Single Ended puzzle

[sepulchre]

Well, it appears I have even more of a problem. I think the PT is gone.

Okay, just checked again - it's history. Only putting out a few volts. Even the heaters are down to 1.2V and that's with no tubes.

crap

But I need to figure out what's going on before frying a new one.

[JazzGuitarGimp]

The control grid; Your schematic indicates its connection comes 'from master'. Does this mean the wiper of the master volume pot? If so, I wonder if the wiper of the pot has gone open, leaving the grid floating?

[sepulchre]

The master tests okay.

Well fellas, I've gotta hit the hay. Gotta work tonight (and every other night) so I'd better get some shut eye.

I'll get back to it tomorrow.

Thanks for your help, guys.

[jjman]

You can put a "safety" resistor on the master pot. Wiper to ground lug. I use 2.2meg or 4.7meg. Mounting it on the tube socket (grid) to ground would be even safer.

[sepulchre]

I may install a safety resistor when I reassemble the amp. While I have the PT out I intend to go through the whole board and chassis confirming everything, reflowing joints and generally straightening up, though the lead dress is not bad now. Still, I have some filter caps coming so I may replace those or at least check them with a capacitance meter (I don't have an ESR meter). I may even lift a leg on each resistor to check it. Whatever I have to do, the problem has to be fixed before powering up again. I won't risk a new PT on a guess.

JazzGuitarGimp earlier speculated that there could be a short in the tube from the screen to the cathode. If such a short existed wouldn't that cause the cathode voltage to be higher instead of lower?

All this is showing me just how ignorant I am about power amps and tubes. I know my way around preamp circuitry pretty well but this... well I guess I've got some studying to do. I'll be looking to hit the books.

Btw, Jana - I built your design on the 2204 and it sounds great. So I do have a nice amp to use while this one's on sick leave. Kudos to you for the excellent design!

[JazzGuitarGimp]

Based on the voltage drops across the two screen resistors, I'd expect the cathode voltage to be about 37V if the tube has a screen / cathode short. I don't recall you stating the cathode voltage. I'd be interested in knowing what it is.

[sepulchre]

It was about 10V, but that was not long before the PT quit. Earlier I had checked the bias current and it was down to 53mA. It had been much higher, around 90mA IIRC.

[JazzGuitarGimp]

There may be something to be learned from taking a few resistance measurements. With the 88 in the socket (no power applied), what is the resistance from the screen grid to GND? And from the screen grid to the cathode?

[sepulchre]

I don't know. I'm currently at work so it will be morning before I can take another look at it. But that sounds like a good place to start. Should try it with and without the 1K resistor connected?

[JazzGuitarGimp]

I'd start with it connected, then if you get a reading that's high enough that it might be influenced by the 1K / 5K, you can always lift the 1K.

[sepulchre]

Got it.

I thought about checking the tube for shorts without it being in a socket - just on the pins, but I haven't done so.

[JazzGuitarGimp]

Something was causing both the 1K and 5K resistors to pass about 50mA. If that current wasn't getting its return path to GND through the tube, then it was finding another path. But the fact that the 5K was not overheating with the tube removed casts a shadow if suspicion on the tube.

[sepulchre]

Well, I don't think it was overheating without the tube. The fact is, when I pulled the tube to check that, well that's when the PT died. But I had checked voltages earlier with no tubes and didn't notice it being hot then. Before you ask I don't remember if there was a large drop across it when I was checking. I might not have even checked that, but the measurements I did take seemed in line.

[sepulchre]

Well, I'm back at the bench. I'm sure you guys, at least some, are asleep or off to your jobs, etc. My schedule is screwed up so this is the time I have to work.

First I measured the cathode to ground with no power. It's at 200 ohms, give or take an ohm. Then I measured the screen grid to ground as suggested by JazzGuitarGimp. It measures 243 ohms to ground. But then I wriggled the tube in the socket (which does apparently have issues) and the reading went down to 200. I measured straight across the tube from screen to cathode: dead short.

So I took the 88 out and tried a new 6550. It's not identical to an 88 but I should be able to learn something from it. Indeed I did. The resistance rose to around 260K then slowly fell settling out a about 186K.

I had been running the KT88 with the bias set low. My figures said it wanted around 100mA and I had it running with about 87. Is that so far off the mark that it would cause problems? Is low bias even a possible cause for such a problem? Or did this tube just give up the ghost. I didn't buy it new; it came in another amp I bought, one very similar to this one. So I don't know its history.

But it's important that I know, could something else have gone wrong with one of the parts that would cause the tube to fail? Just like the PT, I don't want the same thing to happen to a new tube, and then the new PT in turn. Everything I have checked seems fine. The resistors and caps measure out in range.

[JazzGuitarGimp]

It might be a good idea to repeat the screen - cathode resistance measurement on just the tube (out of the socket). If you are still seeing a short, you might want to measure from the screen grid pin to the control grid pin (pin 5) as I imagine it got caught in the crossfire. If you have a short from screen to control grid, it's possible that the master volume pot was damaged as well. Adding a safety resistor, as jjman suggested would be a good idea, though I am not sure it would have kept the amp from melting down in this particular instance.

Not sure what actually happened. With a plate V of 356 and plate I of 0.087, you're right at about 74% dissipation at idle, which is a safe place in most PP amps, though I am not sure about SE amps.