Tweed Deluxe: max clean output 9watts?

[JJman]

I measured the output of my homebrew just before clipping using a 1k sine wave. I show about 8.5volts RMS on an 8ohm dummy load using my Fluke meter. Also see about 24volts pk-pk on my scope, (which is 12volts pk, which is also about 8.5volts RMS.)

Doing the math:

Watts=VA, A= V/O, Watts=Vsquared/O
Watts=8.5x8.5/8= 9

So I’m seeing only 9 clean watts. I see about 11.5volts RMS (Fluke meter) well into clipping which is 16.5watts doing the math.

Here are my stats at idle:

Plate 344V
Screen 339V
Cathode 21V
Current 38ma ea.
Watts ~12ea.

From what I’ve read, the above idle stats are extremely close to Tweed Deluxe figures. The primary on my OPT is 6.6k. I don’t know what a Tweed Deluxe would be. Using 8ohm secondary into 8ohm dummy. I have verified that the PI does not clip before the output tubes begin to clip, and it’s outputs are nicely balanced. The preamp does not clip until even later on the volume control. Output tubes are JJs and are matched within 1ma. 6v6s are clearly running class AB showing cutoff of less than 50% of the cycle, so no crossover notches.

So is a max of 9 clean watts to be expected for this output setup? Are my 1k screen resistors curtailing the output? I think one is supposed to measure output at 10% THD but I wouldn’t be able to estimate that point. Maybe the “official” output is 10-12watts if the clean is 9watts?

Volume is fully adequate for me and actually meets my desired outcome nicely. I understand that the volume difference between 9watts and say 20watts is minimal. I’m just a little surprised to see only 9watts here. After all “18watt” amps use EL84s. Are they really 18watts RMS?

My SFDR uses higher plate voltage (427), fixed bias and purports to be “22watts,” as a reference. So the homebrew should be measurable less than the SFDR. I should have measured it the other day when I had it on the bench. Next time..

Has anyone actually measured the clean output potential on a Tweed Deluxe type setup?

 

[schmidlin]

Next thing I would do is check the output on a 4 and 16 ohm load. You may be surprised. You also may have an in-between output, like it wants to see 11 ohms.

I thought 2 x 6V6's wanted to see 8K on the primaries. 5E3?

 

[jay42]

You're definitely in the ball park. It's hard to notice a couple extra watts. A different set of 6V6's might push it one way or the other. I doubt there's anything wrong with it.

fwiw, my 1974 clone delivers 13W clean with JJs cathode biased. I actually prefer it fixed bias. I get 15W that way.

 

[RussB]

sounds about right. are you running 12AY7's or 12AX7's?

 

[phsyconoodler]

The original 5e3 had an 8k primary,so you will have less overall output with 6.6k into an 8ohm load.Try running a 4 ohm speaker.
If you like the sound,you may as well leave it alone.Totally up to you.

 

[JJman]

I'll probably try a 4ohm load on the 8ohm tap to see what happens. I'm afraid to go the other way because of the warnings about Mr. Flyback and Eddy on the internets. But I'll stick with the current setup for playing.

I'm using a lower voltage Express type 12ax7 preamp and PI but with a 12at7 in the PI at the moment. Also blonde Tremolux Trem. Only the 6v6's are in Tweed Deluxe configuration although I added NFB recently. With the NFB, clipping is now reached a little later on the volume control but the PI still runs clean while the 6v6's are in their early clipping range. The max clean power output didn’t seem to change with the added NFB.

Thanx for the feedback.