Switching 40W to 20W w/EL84 amp

[mbratch]

I have an EL84 based amp with 4 x EL84 power tubes. I also have a schemtic for it. The amp is a 40W amp, but the schematic shows an option for 20W which only involves removing two of the EL84's and changing a resistor value from 51R to 100R.

What I'd like to know is if I can just remove the two EL84s and run at 20W without changing the 51R, and would that hurt the amp, or seriously shorten the life of the tubes.

I have the portion of the schematic in question shown at www.codekinesis.com/files/el84.tif which shows half of the power amp section (the other half is symmetrical to what's pictured, and the 51R applies to all 4 power tubes).

(I don't know if I'm showing enough of the schematic to evaluate, so let me know if you need more info.)

I can modify the amp to switch from 51R to 100R no problem, but just wondering if I can "cheat" and avoid the mod.

 

[John Phillips]

No.

That resistor develops the bias voltage from the current passing through it, and it needs the current from all four tubes to produce the right voltage. If you run only two tubes, the bias voltage will be far lower and the tubes will run extremely hot and have a very short life, even if they don't blow very quickly.

If you wanted to give the option without pulling the chassis more than once, you could either fit a switch, or use two separate 100-ohm resistors and two cathode bypass caps, one for each pair of power tubes (inner and outer, not left and right), so you could pull one pair without having to take the amp apart.

 

[mbratch]

Originally posted by John Phillips
No.

That resistor develops the bias voltage from the current passing through it, and it needs the current from all four tubes to produce the right voltage. If you run only two tubes, the bias voltage will be far lower and the tubes will run extremely hot and have a very short life, even if they don't blow very quickly.

If you wanted to give the option without pulling the chassis more than once, you could either fit a switch, or use two separate 100-ohm resistors and two cathode bypass caps, one for each pair of power tubes (inner and outer, not left and right), so you could pull one pair without having to take the amp apart.

Thanks very much John. Leads me to a couple more questions:

Do the two 100-ohm resistors need to be tight tolerance so the match as closely as possible, or can they be "sloppy" (10%).

Also, suppose I wanted to be able to do the whole thing with a multi-pole switch. Is there an easy way to "switch off" two of the EL84's without having to remove them (assuming I make the change to separate the resistors and cathode bypass caps)?

 

[Blue Strat]

Originally posted by mbratch
Thanks very much John. Leads me to a couple more questions:

Do the two 100-ohm resistors need to be tight tolerance so the match as closely as possible, or can they be "sloppy" (10%).

Also, suppose I wanted to be able to do the whole thing with a multi-pole switch. Is there an easy way to "switch off" two of the EL84's without having to remove them (assuming I make the change to separate the resistors and cathode bypass caps)?

5% would be fine.

To disable 2 of the tubes, you can disconnect their connection to the cathode resistors with a double pole switch.

 

[John Phillips]

Mike - would it be better, or worse, or turn off the filaments on the unwanted tubes?

In other words, is running tubes for very long periods with no plate current (ie causing cathode contamination) a bigger issue than cold-stripping (if the tube is truly cold, ie off, does that even occur?)...

I really don't know.

My guess is that turning off the filaments may be better.

?

 

[Wakarusa]

I've always seen it done by lifting the cathodes. Don't know for sure about the worse of the evils of contamination or cold stripping, but I would be concerned that the filament-off tube might be in a warm enough environment where you'd get some (but inconsistent) emission at the cathode.. which would probably sound nasty.

 

[Blue Strat]

Originally posted by John Phillips
Mike - would it be better, or worse, or turn off the filaments on the unwanted tubes?

In other words, is running tubes for very long periods with no plate current (ie causing cathode contamination) a bigger issue than cold-stripping (if the tube is truly cold, ie off, does that even occur?)...

I really don't know.

My guess is that turning off the filaments may be better.

?

Good question John. As far as I know, this "Cathode stripping" thing is a theory. As with a lot of this stuff it would take about 50 tubes and 1000s of hours to prove this with no one to pay for the research. Would anyone have paid for the time/effort even back in the tube glory days? I have no idea.

 

[mbratch]

OK, did the rewiring job. Not too tricky. Added a parallel cap/resistor cathode circuit to the two outer EL84's, doubling the values of the (now) two cathode circuit resistors. So if I want it to be a 20W amp, I remove the two outer EL84's.

Works like a champ! (Well... not exactly a champ... it's an EL84 amp after all. )

At some point I may add a switch if I find myself swapping too much.

 

[mbratch]

Hmm... I was just reading somewhere that when I disconnect two of the four power tubes that this cuts the required speaker impedance in half, and that I would have to plug the 8ohm speaker into a 4 ohm connector on the back of the amp (if I had one) since the rating of all the impedance outputs would double (compared to what they are labeled). Is that true?

 

[John Phillips]

Yes, ideally. But if you're looking to intentionally reduce power, the mismatch will help, although it will also change the tone too. It's slightly harder on the remaining pair of tubes though.

 

[mbratch]

Thanks for the reply John. So it won't be a problem for the OT?