6L6PP Cathode Bias

[Sir Cuitous]

I hope someone can point me in the right direction. I have an amp that sounds great as long as I don't dig into the strings real hard. When I do, it rasps and the volume will jump up and down. I took some measurements with the amp at idle, but I'm not good at interpreting the results. The 6L6 plates read 440v and the cathodes read 30v to ground. I also checked the common cathode current across a 1ohm resistor which I realize may not be the best way to measure that. It reads 140ma. Is that about 70ma from each tube? Do these measurements reveal anything?

[6267]

First off, you're dissipating 30w per tube *assuming* they are both drawing equally. The description of the sound could be the amp trying to drive the grids positive into A2, and failing somewhat nastily. Both problems would be solved by increasing the negative bias voltage, via increasing the cathode resistor. However, I also imagine that both tubes are not drawing current equally. I'd convert it to use two seperate cathode resistors, you can approximately double the value of the original resistor and install it on each tube. You'll need to put in seperate bypass caps of approximately the same value per tube as well. I would increase the size of this resistor until anode to cathode voltage * current in amps (for your data above, 440 * 0.070) is somewhat less. The current across the resistor includes screen current, but I think 30w is still too high. If you google Weber bias calculator, you'll find Ted Weber's excellent online calculator. Some people like to bias it a bit hotter than it recommends, but you'll not gain much for the added wear on the tubes. Quite a useful resource. I'd try a new set of 6l6s (I hope the ones you have are 6l6GC or equivalent!) before changing a bunch of cathode resistors, it could be that one tube is eating a lot of current and screwing things up.

[Enzo]

I am concerned over your description. You say the cathodes are at 30v, then you say you measured cathode current over a 1 ohm resistor. Do you have cathode resistors PLUS a 1 ohm resistor? Or did you measure the 30v then put a 1 ohm resistor from cathode to ground? You can add the 1 ohm resistor to a cathode resistor if you want, but it goes in series, NOT parallel. Paralle would give you a reading, but it would be high because you'd be killing the bias. (I prefer to simply use the cathode resistor itself. The arithmetic is not hard.)

Or did you mean the grids were at 30v.

What amp do you have? we can all look at the schematic.

[6267]

Hmm...confusing–I assumed the 1-ohm was in series. You're sure the amp is cathode biased, right? If the cathode was connected straight to ground, it's fixed biased, if one or both cathodes are connected through approximately 100-500 ohms, it's at least partially cathode biased. Smaller resistors than that on one or both cathodes (generally 1 or 10 ohms) are usually for bias test points. If you did parallel a 1-ohm with the aforementioned cathode resistor, it's probably fine, just be careful! I once accidentally gave a pair of 6v6GTs about 100ma (at around 350v, a merry 35w out of a 14w tube) of current for about 5 mins, no fires or melting, but the tung sol got very, very, very hot. The RCA surprisingly just began to glow an exciting blue. Neither redplated, surprisingly. Don't try that one at home!

[Jag]

If you did parallel a 1-ohm with the aforementioned cathode resistor, it's probably fine,

No it's not!!! That will effectively cancel the cathode biasing and the tubes will run away. SERIES would be fine but it would be better to just calculate using the value of the cathode resistor as Enzo said.

[6267]

I was hoping to imply from the context of my post that it was 'fine' as in it probably didn't kill the tube, but be careful next time and don't do that! Not necessarily as in fine, run a couple amps through the things and watch the pretty lights!

[Sir Cuitous]

The tubes are 6L6GC. The 1ohm is in series. I measured the cathode to ground and then added the 1ohm. Enzo, I know you have said in other posts to just use the cathode resistor for the calculation - you're right. I'm looking at my notes again which say 33v at the cathode (not 30). I hope I got that measurement right. I have some time this evening and I will measure again. It's a custom design, but I did draw the entire circuit schematic. However, I'm not sure that the amp maker would want the schematic posted. It is cathode biased with a common 250ohm resistor and a bypass cap like some of the older tweed Fenders with 6L6s. Though I think this amp would be better with a fixed bias, I do not want to change the existing design.

The description of the sound could be the amp trying to drive the grids positive into A2, and failing somewhat nastily.

I think that this might be what's happening. The Weber calculator (thanks for the tip) says 61.3ma for class A, and 47.7ma for class AB. This amp is biased higher than those values, so I am going to try increasing the bias resistance. This amp did sound good for year or so. I believe the tubes are not in good shape after being driven like that. I have a new set of tubes shipping to me and a few values of 10watt resistors. Is it possible that the 250ohm cathode resistor has drifted lower allowing more current to pass?

[6267]

I would suspect drifted tubes long before cathode resistors–it should be easy enough to verify the value of the cathode resistor with the tubes removed. If it's a little high, it should be easy to replace, I've found 250-270 and 500 ohm resistors (7w) in a ton of random things I've scrapped, solid state included. Use 10w if you can, although some people would prefer the cathode resistor to die instead of the tube in an overdraw situation. 250 is a pretty reasonable value for the resistor, I'd imagine. B+ unloaded is like 400-450? Seperate cathode resistors (500 ohm per tube) will allow for the tubes to wear more equally among other things, and probably let you use tubes that aren't as closely matched. I'm actually about to begin a similar project–although I may switch output tubes to something cheaper.

[Sir Cuitous]

Thanks to all for the good suggestions and technical info.
B+ is 450 unloaded. I did measure the cathodes again: 34v at idle. I believe new tubes and setting the bias will do wonders for this amp. I just got some 10 watt resistors in various values. Here's the kicker. I'm a guitar maker - not amp repair, but it seems that I'm getting asked to look at amps all the time. I just talked to the wife of the owner of the amp. She remembers "the last time the amp broke down it was the tubes, but they keep blowing". Hmmmm I wished I had known that when he first brought me the amp. I am waiting for a new set of tubes to ship. I should have some results by friday. Anyway, I'm really enjoying building and working on guitar amps and I appreciate the help.

[6267]

It sounds like once the bias is back in check the amp will be a lot more reliable. I started (not too long ago too, I was born in the 80s) as a musician who wanted a tube amp but didn't have a tube wallet. Once you've got a grasp of the basics, it's easy enough to experiment, especially if you have some crappy tubes you don't worry about to test with. If you're going to get into design, make sure to get a copy of the RCA tube manual, even the 1951 has the 12ax7 at least–The RC coupled charts will give you all the component values for various B+ and desired gain values with preamp tubes. I think it's a good idea to get some old tube junk (not the little radios though, they're dangerous) like a rusty PA or something and practice different circuits on it. It's fun too because you can make it sound the way you want, not the way a company expects 75% of people playing 91235 different kinds of music to be satisfied with. I need to get me some 20w rheostats though to find ideal cathode resistor values, it can be a pain to guess and check!