Speedball LEDs not lighting on one side

[tintin]

Just plugged in my amp today and got no sound from either channel but tubes were lit fine. Turned it over to check and the two left LEDs on the small speedball board are not lit. All other LEDs including 12AU7 are lit.

Voltage and resistance checks everything normal on all points including large board except OA small board = 170V = IA


Comparing left and right side the only difference is at R2. R2B measures 150kohm across two legs while R2A measures 0 ohms and increases steadily once multimeter is attached. Desoldering the resistor and checking its resistance it measures 150kohm so the resistor R2 itself is fine.

I redid all points on underside of small speedball board and made sure nothing was shorting. Not too sure where to check next. Broken component?

Incidentally, will it be safe to keep the large speedball board on and replace just the 2 resistors in place of the small speedball board?

 

[Paul Birkeland]

Voltage and resistance checks everything normal on all points including large board except OA small board = 170V = IA

That will make one of the outputs on the big board also bad.

Comparing left and right side the only difference is at R2. R2B measures 150kohm across two legs while R2A measures 0 ohms and increases steadily once multimeter is attached. Desoldering the resistor and checking its resistance it measures 150kohm so the resistor R2 itself is fine.

Can you post some photos of your build?

Generally I expect to see 170V there when the 12AU7 isn't glowing on one side (bad solder joint at A4/A5), there's a broken LED (you would see one socket LED not glowing and about 12V at A3 or A8), or the R1 resistors are swapped between boards (made more difficult by us changing one to a brown Dale).

The only item on that board that is grounded is one end of the 150K resistor, so if you are able to measure 0 ohms of DC resistance on either end of the 150K resistor when it's installed on the board and the board is wired in, that would be a curious thing. 

Was this amp previously working properly?

 

[tintin]

That will make one of the outputs on the big board also bad.Can you post some photos of your build?

Generally I expect to see 170V there when the 12AU7 isn't glowing on one side (bad solder joint at A4/A5), there's a broken LED (you would see one socket LED not glowing and about 12V at A3 or A8), or the R1 resistors are swapped between boards (made more difficult by us changing one to a brown Dale).

The only item on that board that is grounded is one end of the 150K resistor, so if you are able to measure 0 ohms of DC resistance on either end of the 150K resistor when it's installed on the board and the board is wired in, that would be a curious thing. 

Was this amp previously working properly?

Yes my mistake, OA tested 0V on large board, but all LEDs on the large board were lit.

I just tested the resistance across the legs (not the soldered pads) of MJE350 transistor at Q2A and they read 0ohm across all leg combinations. I assume this means the transistor is fried? Not too sure how this happened or how to prevent it from happening again. Further info if it's not the transistor;

All LEDs are lit including both for 12AU7 socket. Heater wiring is fine, glows fine on both sides. I swapped in 2 known good 12AU7s and 6080 and same problem.

The amp was previously working very well until yesterday.

The resistance across R1A measures correctly at 237ohm. Across R2A it tests 0ohms and steadily rises in resistance holding the test leads in place.

Photo not uploading right now, will attach asap.

Would unilateral failure of a transistor cause no sound in both channels? Not sure of what other tests to run to check everything else is ok. As far as I can tell, the right side of the amp all tests perfectly. Would just like to make sure before I order any new parts, since shipping and freight time is expensive and long, and don't want to have to do 2 separate shipments. Thanks

 

[Paul Birkeland]

No, a transistor shoring out on one channel wouldn't take them both down.

It is rather unusual for an MJE350 to just randomly short, I would definitely be interested in seeing some photos of what you're working with.

 

[tintin]

Wiring is not the neatest but I checked every point carefully for connectivity and made sure no mechanical or electrical shorts.

Pin 6 on the socket tests 170V=OA. Is there anything else that might cause shorts across the transistor pads? If not I’ll snip it out and test it. Strange that it’s gone out, surely its some other part

 

[Paul Birkeland]

It would definitely be worth reheating all of the joints on both of those boards other than the joints for the LEDs.  I can see many joints where the solder hasn't flowed through the board, which indicates insufficient heating.

For the resistors and MJE350, this may mean holding your soldering iron on a joint and counting to ten, then move around and heat a different leg on a different part for 10 seconds, then move back to the first part and reheat the next joint for ten seconds.  The back and forth soldering will minimize the heating of the part itself.

 

[larcenasb]

Wow, Paul, your quick and frequent support on these boards is truly exemplary. I just wanted to make sure to say, "Thanks so much, Paul!"

 

[Paul Birkeland]

You're killing me with the suspense, is your amp back up and running?

