Wattage...

[tubby]

Well, tonight I got bored and measured the output voltage of my single-ended, cathode-biased KT88 amp at the speaker jacks while a friend was playing away full-blast through a 4 ohm cabinet. I read 10.9 V peak which would equal about 30 Watts if I'm correct. Of course the amp is far from any clean sound at that point, but this is still....um, surprising. It sure sounds like 30W, it's just that I thought you can't get more than ~20W from a single KT88 (at least in theory).
I'm sure I'm missing something here. If someone could enlighten me....

Thank you.

[Enzo]

10.9 peak is 7.71 RMS - applying sine wave rules - square that and divide by resistance gets on my calculator 14.8 watts

[tubby]

ahh, i see...this sounds more realistic. i used a true rms meter and therefore thought i could skip the x 0,707 step. what exactly is the advantage of having such a meter then?

[Steve Conner]

Well, if you used a true RMS meter you shouldn't have said "peak" in your original post, you should have said RMS. Enzo probably thought you measured the actual peak voltage with a scope.

If you did use a RMS meter then your calculation was roughly right. Cranked single-ended amps can go all the way into Class-C at which point they're basically choppers delivering a square wave to the speaker. This can amount to a good deal more power than the datasheet figure for the tube, which is for undistorted sine wave output.

[tubby]

Well, if you used a true RMS meter you shouldn't have said "peak" in your original post, you should have said RMS.

well, that was just the highest value i observed (i used the max function of the meter). i guess i should have said maximum.

i'm still confused: i just measured an almost cranked fender bassman export (2x6l6gc) into a 4 ohm cabinet: 18v max with the same true rms meter. this would mean approx. 80 watts. could it be that the true rms meter is overtaxed with the rapidly fluctuating signal or it's complex frequency content and thus can't display a decent rms value?

[Steve Conner]

If it's true RMS, it should be true RMS. Audio power is a complicated subject that marketing guys go to town on. Just two things I can think of:

Amps can deliver more power in short bursts than continuously, because the filter capacitors store energy.

A 4 ohm speaker does not measure 4 ohms at all frequencies. At its bass resonance it can be more like 40. So it's not necessarily sucking a current corresponding to the measured voltage divided by 4.

[Enzo]

When someone says "peak" I think "peak."

[Chuck H]

No means yes, orange is the new black, and peak is the new RMS.

Chuck

[Robert M. Martinelli]

Maybe I'm wrong, and, if so, be patient, but I think we're missing one step in calculating the RMS power.

As Enzo knows ( he kindly posted here the schematic diagram at my request ) I recently repaired an Ampeg BA115HP. When I finished fixing it I bench tested it, and measured its power, using a signal generator, an oscilloscope and a resistive dummy load.

With a 1Khz sine wave, at the conditions specified by the manufacturer, the amp started clipping at 81 Vpp on my Tek 2335 on a 4 Ohm dummy load, and, to get the true RMS value, I had to divide the pp value by 1,4142 first and then divide again by 2 ( our missing step, which can be thought of just like "graphically" taking the lower half of the waveform and moving it on the positive side -output power can't go "negative" ). That yielded 28,63 V(RMS) which, divided by 4 Ohm gave 7,16 A, and this, multiplied by 28,63 V(RMS) gave me 205 W(RMS) which matches almost exactly the approximate formula W=((VRMS)^2)/(Z) ===> (28,63^2)/4=204.91 W(RMS), pretty close to what Ampeg states to be the BA115-HP RMS power ( 220 W ) ( I didn't find matched MOSFETS and this could be the reason why the power was slightly less ).

JM2CW

Cheers

Bob

[Steve Conner]

Maybe I'm wrong, and, if so, be patient, but I think we're missing one step in calculating the RMS power.

the amp started clipping at 81 Vpp

That's peak-to-peak, which is twice the peak value, so yes you need to divide it by 2. It is not the new black or the new RMS

Also an explanation for the lower power might be that your line voltage was lower than whatever Ampeg thought the export model ought to run off. If it's 210-220, that would hurt the power for sure.

[Robert M. Martinelli]

Hi Steve,
I used that phrase to "soften" my way to chime in and express the concept, he was aware something was wrong in his way to calculate the output power, ( remember his doubts about the "80W Bassman" ) but probably couldn't understand what, so I felt I had to find a "gentle" way to express the concept....and that's also why I used the "graphical analogy", to make his understanding easier.....as you rightfully said, nothing new under the sun .

As to the BA115HP, I measured the line voltage before starting to test the amp on the bench, and it was dead on 220VAC ( matching the amp's requirements ) that's why I stated that the "missing" power is probably due to slightly mismatched MOSFETS. ( That amp is a "complementary symmetry" design, so matching matters ).

BTW, how about some vacation? Hope your situation at work is not as bad as it was the last year ( mine's worsened , been working a solid five weeks with some calls at the night too, no wonder my mental clarity is going down the drain ).

Cheers

Bob