Biasing question

[grindfix]

I'm revisiting my HK TA5000x rebuild and since I could never decide what to bias output tubes at, the receiver was on the "to do someday" shelf. The output stage is basically cathode biased but I made it adjustable to be able to use "best matched" pairs of 7355 tubes I have. If I bias treating it as cathode biased stage, at 46mA of cathode current per tube, which is a little cool for P-K=346v, the receiver runs like a stove. I'm looking for suggestions for bias point. If I were to treat it as AB1 and bias at around 35mA of cathode current, what is the best and most accurate way to catch crossover/notch distortions if it ends up too cool? Or, is there a way to determine the optimal bias point judging by the cathode current difference between idle and max power? Here is the original schematic:




Here is one with my modifications, including Zeners across the heater string:



Thanks for the suggestions!

 

[BrassTeacher]

Well, if it was a solid-state amp and I wasn't sure of the bias point, I'd run a sine wave through it, start with it at the lowest voltage bias point, watch the sine wave, and slowly raise the bias point until the crossover distortion becomes invisible, then go just a wee bit past it. Let it run for awhile, check it again, and touch the output transistor heat sink to make sure things aren't getting out of hand. In this case though, I would not recommend touching the output tubes

Afterwards, I would run square and sawtooth waves through as well, just to double-check

 

[gadget73]

with the zeners in there its more like fixed bias, so it should be ok set to a lower current level than stock. Some of that was because of the shunt resistor to limit swing across the tube heaters, and some was probably to limit the effect of the bias shift that happens with cathode bias.

 

[grindfix]

It took a lot of fiddling with the bias circuit. I didn't have enough adjustment to go below 40mA, so the bias adjustment tail resistor was eliminated and one of the 11v zeners was replaced with 12v unit. Now I have enough bias voltage to go below 40mA on one pair but the other pair is maxed at 33mA so the head resistor had to be lowered. I now have enough of the bias voltage swing to get all tubes biased to draw 36.5mA each. I think I see some crossover distortion at full power. Trace is not very straight. I think I need to go higher, perhaps above AB1 limits, because zeners don't clamp to the same voltage under full load and my cathodes from 34.5v at idle become 37.6v at full power. That applies more bias and cools tubes off at full power outputs. Is that what is happening? If that is the case, I think I should put 11v zener back and bias tubes at 40mA. This should minimize cathode voltage rise and the effect of the increased bias supply will be lower at the max power output. Does it make sense?

 

[gadget73]

there will be some swing to the voltage since its not truly fixed bias, but the 3 volt shift you get now ought to be a whole lot less than it would be without the diodes. If you want to see how ugly it gets without anything to limit it, just unhook one end of the string and give it the beans. Fair bet you'll see crossover distortion at a much lower level.

and yeah thats whats happening, the K is higher than G1 by a larger amount so the current flow drops. The B+ is likely sagging, and the screen supply is probably sagging even more, both of which also contribute to less current flow. Its all the goodness that comes with tube amps that do not have regulated power supplies or any mechanism to force the voltages to track at a relative % of one another, which is most of them. None of this is ever perfect and if it ends up sacrificing a couple of watts at the very top end frankly I don't think it will really matter. Fair guess its making more power now than it was in stock config anyway.

 

[grindfix]

Fair guess its making more power now than it was in stock config anyway.

I don't think I will ever know unless I stumble upon a set of strong and matched tubes. It's getting close to 25wpc with mismatched tubes I have

 

[gadget73]

Thats probably about all its gonna do, but if you lift the zeners I'd expect the output before it starts getting some sort of obvious distortion to be less than 25.

 

[grindfix]

How do I find the least sacrificial compromise?
I have a wood case for this receiver that from searching online looks to be original. The output tubes are in the middle of the receiver chassis but the metal screen is in the back section of the wood case. I’m a little uncomfortable having a stove inside the wood box, getting it to run cooler would be nice