 

[tintin]

Had a busy day with yesterday, got back to it today with a fresh approach. Can't believe I didn't check the D1/D2 LEDs first. Turns out they fried somehow which was causing the 'short' across the transistor legs. Took them out and they had bidirectional continuity. Luckily have some spares when I ordered other parts a while back and was reading about how fragile they are. Put them in and functioning well

Thanks Paul for your prompt responses as always! Definitely gave me extra motivation to fix it up. While the soldering iron's out I'm going to change out R1 for some 470ohm resistors and try out some E80CC and ECC32 tubes.

I wonder what fried the LED's? I just received a voltage regulator to keep my AC at constant 240V so hopefully that helps; my wall voltage was previously fluctuating between 230-270V so I guess that could've been it. 

 

[Paul Birkeland]

I don't think I've ever seen a shorted LED, that's really odd.  They can fail if they are allowed to sit in a humid environment for a long time, then when they are soldered the absorbed moisture blasts out of them and opens them up. 

 

[Doc B]

Agreed, usually a short measured across the transistor legs means the transistor is shorted. LEDs usually fail open. Are the LEDs you took out intact enough to remeasure their resistance out of the circuit? It would be educational to know how they measure.

 

[tintin]

Thanks again Paul and Doc for the patient and ongoing support.

Broken LED that was at D1 measures 3.7kOhm and D2 measures 36.9kOhm both directions outside of circuit. I removed both transistors on the left side of small board from the circuit to test, and there are no shorts. Once in circuit there are also no shorts after replacing LEDs. My previously drifting measurement across R2A is now stable at 150kOhm too, and cant see any other abnormalities after replacing the LEDs.

I dont think my environment has been all that humid lately, only a couple rainy days the last month and mostly sunny. My house tends to be quite dry if anything. The new LEDs Ive been careful to keep in their sealed pack with moisture absorber and the blot card that indicates humidity is all blue still.

 

[tintin]

For what it's worth, I ran some measurements again this morning. Checked wall AC and it was very low at 170VAC (vs what I usually get ~248V). Guess I have significant line sag as well as swells. Plugged in amp and tested through 240VAC voltage regulated. Swapped out R1 for 470ohm and with an E80CC that I tested at 2700/2600gm both triodes

OA=OB=66VDC, IA=IB=185VDC

Large board OA=OB=101VDC
B+=185VDC

A used ECC32 yielded OA=76VDC OB=83VDC, IA=IB=173VDC
Large board OA=OB=107VDC, B+=185VDC

Nothing problematic hopefully? Sounds great to my ears 

 

[Paul Birkeland]

IA, IB, and B+ are all wired together. They will not be different by such a large amount unless there are iffy solder joints in the amp.

 

[tintin]

just reheated and retested again, I may have just been holding the test lead poorly previously. Need to buy some new leads as these are quite old and sometimes dont register. All stable at 185VDC now.

Another question.. does the plate voltage at OA/OB directly correspond to plate current, and is there a way to derive that from the circuit?

 

[Paul Birkeland]

The plate current is fixed by the C4S board.  It is set by the R1 resistors.

 

[tintin]

Ok, I am just trying to learn some more about how to read and interpret tube traces. I am looking at the E80CC data sheet, and I understand the cathode bias by the LEDs is 1.5V so does that mean I should be trying to stay between the 1V and 2V curves? If I'm reading 80V plate voltage that would correspond to about 3mA, is that correct? And if it is fixed, is there a way to find the circuit's plate current first to find the appropriate voltage range, or do I have my order around the wrong way 

 

[Paul Birkeland]

Yes, you have correctly determined the plate current by looking at the curves.

Since the cathode bias voltage is fixed and the plate current is fixed, that means the plate voltage will wander around quite a bit between different tubes and as a tube ages.  With plate loading resistors and cathode biasing resistors, you can lock down the plate voltage a bit better, but the amp won't sound as good.

 

[tintin]

Ok thanks! Is there a way to directly calculate the plate current set by the C4S board, or is it correct to always assume a good operating for any tube is 55-90V? Just not sure now on which target number or parameter (current/voltage) I should be aiming to measure/get right  first. 

 

[Paul Birkeland]

Yes, the plate current is 0.855V/R1.  No, it's not correct to assume that 55-90V is a proper plate voltage for any tube.  This would be far too high for a few tubes, and far too low for many others. 

In the Crack circuit specifically, we would like to see the driver plate voltage to be roughly 70-90V.  This leaves enough of the B+ for the 6080 to be able to do its job and keeps the dissipation across the big Speedball board reasonable (or the 3K resistors in the stock circuit).