Weird problem or crappy PT?

[nickb]

Oh! Hey! I just reached 10,000 posts at 8:40am on #40. Can I get a WOOT! WOOT!

..and you're STILL talking! Sheesh!

Seriously... a big WOOT from me. I always enjoy reading what you have to say.

[Leo_Gnardo]

Oh! Hey! I just reached 10,000 posts at 8:40am on #40. Can I get a WOOT! WOOT!

How's about a HARRRRUUUUUMPH! Hey can I get a HARRRRUMPH from that guy?

Congrats on your 10,000th. Keep the fingers flying!

[tedmich]

Congrats Chuck,

But since RG and Leo appear in every one of your posts, they have more posts by the commutative property I believe...

[Enzo]

10,000 posts? Hey, great, kid... DOn't get cocky.

OK, so I am unclear, did we or did we not sub in a different OT?

[Chuck H]

We did sub another OT. I have a few around. I used one half primary on a PP 20W unit that just happened to be the nearest hunk of iron. This presented a 3300 ohm primary and that did improve output a bit. The distortion increasing with frequency and the extreme screen dissipation remained. BUT...

I need to confess that there is some dirt in my testing. Testing started with a 1k 70mV input signal. Like this the amp makes about 5W. When sweeping and watching the distortion increase with frequency I would then reduce volume until the distortion was gone at 5k then return the input frequency to 1k and THEN test the amps performance. My assumption being that for the amp to be clean at all audio frequencies THIS is as loud as it goes. Well... I decided to measure for max clean watts without returning the input signal to 1k. It turns out that with the same test, but leaving the frequency at 5k the amps output is about 9W. So, basically, with the amp volume adjusted to make 9W at 5k it is only making 2.5W at 1k. I was erroneously taking the lower number to be the maximum output across the sweep. But this is now looking like pretty normal behavior. Why the amp isn't as loud as it should be remains a mystery.

I also experimented with a hotter bias and that did improve the plate and screen current behaviors a little. Reading about this sort of circuit on Merlin's site the potential for over dissipating the screens badly with clipping is discussed. This is apparently also normal behavior. As it happens, tuning a single ended el34 circuit requires a lot more idealizing than a typical AB1 guitar amp. Get one thing wrong and performance is spoiled. Screen current requires managing, bias is critical and if you plan to clip the amp then the screen voltage MUST be managed and kept below the plate voltage by a margin that doesn't allow it to be positive more than momentarily WRT the swing at the plate. Phwew!?! Learning a lot with this one. My usual low tech approach is just biting me on the ass I think.

[tedmich]

10,000 posts? Hey, great, kid... DOn't get cocky.

Enzo's posts per day is ~2.2 times Chucks so barring any finger injury King Enzo (Kingzo?) can never be caught...