5E3 only puts out a couple of watts

[Jon Anderson]

I just recently finished building a 5E3 from scratch from some parts salvaged from various places. Besides a few minor tweaks in the preamp, I also have the power supply running at a much higher voltage. The 6V6 plates are running at ~415, the screens at about ~380, and the preamp at about ~270. I've got the power amp biased to about 30mA a piece for the two tubes. (cathode resistor is 500 ohms, with a cathode voltage of about ~30V). The OT is a 10k:8ohm.

Despite all this, when I measure output with my scope and a 10 ohm dummy load, I'm only getting about 1-2 watts clean. Heck, even after distortion hits and I crank everything to ten, I'm still only getting about 1-2 watts. (Point to point on the scope is about 12V across the dummy load. Converting that to RMS = (12/2)*.7 = 4.2; 4.2*4.2/10 = 1.7 watts.

At first, since I was using a salvaged cap can for the PS caps, I thought maybe I had a bad cap for the preamp/screen, which was causing it to run like a triode strapped pentode. But I've replaced and rewired the hole PS, and still am getting the low power.

I just tried wiring 1 ohm screen grid stoppers on each screen so I could look at the screen current. The current is basically non existant. Something like 1mA per screen. Does that sound correct?

Anybody have any other ideas? I've checked the preamp, and the output from the cathodyne is plenty to drive the 6V6s. In fact, watching it on the scope, the 6V6s are starting to clip before the cathodyne is.


Jon

[dobsont]

I don't know much about reading with a scope or what effect the (static) dummy load might have, but your numbers look ok, if just a little cold.
380v*(60ma-2ma) = 22 watts dissipated between the tubes so 11 apiece. Certainly reasonable. I'm running my salvage jobbie at ~73ma, 17v across 220ohm cathode resistor (-4ma best guess at screen), ~335 on the plates. for ~12w dissipation per tube.

I think you can take them to 14w anyway, so you might want to drop the cathode resistor value back a little and see where that gets you.

No doubt though I'm telling you absolutely nothing you don't already know. Have you connected it to a speaker?

[Jon Anderson]

I think you can take them to 14w anyway, so you might want to drop the cathode resistor value back a little and see where that gets you.

I had it biased hotter for a while, but I had to connect a bunch of 100 and 50 ohm resistors and series to get it where I wanted. (about 400ohms). I decided to just swap in a couple of 1k in parallel until I can everything situated.

Have you connected it to a speaker?

Yep, and it absolutely kills. I'm just paranoid something is going to go up in smoke since I don't know why its low power. If I can figure out why its low power, then I should be able to determine if its safe to keep it that way.

Jon

[dobsont]

Weirdness. So it actually sounds like its only putting out 2 watts in the real world? With those kind of dissipation numbers it ought to be pretty dang loud. When I had mine running all cold -- like 8 watt per side it was still loud. Sorry I can't be more help.

[pdf64]

Are your 6V6 known good tubes and a reasonable match?
Have you tried measuring the output with a known good voltmeter (there are a lot of things to twiddle on scopes)?
A 500 ohm cathode resistor will greatly limit the current through the 6v6s, and the bias voltage will go very high with a signal, causing cutoff distortion (2 sides of the same coin). How hot do the 6V6s run with 330ohms?
You are dropping a lot of voltage between the anode and screen PS nodes - what value is the resistor? How much pre amp current are you pulling?
Peter

[Jon Anderson]

Are your 6V6 known good tubes and a reasonable match?

Yep, and I've tried a couple of different pairs. I've current got some JJ 6V6's in there, and they are running at 29mA and 25mA a piece. (Measured by shunting across the ot)

Have you tried measuring the output with a known good voltmeter (there are a lot of things to twiddle on scopes)?

I'll try that. I guess that would be a good way to double check everything. Do be honest, though, I'm pretty sure the measurement is right. It's pretty close in volume to my 5 watt SE amp, maybe a little quiter. It is no where near my 6V6 Plexi style amp , (which is actually running with a lower B+, albeit fixed bias)

A 500 ohm cathode resistor will greatly limit the current through the 6v6s, and the bias voltage will go very high with a signal, causing cutoff distortion (2 sides of the same coin). How hot do the 6V6s run with 330ohms?

I don't recall my actual measurements, but I stepped the cathode resistor up in steps of 100 or 50 at first. I was up to about 350/400 before dissipation was less than 14 watts a tube, which I felt was pushing it a little too much.

You are dropping a lot of voltage between the anode and screen PS nodes - what value is the resistor? How much pre amp current are you pulling?

It's the typical 5k/22k PS resistors. I've also played with 6K/33k. The voltage drop is in line with what I've been able to dig up. Well,those listed here anyway (http://www.kilback.net/homebrewtweak...ematic_r3v.gif)

The preamp is pulling about 3 mA between the 12ay7 and 12ax7, and the screens are pulling about 2ma total.

Jon

[Bruce / Mission Amps]

Assuming you don't have a signal generator, turn it up to about 1/3 and have someone measure the AC signal voltage on the grids of the power tubes.
My guess is you have a signal loss before the power tubes... this could be caused by something as simple as using a pair 22K resistors on the grids loads instead of 220K... etc.

[stokes]

Despite all this, when I measure output with my scope and a 10 ohm dummy load, I'm only getting about 1-2 watts clean. Heck, even after distortion hits and I crank everything to ten, I'm still only getting about 1-2 watts. (Point to point on the scope is about 12V across the dummy load. Converting that to RMS = (12/2)*.7 = 4.2; 4.2*4.2/10 = 1.7 watts.

Without having read all the responses,I am going to say your math is all fouled up.Your scope should be reading RMS voltage,no?Even if it isnt to get RMS you would multiply 12X .7 you get 8.4 now 8.4 squared is 70.5 divided by your 10ohms is 7 watts.Try reading the ac volts across your 10ohm resistor with your DMM and see what you get.Whatever you do to get RMS volts dont divide by 2,just multiply by .7.

[Jon Anderson]

Without having read all the responses,I am going to say your math is all fouled up.Your scope should be reading RMS voltage,no?Even if it isnt to get RMS you would multiply 12X .7 you get 8.4 now 8.4 squared is 70.5 divided by your 10ohms is 7 watts.Try reading the ac volts across your 10ohm resistor with your DMM and see what you get.Whatever you do to get RMS volts dont divide by 2,just multiply by .7.

I measured with an RMS volt meter, rather than my scope, and came up with 10 vRMS. I remeasured by dummy load, and found it was actually about 8.3, rather than 10. (Stupid 20% tolerance. )

So, 10*10/8.3 = ~ 12 watts. Sounds about perfect. So I guess my math was fouled up, but maybe my ears as well. But maybe that's one for those psychoacoustic effects. Or else a inefficient speaker.

So, one more question. RMS is supposed to be 0 to peak, not peak to peak. When I see a signal on my scope across the dummy load, its peak to peak. So why don't I have to divide by two?

Thanks!

Jon

[pdf64]

A 10V rms sine wave will have a peak voltage of 14.14V and p-p of 28.28V. Maybe your scope y amplifiers need re aligning (or have you inadvertently left an attentuation pot turned down)? Try measuring some known voltages as a sanity check ie 9V battery - Peter

[Chris - CMW amps]

Did you connect the screengrids-supply to pins 4, if connected to pins 6 the power is very low...?