The part that's cooking my noodle is the data sheets for the 6L6GC and 6L6GB seem to have the same data: voltage, load, etc. But one's a 30W tube one's a 19W.
You found a key point here. The data sheets
do have the exact same information.
Why? Bottom-line up front: it's just too much work to re-do everything for the new manual.
If you read the front matter of an RCA tube manual, it gives design examples for class A and class AB output stages, but uses a 2A3 triode and an unspecified pentode. For the triode, there are convenient equations to use as a starting point, given a tube and a supply voltage. However, with a pentode, they recommend drawing multiple load lines arbitrarily and calculating the power output and distortion for each. Then you look at which line gave the best results in the end; that also implies different supply voltages and idle points. What a pain!
Say you just wrote an 80-page paper for work or some college class. A reviewer reads it over and says you need to change what's written on pages 56-58. Do you rewrite the whole thing? Hell, no! You makes the changes on those 2 pages and call it good.
Tube manuals came out about every other year. They didn't rewrite all the sheets, or even use an updated tube for the examples in the front (the 2A3 probably was considered a relic in the early 50's). They just updated the relevant specs on the sheets that needed it. But the typical conditions took somebody time and thought to come up with a set of optimum circumstances to show off the tube. I believe they didn't take time to rewrite those, even though the newer version's capabilities were much greater.
So what's the deal with 70%? I feel the issue at this point is that you're trying to reconcile differences among a lot of sources of information. They're all correct, right?
Maybe.
Ponder the
1937 6L6 data sheet.
Look carefully at the conditions given for class AB1. The zero-signal plate currents are for 2 tubes, so divide them in half. 4 conditions are given:
1. Self-bias, 400v plate, 48mA per tube: 19.2w idle, 100% dissipation
2. Self-bias, 400v plate, 56mA per tube: 22.4w idle, 117% dissipation
3. Fixed-bias, 400v plate, 44mA per tube: 17.6w idle, 92.6% dissipation
4. Fixed-bias, 400v plate, 51mA per tube: 20.4w idle, 107% dissipation
Way over 70%, but screen voltage is also less than plate voltage in each case.
We are also not talking about over-volted guitar amps. The later trend became to raise the plate voltage for more output power, and have screens near plate voltage for ease of power supply design, but that means you have to increase the bias to keep the tubes from melting at idle and/or during the signal cycle.
Where did "35mA" or 70% come from for guitar amps? Gerald Weber wrote in one later book that people always asked him for a specific idle current to bias their tubes (note, this is also after he had a different book out saying we needed to "blackface" every amp, to have an adjustable bias pot instead of a bias balance pot). He said he picked 35mA because it was likely low enough not to cause problems in the majority of amps.
I don't remember exactly when the "70% rule" became popular. I notice that if you assume a 19w 6L6 and a 24w 5881, then with 420v plate and 35mA idle, that corresponds to 77.4% for 6L6 and 61.3% for (original) 5881. I think someone got smart and said "halfway is the place to be." But PRR's explanation is probably most-correct, since it is an interesting coincidence that the square-root of 2 is .7071, and you could turn that into talking about working at 70.71%.
Rambling aside, there are too many variables at play to give such a simple rule as 70%. The idea is to be able to give a safe recommendation for those folks who demand a simple answer. "Bias it to ...."
Topic: 6L6/6L6GC/6L6WXT (Read 22879 times)