Problem - muffled sound

[Paul Joppa]

Meanwhile, back at the ranch ...

I finally marked up my schematic and checked the voltages.

Terminals 11 through 15 are the Zeners, which are inactive after the shunt regulator kicks in - so there's not much point in measuring the voltages there. Since the board is misbehaving, it's only T16 and T19 that have anything useful. But the tube driver voltages are critically important to determining what's wrong.

The LEDs on the board are part of the C4S (not the shunt regulator) and the B side C4S feeds plate voltage to the driver. So - without knowing the tube voltages - I can only guess that the connection from OB on the board to B4 on the tube is the problem. If you did not remove the heat sink before re-heating the solder joint at OB, which goes to the middle lead of the large transistor on the B side, you may not have gotten that joint hot enough. That middle lead is connected to the metal flange which clips to the heat sink.

If that's not the problem, then we need voltages.

[Paully]

Ok, I will measure voltages tomorrow.  We didn't resolder anything with a heat sink.  Just the resistor we mentioned, 147K, and the 2N2907A transistor.  Though the second time I did reheat a bunch of other resistors for good measure.  But the first time I just reheated those two.  Regardless, I will get you the voltages tomorrow and also take a look at the spots you mentioned.

[2wo]

Sounds to me like you have bad connection that by luck got a bit better while you were poking around nearby.Time to tug on the wires, poke at the connections. Look for a cracked resistor, or something like that...John 

[Grainger49]

I want to point out that the problem appeared after the installation of the soft start board.  It has continued through the installation of the new iron and the pot & 2 resistors for top plate bias adjustment (associated with the soft start).

[Paully]

Thanks for the suggestions everyone, they give me some great places to start.  I am at work now but I will take a look at the connections and the voltages today and post them.  Just turning it upside down to resolder might have been enough to make something shaky reconnect that disconnects shortly after I turn it back over.  Will post back soon.  Thanks!

[Paully]

These are the voltages I was able to get:

Pins
1    0
2  3.5
3    0
4  192
5    0
6  194
7  Can't get to
8  6.6
9    0

Terminals
16  495
17     0
18     0
19  192
20     0

I am of course looking for questionable solder joints and loose connections.  I will go back and search again.  And I will start getting into the actual wires as suggested if nothing obvious jumps out.

[Paul Joppa]

This one is a puzzler.

When LED D2 does not light, but D1 does, that usually means there is no load on the C4S - which would indicate that the driver section of the 5670 is not connected to the board OB. But the voltages on pins 2, 3, and 4 indicate that the driver is drawing all the available current, limited only by the other C4S (A side) which feeds the shunt reg. That would at least explain why the shunt reg is not regulating (pin 6 should be 300v, regulated) - there's no current left to shunt, the driver side is hogging it all.

So the B side C4S is getting bias power (D1 lights up) but is dumping too much current, meaning the 215 ohm current set resistor is not doing its job. My best guess right now is that Q1 (2N2907) has been damaged, possibly by an intermittent connection to D2. If D2 opens up, you could easily apply too much voltage to Q1. I would at the least replace D1, D2, Q1, and Q2 on the B side of the board.

However, I can't make sense of seeing only 6.6v on the cathode of the shunt regulator, pin 8. I would expect it to be higher, if in fact it is not regulating at all. So there may be something wrong with one of the 149K resistors, or the B side 431 chip - or something else.

You can try replacing parts, but at this point I recommend that you simply get a new board and components, build it up carefully, and install it. The old board can become an object lesson on display, if you want to keep it out of the landfill (it's only other application).  :^)

[Paully]

I replaced Q1 and it is playing now without incident for a little while.  But we have been there before.  So I will look into replacing the board.  If one of the things you suspect fails, will the failure be catastrophic, i.e. is it going to take something expensive with it so I shouldn't play my amplifier until I replace the board?

[Paul Joppa]

While I can't predict EVERY possible failure mode, it's the solid state parts that are easy to damage. Tubes are tough!

A failure of everything could drive the 45 grid pretty high, and might damage that tube. Don't use the pricey ones until you are confident that everything is running well.

[Paully]

I was more worried about having to replace a (Magnequest) transformer or some such.  Anyway, I am going to see how it goes for a little bit then.  I suppose I should actually check the voltages now that it is running.  Thanks for the help as always!

Paul

[Grainger49]

Let's remember that rewetting solder has helped 3 times before for only hours then a failure.  We can't declare it fixed till Paul gets through the weekend and it works for a week.  I'm thinking there is something we are overlooking.  I hope I'm wrong.

While I can't predict EVERY possible failure mode, it's the solid state parts that are easy to damage. Tubes are tough!

A failure of everything could drive the 45 grid pretty high, and might damage that tube. Don't use the pricey ones until you are confident that everything is running well.

[Paully]

The Queen is getting a board together for me so I think I am just going to heed the Man's advice and replace the one that is in there.  We have spent so much on other tweaks that it is probably the wise thing to do so we know we have an amp that is operating properly.

[Paully]

Ok, so I went in and replaced some parts on the driver board two months ago and at the same time ordered a new board just in case.  The amp decided to work perfectly for the next two months until yesterday wen the problem resurfaced.  Sound is fine for a few seconds then goes bad.  Flipping it over and watching I can see that two sets of LEDS light up instantly, the last two never do.  So I went ahead and put in the new board and all new parts.  Didn't fix the problem.  I suppose I can replace all of the wires on and into the board in case one of them is bad.  Any other thoughts?  Very frustrating.

[Doc B.]

I wouldn't necessarily pull all those wires, but it might be worth touching up the connections at both ends of each wire and checking their continuity. If the board itself is good then the problem must lie in one of the connections to it. You might try monitoring the plate voltage of each half of the 5670 as the amp goes thru the "Sound is fine for a few seconds then goes bad.  Flipping it over and watching I can see that two sets of LEDS light up instantly, the last two never do" cycle. That might help to understand what is going on.

[Paully]

I did just check continuity, that was fine.  But it wouldn't hurt to touch up the solder joints to be sure.  Will do that tonight and see what I can figure out on the 5670 while I am at it.  Thanks Doc!