An interesting graph

[Kap'n]

Thought some of you tube geeks might find this one interesting. It's from the Sylvania Technical Manual (c) 1955, and applies to 6L6, 6L6G and 6L6GA tubes. It's the only place I've seen this data.

Want to design the output distortion of your amp?



Funny, how folks talk about the even order harmonic distortion of an output section, when between the load (4K in most cases) and the push-pull, second harmonic distortion is minimized, and third and fourth harmonics are much more prevalent.

And, no reducing the load on an output section of a tube amp does NOT increase your power output.

[Kap'n]

More data from the same manual

[Wilko]

What is their definition of the harmonics?

And the wattage ratings are very low for that kind of distortion.

My McIntosh amp claims 40 watts for a pair of 6L6s with max .5% THD.

[JAM]

I looked at the power/distortion graph from the RCA manual when I made the decision to use the 2 ohm tap (with the 3x10s) in my 5F6-A clone. The RCA graph doesnt go below 3K, but goes higher than the Sylvania shows. It only shows the total distortion, it's not broken out 2nd, 3rd etc., but it's for a 6L6GC with more typical voltages found in guitar amps. (450V plate)

link

[Kap'n]

What is their definition of the harmonics?

And the wattage ratings are very low for that kind of distortion.

My McIntosh amp claims 40 watts for a pair of 6L6s with max .5% THD.

I suspect the measurements are made with a 1kcps (this is the 50's y'know ;-) ) signal, generally considered a 'standard test signal.'

Wattage ratings, and all other measurements, I'm sure are for single ended applications.

I suspect that your Mac is spec'd around 6L6GB or GC tubes, with 23W or 30W plate dissipations, respectively. Back in '55, the 6L6-type tubes were still 19W jobs.

FWIW, with those ratings.....

A single-ended 6L6(G, GA) at book voltages of 360V on the plates, 250V on the screens, and a fixed (!) bias voltage of -18V gave an output of 10.8W, with a distortion of 15%. (on tube data sheets, drive voltages are spec'd as having a peak of the bias voltage)

A push-pull pair with 360V / 270V / -22.5V gives 47 watts at 2% distortion.

I'm certain that these numbers are obtained with standard mid- to low-fi transformer-coupled outputs of the day, given the target audiences of these spec sheets. Mac, Leak, or any of the other true hi-fi companies of the period were able to obtain better results.

[Kap'n]

I looked at the power/distortion graph from the RCA manual when I made the decision to use the 2 ohm tap (with the 3x10s) in my 5F6-A clone. The RCA graph doesnt go below 3K, but goes higher than the Sylvania shows. It only shows the total distortion, it's not broken out 2nd, 3rd etc., but it's for a 6L6GC with more typical voltages found in guitar amps. (450V plate)

link

That's a very interesting curve, too.

On the curve I posted, power output maxes out with a load almost exactly where THD bottoms out, whereas on your curve, the load is slightly higher for max power output vs. THD.

Is this due to push-pull, differences in tube construction between a "GA" and a "GC", voltages on the tubes, or a combination of some/all of these?