Why are my tubes running so hot?

[joshdfrazier]

Hey thanks for all of the suggestions. The PT is not warm ever, and when the master and gain are maxed out, there is barely any ac voltage on the output. Just a tiny bit from hum.

I am determining the current across the 1ohm resistors. And yes, they are definitely 1ohm exactly! I had another amp that had terribly uneven current draw and one of my 1r turned out to be over 6 ohms! For the record, every resistor was triple checked before going inside this amp.

The amp sounds good to me. Not a whole lot of character, but good dynamics. Which is what I want. Just a very loud, clean signal for my pedals. Emphasis on loud. So, I really need as much power as I can squeeze out of these 6L6s and 60 watt iron.

[potatofarmer]

Only -30 V for bias seems pretty hot to me. I know that's not really a reliable method for determining plate dissipation, but with 400V on the plates I'd expect something in the -40 to -50V ballpark.

Have you tried any other methods of determining idle current / dissipation? I like this one:

The plate current can also be measured by first measuring the resistance across each side of the output transformer primary (it will usually be different on each side) with the power off. Make a note of the resistance on each side, and then, with the amplifier on, measure the DC voltage drop across each side of the output transformer. Divide this number by the previously measured resistance, and you end up with the plate current for the tubes on that side. Again, if there is more than one tube on each side, you must divide the total current by the number of tubes. This method is extremely accurate, and much safer than the shunt current measurement method, because a slip of the probe won't short anything out due to the high resistance of the voltage measurement setting on the meter compared to the very low resistance of the current measurement setting. You can also make a safer measurement by clipping the negative side of the voltmeter on ground, and measuring the center-tap voltage of the output transformer and the voltage at the plate of each output tube. Subtract the plate voltage from the center-tap voltage and you have the voltage drop across each side, and can then use this to calculate the current in each tube, again dividing by the number of tubes on each side.

http://www.aikenamps.com/index.php/the- ... on-biasing

Get a pair of insulated alligator clips for your multimeter probes and it's completely hands off.

It also might be worthwhile to double check your idle screen current by checking the voltage drop across the screen stoppers.

One (probably completely unrelated) concern is that 120R resistor for your screen supply - that's not much filtering. I would up that resistor to 1k. Can't imagine you'd need more than a 10W there. I don't feel like doing the math at the moment but you can probably get away with a lower rating than that, even.

[tubeswell]

What potatofarmer said. You'd normally expect to see around -30V to -35V bias on a 6L6 where the plate voltage was around 360, like in a brown tremolux or vibrolux. With higher plate voltages (like 400 to 460), you'd expect to see more like -45V to -55V bias.

Are you certain you have 1R resistance for measuring the tube current? It's difficult to measure 1R accurately with a multimeter, because there is usually more resistance in the meter leads than there is in a 1R resistor. You need a really accurate meter. You also need to ensure the battery isn't going flat and giving you bad readings. And you need to compare the reading for the supposed 1R resistance with a reading where you short the two multi-meter leads together.

The best type of resistor for this application are the lowest tolerance you can find (like 0.6W 1% metal film resistors). Using two or more parallel resistors ( e.g.: 2 || 2R 1% ) ensures even better tolerance. Otherwise go for a higher resistance, like 10R, and drop a decimal point off your meter reading to get the mV.

[joshdfrazier]

True about the bias voltage. But this is something that puzzles me: when you increase the bias voltage to say, -40, the plate and screen voltages drop considerably, from 400 to say, 360-370. These are just estimates, I don't h e the amp in front of me to test. Is this normal? Is it just a matter of finding that "sweet spot"?

[Tillydog]

...when you increase the bias voltage to say, -40, the plate and screen voltages drop considerably...

...Is this normal?

Noooo.... (IMHO)


Normally, the plate voltage (and general B+) would rise as you make the bias more negative (-30 to -40, say) due to the lower current draw.

Something seems a bit screwy....Pictures?

[joshdfrazier]

Oh no... Could I have my bias voltage and grid voltages on the wrong screens?! I will check when I get home and post pics.

[Stevem]

Yea, that is not right and if that turns out to be the case I am quite surprised that the screens in the outputs have not blown open!

[Tillydog]

Oh no... Could I have my bias voltage and grid voltages on the wrong screens?! I will check when I get home and post pics.

Maybe! :

I wanted the amp to run 6L6, so I switched the grid resistors to 470r...

I hope you meant screen grid resistors (R14 & R15) ?

Your signal / bias voltage should end up on pin 5. Your screen resistors (R14 & R15) should go from B+ to pin 4 (per the schematic).

Note that there is no connection to pin 1 for the 6L6 (there is an internal connection instead), but I don't think it will affect anything if you have wired pin 1 up - I guess the 6L6s probably don't even have a pin in that position.

[joshdfrazier]

Tilly, yes it was r14 and r15. I've checked my work many times over, I should have known that I couldn't have made such a stupid mistake as wiring the power tubes incorrectly. Wishful thinking that it would be something so easy.

Here are some pics, quite a few since the space is small and its a little messy. The circuit board in the center is the bias supply.

[Tillydog]

Well, I can't find any errors in your wiring. A couple of solder joints that look a bit iffy from the angle that the pictures were taken, but I don't think they're particularly relevant - FYI: The 470R to the 6L6 on the right of the photo, and the junction beteen the 100k grid leak resistor and the coupling cap after the PI (the LH or 'back' one).

It might be worth examining / reflowing the joints as necesary, in case there is some intermittent contact somewhere.

Are you measuring a steady bias voltage at the 6L6 socket?

[joshdfrazier]

The bias pot was a little wonky. I completely rebuilt the bias circuit and used exact values, and a brand new 2 watt locking pot. I get about 28 volts of AC current on the output at peak, and I'm going to guess around 1.2 to 1.5 amps. which would be around 38-40 watts of power, i think. again, current is just a guess. Still, at -32vdc (exactly) I'm getting 18ma and 20ma of current draw at idle using the shunt method. i will try the other method that potatofarmer mentioned earlier.. Also, is there any reason my grid 2 voltage is always about 2 volts higher than my plate voltage? it should be about 5 volts LOWER than the plates.

Anyway, I ordered a fresh set of tubes for this guy a few days ago. Fingers crossed!

[joshdfrazier]

Alright, my new JJ 6L6GCs came into today. Popped them in, and hoped for the best. They draw more current, 31ma a piece, but my b+ plate and screen voltage dropped. At -31vdc the plates are at 380. I am getting just as much if not more distortion.

So at this point, I think I'm going to grab a pair of EL34's, put the 1k resistors back on g2, and see if this improves anything. The amp was designed to run El34, so I'm sure there are just some modifications required that are beyond my knowledge.