6505+ blows output fuse on standby. Does not with no power tubes. (

[Biffstar]

Well, I got everything that I'm hoping I'll need...



After I have a bit of solderfest tonight, we'll see if it's the moment of truth...

[Biffstar]

So I've soldered in the two new bias resistors (blue on the left and blue in the middle), and the new screen grid resistor (white block, vertical middle).



The issue I have now is the socket itself. How in the hell am I supposed to desolder this thing? The pins on the underside are quite thick and huge, and have boatloads of solder. I can't soften one up at a time, because the other 7 pins still hold it down. And I obviously can't heat them all up at once to remove it. The solder sucker isn't helping much either since the solder goes pretty deep through the board, and my iron can't really heat it all up through and through.

Any tips/tricks on how to get the stupid thing out?

[The Dude]

The sucker should work with enough heat. If not, try some solder wick. Another option, if you can get in there, would be to cut the pins topside with a side cutter, get the socket out of the way, and desolder each pin separately. You'll, of course, destroy the old socket, but it's bad anyway, so who cares?

[catalin gramada]

A twisted stranded wire will works aso

[Biffstar]

So after some wrestling and solder sucking, I got the socket out:



A look at the backside of the supposed bad socket reveals nothing. Looks ok. I tried to split it open, but these things are one molded piece. I'd have to chop/dremel it apart.



Soldered in the new socket, tested all connections... perfection. All pins read the exact same resistance readings as the other sockets:



Now, the test...

With no tubes installed at all, the amp fires up just fine, and goes from standby to play without issue. Same as before.

Now I begin plugging in one tube at a time, starting with socket #4, then #3. Standby goes to play and seems stable. Same as before. No issues with those two.

I finally throw a tube into the new tube socket, and voila, tube no longer arcs, standby goes to play without popping a fuse. HOWEVER, tube begins glowing red-hot, and smoke now appears from socket #1! Which was previously a good socket! Wtf?!

The socket is likely still fine since I have no more arcing anywhere, but something somewhere is clearly still pushing too much current. It seems to be affecting sockets #1 and #2 only; those two are connected in some way. But I have no idea where the smoke would be coming from with no tube installed... A resistor below perhaps?

At this point, I'm at a loss. The power board completely checks out fine after I swapped a bunch of other components in addition to the socket, so I have no idea where to go from here. My guess is something on the mainboard now? A toasted resistor there perhaps?

I could use some guidance on where to go from here...

[g1]

A coupling cap from the PI that feeds #1 & 2 perhaps has gone leaky?
When you said you were putting them in one at a time, you were leaving them in and adding? So when 4,3,&2 were loaded #1 was empty and there was smoke?

[Jazz P Bass]

Me, I would have resoldered that board by now.

[Enzo]

Pull the tubes. Turn on the amp and standby. The amp is now ready to play except for the tubes.

Now measure DC voltage on ALL the sockets. Specifically, is there B+ (480-500 or thereabouts)on pins 3 and 4 of each? Is there -50 (more or less) on each pin 5?

WHen you plugged a tube into that socket, I would wager your bias voltage was missing and the tube ran away.


Please don't forget your meter never puts any voltage on the board, other than one volt. Things that happen at 400v won't happen with your meter. Also, don;t focus on the parts, it isn;t about the parts, it is about the circuit. You can check the 2.2k resistors, but that doesn't check the circuit they are in.

What you are calling #1-4 are actually V5,6,7, and 8 on the schematic. And on the board they are numbered thus: V5-7-6-8. In any case, the two on the left are a pair, and the two on the right are a pair. SP power off and discharged. Measure resistance from pin 5 of the first socket, to pin 5 of the second. That path is through one 2.2k and back through the other to the other socket. You should measure about 4.4k. That tests not only the 2.2k resistors, but the circuit they are in as well. Now do it with the remaining two sockets. Note the two pairs are not connected together by pin 5.

The screen resistors connect to pin 4 of each tube. They are 100 ohm. Those ARE all wired to the same place. So stick a probe into pin 4 of the end socket, then measure resistance to pin 4 of each of the other three. In each case you should see about 200 ohms.

In both the above tests, if you do not get the readings I suggest, then it would seem you have a problem on the board, if not an open resistor, then maybe a broken copper trace.

