the past and it idled around 1.5A which I believe was abnormal, but even if it wasn't 4.1A is way too high.

Also, while idle current is less in High Current mode (lower rail voltage in this mode) than in High Power Mode on these amplifiers, on this unit it was exactly the same.
So we checked the rail voltage in both Modes, and it was 0.00 volt on both the positive and negative rail.
We ran out of time so we didn´t find on the schematics where to measure the 15V so we haven´t done it yet.

Anyone has an idea which parts(s) could be faulty ?
There´s a chip right next to the SG3524 (can´t be seen on the pic), maybe replace that as well ?
Didn´t notice anything peculiar when performing an ohms test on the rectifiers with the DMM.

Anyway besides replacing the parts we did a nice job on the amp, cleaning it up. Looks very nice now. Would be nice if I can put it to good use, too

[azvrt]

I'm sorry I see I mentioned Ref 400 in the title, it's a 500 as mentioned in the text.
Also, I forgot to mention that Q101 and Q106 become excessively hot after seconds of being powered on.

With high bias models like for instance the Class A 6.0, 10.0 etc is very common for the transistors in the outputs to generate much heat while the amplifier is idling (due to the high bias) but the transistors in the PS section normally are cool while idling.
On this one, two of the six become very very hot in a very short time.

[jol50]

For me the best thing to do with a blown PS is to replace all the mosfets and the resistors attached to them. That eliminates a lot of problems and typically parts are not expensive anyway. If the IC is running the supply but still problems, its handy to look at things with a scope. If you didn't put all the resistors in try that as they do affect the balance of the PS. Sometimes you can see different voltages on the gates to indicate a problem but not the proper way to test. Did you pull the rectifiers to disconnect the amplifier? Need to know if its the amp or PS pulling current.

Only some of the fets (transistors) getting warm usually indicates they or the resistors have an issue, or if something is between them and the IC.

[Perry Babin]

The damage to the gate resistors can happen any time the FETs fail. If you install them standing 1/4" off of the board, it will prevent the board from being further damaged if they fail again.

If 2/3 FETs are overheating in one bank, confirm that you have a good connection (with your meter) between the gate resistor and the gate leg of the corresponding FET.

If all 3 FETs are getting hot in one bank, the problem is generally a defective driver transistor (most often the PNP driver).

[azvrt]

Thanks for your replies.
No we did not desolder the 4 rectifiers. Would that be safe on this model, powering on the amp without them ? After desoldering them I could replace them with higher rated ones I have in stock, FED16FT.

Hi Perry we installed all new parts "above" the board to prevent unnecessary damage to the board, yes. Thanks.

In both groups of 3 transistors each only one transistor per group was getting hot.
So one transistor on one side of the capacitor bank and one transistor on the other side.
1/3 x 2.

[Perry Babin]

If you can't determine why one is getting hot in each bank, swap the hot transistor with one next to it. Does the problem follow the transistor or the location?

The problem may be less significant in high-current mode because the duty cycle is less. That gives the transistor's drive circuit more time to get the transistor to switch off.

Look at the waveforms on the gate legs of the power supply FETs. Does the signal look different on the hot transistors?

[azvrt]

My guess is the problem will follow the location, not the transistors.
It would be too much of a coincidence that two transistors are malfunctioning while in the same (opposite) locations. We could always swap them just to be 100% sure, though.

We´ll check the waveforms and will let you know our findings.

[azvrt]

I ended up desoldering the rectifiers.
Idle current draw went from 4.1A to 0.6A
I soldered in new rectifiers and the amplifier functions perfectly now.

1.1A idle in High Power Mode and 1.0A in High Current Mode.
Rail voltage +/-39.3V in HP Mode and IIRC approx. +/-33V in HC Mode.
+/-15V is +16.xV and -15.xV, a bit on the high side and different on the pos and neg.
Would that be a problem ?

Apparently we measured the rail voltage at the wrong locations before so when I said there was 0.00V rail voltage that probably was false information. When we measured it now with the amp fully operational it was still 0V that's when we found out we had to measure elsewhere.

Q101 and Q106 do not become hot anymore.
Q103 and Q104 remain ice-cold.
Q102 and Q105 slightly warmer but still cold.
Q101 and Q106 slightly warmer.

Thanks guys ! Another amp saved

[Perry Babin]

The FETs are operating cooler because the pulse width is less when there is less current draw. I think that you need to determine why all of the FETs aren't operating at the same temperature.

[azvrt]

Originally Posted by Perry Babin

The FETs are operating cooler because the pulse width is less when there is less current draw.

Yes of course, the idling current was 4 times more when it was faulty, that will generate some heat.

Do you think the fact that Q101 and Q106 get warmer than the others is abnormal behaviour ? Could be, yes. I will ask around.
Also, I should have another REF 500 in a week or two. I don't know yet whether this one is faulty or functional. If it is functional, I'll power it on without the heatsink and see what happens with the FETs and will let you know.

Thanks again.