Phase linear 700b help!

[6267]

I posted this over at diyaudio too, but hopefully there's someone here that can help.

I got a phase linear 700b, it was trashed. I got one channel working, and all was well until I shorted out the B- trace to ground accidentally. Tada, blown driver board. I've replaced the input pair with original MMPS5172, and the second pair with BF421 I scrounged. The VAS (was an RCA part) I replaced with a 2sd669A which has held. All other transistors test good.

I have all the outputs removed except the big drivers (also RCA parts). The amp is sitting on the + rail, and the first diff pair has the same voltage for collector and emitter. Yuck. To make matters strange, my wiring harness is not matched to the factory color codes, so I had to trace each wire and compare to the schematic (so 82% guarantee!).

Some voltages
Speaker output +94v
Input pair emitters 2.5v
Input pair sources 2.7v
Input pair gate #1 0.01v
Input pair gate #2 3.5v

I think there's a problem with the feedback, but it might be a grounding issue? if it matters, the leads to the meter board are disconnected. I've got it right here if there's any measurements that could be useful. I'm kind of a newbie at solid state stuff, so this is a learning project!

[Enzo]

Normally, I wouldn't expect shorting a power supply rail to ground would hurt any of the channel parts. SO the first thing I want to ask is this - is the V- fuse intact? The -100v would still be available for the small parts, just not for the output stage. And a loss of V- would allow the output stage to snap up to V+.

And check for V- at the ballast resistors under Q14,16,18,etc. I'd say R42,44,46, but two of those numbers were used twice on the drawing. When you shorted V- to ground, it might o0r might not have popped the V- fuse, but it could also have burnt open a trace somewhere in the process, so even with a good fuse, check for V- out where it is headed, those resistors.

The small parts stages would try to compensate, but would be powerless (tee hee) to do so.

perhaps we ought not refer to Q1,2 as having sources and gates.

[Gtr_tech]

This really isn't the right place for this type of amp, although you may get some help here. Audio Karma is the place for vintage home audio type stuff like your "flame linear"

[6267]

Gtr_tech, I wholeheartedly respect your opinion, but it's my understanding that the flame linear was used as much (if not more) for live sound and/or studio bass amp work in the 70s than for listening to vivaldi records on sansui turntables! I figured maybe someone here'd worked with them before.

Enzo, the worst part is that all the fuses hold, even underfused to 2-3A now. I've also seen it rail negative, which is odd.

When the trace shorted, it blew itself off the board. I had to lay resistor legs across burned traces and solder them down. I think what happened was an amp-killing 100-200v transient in the preamp. Somehow, the output transistors survived, but the 150 vceo second pair ate it, as did the RCA40412VI. I replaced those and all the 100v electrolytics (replaced with 250v electrolytics just in case). I have the current limiting circuit removed by lifting D9/D10 which the factory recommended during testing.

Everything else survived, but it looks like something is funky with the +22 B+ for the input pair. I've checked and rechecked transistors in-circuit and out, and verified B+ & B- are coming into the board. Things start to get odd with B+ around the BF421 second pair, I have +18 at the top of R9, I believe I measured a *negative* value at the bottom.

[Enzo]

"A dollar a watt." The Crown DC300 was the holy grail amop back in the late 1960s early 70s. The the dollar a watt PL700 showed up. Everyone wanted one. When I toured with Stanley Steamer, we had a couple for our PA.

Just a thought, lower left, see the ground to ground resistor R2? Any chance it is open? My spider sense is twitching loss of input ground causing loss of reference. Like any number of power amps that must have the signal ground attached to its input.

Divide and conquer. Q7,10 are your predrivers, The bias circuit (Q6 and friends) keeps their bases a couple volts apart. Q5 then moves that whole mess up and down to drive them with signal. So. The predrivers show at +1.6 and -0.7v on the schemo. SO 2.3v. Regardless of what the voltages are, are those two points about 2.3v apart? Your thing is not blowing fuses or drawing heavy, right? That leads me to believe the half to the right of Q6 is OK. And if it is, then it will simply go where Q5 tells it to go, in your case +100 or -100.

Assuming now... assuming the power stage is indeed OK, then what is Q5 doing? If it is not conducting or is open, then it can;t pull down on the bias string, and the whole output stage will snap up to V+. So, is the base of Q5 driven at all?

How about the little supply up top at D3,C3? If that is not present, then there can be no current through R12, Q4 to pull on the base of Q5.

[6267]

Must have MOSFETs on my mind, meant base & collector! Time for another brain transplant

Base of Q1 reads 6.5k to ground with volume wide open. Bottom of R3 measures straight to chassis ground. I installed a 3-prong cord and attached the safety ground to the chassis, if that matters. This thing had a cut cord when I found it. In circuit, it reads 0 volts to ground with power on. The base of Q2 generally hovers around 2.6 to 3.5v. New to solid state, this is the feedback input, right?

Collector and emitter of each generally is the same value or close. They are MMPS5172 as Bob Carver intended.

Q6 and Q5 seem ok, base of Q5 reads +0.4v higher. It's a 2SD669A and has survived + and - railing of the amp. Emitter is always around -100.2 depending on line voltage. I'll remeasure the intersection of R20 and the collector of Q5, it has varied depending on which way the amp railed.

Q6 tests ok, D6-8 test ok. I'll remeasure the top of D6.

It seems like when it's railed only half the power transistors are on, and the driver (Q11 Q12) for the opposite side gets very hot. It's like one side isn't pushing ground away from it properly. I have seen it rail positive and negative, without much input from me.

Thanks for the help Enzo! I was starting to wish I hadn't gotten it to work once.

[6267]

Q5 base is -100.0, emitter is -100.7, collector is -100.4
Q6 collector is at -98.6 gate at -99.7 and emitter at -100.4
1.8v Vceo there, looks right.

Speaker terminals read -100.

Checked Q5 as a "diode", both collector and emitter show about .5-.6 from the base.

Installed RCA drivers and 1-pair of PL909, fuses still hold.

I should note the most extensive board damage was near the resistors associated with Q5 (and it was shorted) originally. I'll make sure I have a solid connection on all the repaired traces.

Update -
D4 shorted for second time. Hmm. Any diode with the appropriate drop is okay here, right?

[6267]

2nd update -
-98.3v on base of Q1, -99 on emitter. There is no R2 on my board, and the solder pads look untouched. This confuses me. I should point out that I found this with 14amp fuses in every holder, a cut power cord, and badly soldered volume pots. It may have been removed by a previous "owner".

Resistance from the signal ground lead to the board to chassis is 0. Voltage on signal input ground is 0. Voltage on signal input is -98? Output is -100 or so.

Hmm...is that just a "direct coupled" thing? B+ is available to both the board and outputs, B- as well. All fuses hold. One TO-3 driver gets hot, probably because it's got 100 volts across it and there's no "sharing".

3rd update!
Q1 & Q2 were short, and may have been for longer than I thought. I think I was getting the diode drop of the two diodes between the collectors, or some other way. Explains the voltages, anyhow!

[Steve Conner]

With 100V rails and 700 watts, and probably terrible crossover distortion at living-room power levels, I think this hardly qualifies as a "home stereo" :-) But these amps do look great in any audio geek's living room, with their handles, heatsink fins and big VU meters.

Richard Burwen's famous 169-speaker stereo is powered by racks of Phase Linears, but he replaced the driver circuitry in them with something better. I think he eventually canned them and went to QSCs, though.

This would be a good job for our high-voltage transistor friends, the MPSA42/92, MJE340/350, MJE15032/3, and MJ15024/5. I imagine every transistor in the thing could be replaced by one or other of these.