[Biffstar]

Enzo, you've been fantastic with your detailed help. Thank you plenty! (the rest of you have been helpful as well, of course)

So stick a probe into pin 4 of the end socket, then measure resistance to pin 4 of each of the other three. In each case you should see about 200 ohms.

Looks normal:

Measure resistance from pin 5 of the first socket, to pin 5 of the second. That path is through one 2.2k and back through the other to the other socket. You should measure about 4.4k. That tests not only the 2.2k resistors, but the circuit they are in as well. Now do it with the remaining two sockets.

Looks normal:

Now measure DC voltage on ALL the sockets. Specifically, is there B+ (480-500 or thereabouts)on pins 3 and 4 of each?

Pin 3 of each socket. A little higher than you specify, but I'm assuming this is normal for this amp:


Pin4 of each socket. Same story here, looks ok:


And the bias voltage at the external test points reads 50 volts as expected:



Also, the bias adjust knob/pot works correctly, with a sweep of around 43 volts to around 58 volts. All seemingly normal so far. But this is where it starts getting messed up...

Is there -50 (more or less) on each pin 5?

Pin 5 readings:


So the two rightmost sockets read zero on pin 5, and the two left sockets (the ones that smoked and arced) are reading the full 500+ voltage on pin 5 instead of the 50 volts... Clearly not right.

Thankfully, the 50 volts bias voltage is being produced just fine (happens on the main board), but it's just not making it to the power board/tubes correctly. The power board has to be ok, no? I'm assuming at this point that this would be a resistor or trace on the mainboard? Any suggestions on where on the mainboard to start?

[Jazz P Bass]

It appears that you may have a phase inverter issue. (V4).

Phase Inverter.pdf

If you see the same conditions at the preamp board header (Pin J57/ J58) then you have to move in towards V4.
As G1 surmised, you may have a bad coupling capacitor. C29 or C28. (The full plate voltage will pass through it)

The lack of bias voltage on the one pair could be a failed connection to the bias point.

[Biffstar]

If you see the same conditions at the preamp board header (Pin J57/ J58)

What conditions do you mean? What sorts of voltages should I (or shouldn't I) be seeing at those pins?

As a side note, could it even be something as simple as just a bad 12AX7 in the phase inverter slot? Would a bad PI tube cause this gigantic mess?

[Pedro Vecino]

Check value of R54 (220K) associated with pin J58. Like the two you replaced, it is discolored.

[g1]

Good eye Pedro! That would account for the 0V at pin5 of one pair of power tubes.

At pins J57 & J58 you want to see the -50V bias voltage.
C28 and C29 should each have PI plate voltage on one side, -50V bias on other side.

[Jazz P Bass]

What conditions do you mean? What sorts of voltages should I (or shouldn't I) be seeing at those pins?

As a side note, could it even be something as simple as just a bad 12AX7 in the phase inverter slot? Would a bad PI tube cause this gigantic mess?

The 'same conditions' that you have on the output tube board.
546Vdc on pin #5 of one tube pair and zero Vdc on pin #5 of the other pair.

J57/58 should be a straight shot to the output board pins.
Just trying to verify the problem.

If those odd voltages appear at the preamp board header, then the problem is before that.

Look at the schematic I attached.
V Bias comes up to the middle of two 220K resistors.
It then branches off to each 12AX7 half coupling cap and then to the header.
If you have V Bias at the junction of the 220K's, then you should have it on the other end of each resistor.
If not, then, (the one that reads zero), the 220K resistor is open.
If you see 540Vdc, then that leads to the coupling cap being bad. (the cap blocks the dc plate voltage while allowing (coupling) the ac signal voltage through.)


No. The 12AX7 cannot exhibit the symptoms by itself.

[Biffstar]

Thanks, Jazz B Bass. I will test for these things tonight. Nice catch, Pedro. I didn't notice that either. I'll test that resistor tonight as well.

I'm quite novice at reading/following schematics, so I apologize for not catching on to everyone's suggestions immediately. I hope I'm not frustrating you guys... I really am trying to follow along. It doesn't help that I only have around 1.5 hours every night to tinker with this (young baby daughter to take care of and put to sleep).

So if I'm reading the schematic right, C28/29 are the two yellow capacitors in the photo above? Each leading to pins J57 and J58?