Premier Rock Forum

rocco wrote:You're right. I assumed that if they are sold as matched then they should run usually within 5mA of each tube. But if you bought a smaller amount, I'm sure the guy would have spent more time finding closer pairs.

Well, at the 280V that this amp operates at, in the most extreme example within the 5ma range, if the tubes were 35ma and 35ma on the push half, and 40ma and 40ma on the pull half, that would work out to 19.6W on the push, and 22.4W on the pull. You'd be better off with one 35ma and one 40ma on each half, with an even 21W on each side. No?

I've read stuff that would suggest that at the very least, an imbalance on the output transformer really isn't good for the bass response. And can't it be stressful on the OT in general?

Hell, maybe I just like geeking out and working with my amps a little too much. Honestly, I can't wait for the day when I know more than just enough to be dangerous.

scott wrote:Well, at the 280V that this amp operates at, in the most extreme example within the 5ma range, if the tubes were 35ma and 35ma on the push half, and 40ma and 40ma on the pull half, that would work out to 19.6W on the push, and 22.4W on the pull. You'd be better off with one 35ma and one 40ma on each half, with an even 21W on each side. No?

Yeah, you should be making sure that the current draw on each side of the output transformer is balanced rather than two tubes on the same side being balanced. If its uneven like your scenario you'll produce crossover distortion which is probably undesirable but could be slightly harmonically interesting. I was imagining a scenario where the purchased matched quad is pulling 29mA, 31mA, 32mA, 34mA. So you'd obviously put the 29 and 34 on one half and the the 31 and 32 on the other and not buy 4 more tubes.

I don't know much about el84's but aren't you biasing them too hot? Most people say to bias between 50 and 70 percent of max plate dissipation and for an el84 its 12watts. If you've got it at 21watts a side, then its 10.5 per tube which is close to 90%. Can they be run like this or do they burn out quickly?

rocco wrote:I don't know much about el84's but aren't you biasing them too hot? Most people say to bias between 50 and 70 percent of max plate dissipation and for an el84 its 12watts. If you've got it at 21watts a side, then its 10.5 per tube which is close to 90%. Can they be run like this or do they burn out quickly?

It's "Class-A" (like the AC30 supposedly is) so the tubes idle at full, right? It's cathode-biased, so I'm not biasing them myself. There's no internal adjustment either, short of actually changing the values of stock resistors.

The tubes I took out of there are unlabeled and it wouldn't surprise me if they're the original tubes from when the amp was made (I found out from the factory checklist inside the chassis it was made on my birthday in 1995! So neat!). The guy I bought the amp from apparently had no tech and was not himself gear-savvy at all... because there were two issues with the amp that he graciously mentioned in the eBay auction... the "power on" LED was pushed inside the chassis and was barely visible, and one of the input jacks wasn't working.

Once the chassis was taken out (like 7 screws) and the soldering iron warmed up, it took me all of about one minute to fix both problems. So I'm guessing if he couldn't get stuff that simple taken care of, he probably wasn't dropping in new tubes either. Who knows?

Anyway, I only replaced them cause I figured it was probably due... I played it for many months with no problems at all before popping in the new tubes. And so far, they're slightly gassy (just a touch of blue glow in a couple of them) but haven't shown any signs of weirdness after a couple hours of runtime during biasing and another several hours of actual play. The amp sounds great, always has, and I have no complaints so far with the new tubes. If they fry out quick, then I'll know that something's whack.

I wonder also if the 1-ohm resistor in the bias meter (a Weber BiasRite) isn't tweaking out the readings somehow, too. Pretty much all of my tube-tech experience has been with fixed/variable bias amps, not cathode-biased... so my numbers might be messed up in the first place. But as long as they're consistent on the two halves of the OT, I'll sleep easy.

And hey, any info about testing the bias of a cathode-biased 4xEL84 amp would be great. If I'm going about it all wrong, I wanna know!

Oh, okay, that makes sense if its class A since it will draw max current for the full cycle. Everything I was talking about was with respect to push-pull class AB. I'm not even gonna try to wrap my head around an amp like the ac30 that's sort of class A but really class AB. I guess all you have to do is find a good tube compliment and measure it properly. If you think your bias measuring tool is lying you can measure it yourself by directly taking a reading of the plate voltage, cathode voltage, and plate current. Subtract the cathode voltage from the plate voltage and multiply that number by the plate current.

scott wrote:

rocco wrote:You're right. I assumed that if they are sold as matched then they should run usually within 5mA of each tube. But if you bought a smaller amount, I'm sure the guy would have spent more time finding closer pairs.

Well, at the 280V that this amp operates at, in the most extreme example within the 5ma range, if the tubes were 35ma and 35ma on the push half, and 40ma and 40ma on the pull half, that would work out to 19.6W on the push, and 22.4W on the pull. You'd be better off with one 35ma and one 40ma on each half, with an even 21W on each side. No?

I've read stuff that would suggest that at the very least, an imbalance on the output transformer really isn't good for the bass response. And can't it be stressful on the OT in general?

Hell, maybe I just like geeking out and working with my amps a little too much. Honestly, I can't wait for the day when I know more than just enough to be dangerous.

Simply multiplying the current times the voltage wont give you the wattage of the amp, it will give you the wattage of the power dissiption on the plates of the tubes.

While it's a good way to bias an amp, make sure the tubes aren't runnign too hot, etc, a lot of that power on the plate is dispersed as heat, as opposed to actually pushing the speaker.

To match sides of your transformer accurately, you'd neet to send a sine wave through the amp and measure the signal separately on each leg of the transformer. Remember, the phase inverter might not be amplifying evenly, or the values on one of the plate load resistors of the PI might have drifted, caugin uneven level to be sent into the power section.

Tube amps can never be an exact science. They will all eventually wear out. As long as no components are run dangerously, a slight mismatch on the output transformer is fine and pretty much normal. In fact, dod you take out the transformer and hook it up to a voltage source and measure if each side of the OT is electrically identical. It''s pretty much impossible to get it perfect; to get the center tap right in the "center".

But yes, you're tubes weren't matched if every plate was getting the same voltage. I know people who put separate bias adjustments on each half of the output section. Seems crazy to me. I'd just rather get it sounding good, running safe, and then move on to playing or the next project.

Cheers!

Ben Adrian