intermittent DC in PL400

[Othello]

Loud popping noises in one channel of my Phase Linear 400 alerted me to potential trouble, the meter was banging violently against its stop.
I disconnected the amp which has an original set of semiconductors and tried to measure the output dc.
Unfortunately it is an intermittent phenomena, the amp works like a charm on the bench, but I am reluctant to put it back in service.
The amp seems older, it has the serial number 2740 and the board has the designation 14A.
Since the intermittent behavior is probably not due to a semiconductor, I wonder what should be changed in the amp to regain confidence, so to speak.

Uwe

[wg_ski]

Check the solder joints on the VAS transistor. It's the TO-39 that's permanently welded to a small heat sink. It gets pretty hot and eventually develops an intermittent connection to collector (case). I had one put DC to the speaker and kill it because of this.

[Othello]

Originally posted by wg_ski
Check the solder joints on the VAS transistor. It's the TO-39 that's permanently welded to a small heat sink. It gets pretty hot and eventually develops an intermittent connection to collector (case). I had one put DC to the speaker and kill it because of this.

I have to ask, what is the VAS transistor?
And do you mean it got so hot it un-soldered it's leg from the circuit board or did the damage occur internally to the semiconductor?

Uwe

[wg_ski]

VAS transistor = voltage amplifier stage. It doesn't really come unsoldered, but the heat causes the connections to oxidize. Eventually the connection from the heat sink (collector connection) to the PCB fails. The 7.5k 1 watt resistor in the collector circuit gets fairly warm, too so that might be worth checking.

The Series 2's have a similar problem with the shunt regulators that feed the LF356 op amp. They get hot enough to discolor the PCB, in addition to the failures due to oxidation.

Any amp that's been around this long and seen a lot of use, and isn't forced-air cooled is going to have similar problems. But PL's topology has a really nasty habit of putting out large DC spikes when this stuff happens.

[Jan Dupont]

You can get the schematic here:
http://www.audio-circuit.dk/Schemati...00-pwr-sch.pdf

As saud, it's most likely a bas solder joint. With power on, try tapping carefully on the circuit board with the plastic back end of a screwdriver or similar tool to find the spot............

[Othello]

Well, I still don't know what you mean.
Many transistors on the pcb have their own little heat sink, and Q6 is off the pcb bolted to the chassis.
Nowhere in my amplifier is a 7.5k resistor (or I am blind!), so I am at a loss what you refer to, are you talking about a PL400 with the pcb 14A?

Uwe

[Othello]

Well, I still don't know what you mean.
Many transistors on the pcb have their own little heat sink, and Q6 is off the pcb bolted to the chassis.
Nowhere in my amplifier is a 7.5k resistor (or I am blind!), so I am at a loss what you refer to, are you talking about a PL400 with the pcb 14A?

Uwe

[wg_ski]

My bad, not all the boards are the same. It's 18k on one of my PLs, and 7.5k on the other two. And I've seen three different input stages over the years. Two diff stages, one diff stage, and an op amp. But everything following that is pretty much the same and I've seen the VAS (the 40327 or 40412) open up twice - once fatally. Any solder joint can fail upsetting the DC balance, but the usual suspects are those components that sit there and cook when the amp idles.

[Othello]

Originally posted by Jan Dupont
You can get the schematic here:
http://www.audio-circuit.dk/Schemati...00-pwr-sch.pdf

As saud, it's most likely a bas solder joint. With power on, try tapping carefully on the circuit board with the plastic back end of a screwdriver or similar tool to find the spot............

Boy did I find the spot!
I used your tapping technique as well as a temperature probe to see what ails one of my channels. I noticed Q5 was very hot (compared to the trouble free channel), about 70 degrees Celsius.
When I tapped Q7 there was a strong spark. It looked as it I had made a short between the transistor and its cooling fins, but of course that doesn't make sense, they are connected anyways. So I don't know what sparked there, the transistor was not bent or touching anything else. But now my channel shows only half a waveform the other half is lost, Q7 seems dead.

It wasn't the original anyways instead of the MM4003 I found a 2N5415, the other channel is original.
Also the AGX 5 fuses seem to have been changed and increased.

Now I need a cigarette and advise what to attempt next.

Uwe

[Jan Dupont]

A reason to that Q5 is getting so hot could be that Q7 was the reason to your problems. If Q7 had an intermittent internal short circuit, you would have the problems you described in post 1.

If you got a spark tapping on Q7 (originally a MM4003 in TO-39 case), you should not get a spark between the TO-39 case and the heatsink (unless the heatsink are touching other components or metal parts). The spike most likely came from one of the pins at the circuit board due to bad soldering.

Change Q7 and make a new test. I have attached the data for the MM4003. There are many Phase Linear threads in this forum with suggestions for replacement Q's.

If needed, contact me through PM.