How does bias and +B voltage affect the output power?

[Alan0354]

There are really three questions. This is mainly on my Pro Reverb.

1) I put in an adjustable bias and I like the sound when I lower the grid bias from -47V to -44V or even to -42V. This I run the tube a lot hotter. How does this affect the output power. That is originally it's a 40W amp, my guess is the power is lower a little by running the amp hotter.

2) I am putting in a Zener circuit and a bypass switch so I can either run the amp like original or lower the +B by 40V from 440 to 400V. At the same time, I have the grid bias at -44V when switched to +B=440V, then lower to -42V when the +B is switched to 400V. If I use 6L6GC alone, How much have I decrease the output power from 40W when I lower the voltage?

3) I mainly do the switchable lowering the voltage so I can change to 6V6GT. Will I be lowering the output power to 20W or so? The reason I asked is because even I lower the +B to Deluxe spec and use 6V6GT, but the power and output transformer are still the Pro Reverb and be able to provide the power and output power. Will changing the +B and changing the power tubes lower the output down to 20W? The advantage of this is I lower the power by creating an impedance mismatch changing to 6V6 without changing the OT and speakers configuration that is designed for 6L6GC.

thanks

[soundguruman]

There are really three questions. This is mainly on my Pro Reverb.

1) I put in an adjustable bias and I like the sound when I lower the grid bias from -47V to -44V or even to -42V. This I run the tube a lot hotter. How does this affect the output power. That is originally it's a 40W amp, my guess is the power is lower a little by running the amp hotter.

It does not affect the output power.
Cooler bias= more distortion - longer tube life - poorer sound quality
Hotter bias = less distortion - shorter tube life - better sound quality

2) I am putting in a Zener circuit and a bypass switch so I can either run the amp like original or lower the +B by 40V from 440 to 400V. At the same time, I have the grid bias at -44V when switched to +B=440V, then lower to -42V when the +B is switched to 400V. If I use 6L6GC alone, How much have I decrease the output power from 40W when I lower the voltage?

About 10% - really UN-noticeable to human ears. All you have really done is decrease the headroom.

3) I mainly do the switchable lowering the voltage so I can change to 6V6GT. Will I be lowering the output power to 20W or so? The reason I asked is because even I lower the +B to Deluxe spec and use 6V6GT, but the power and output transformer are still the Pro Reverb and be able to provide the power and output power. Will changing the +B and changing the power tubes lower the output down to 20W? The advantage of this is I lower the power by creating an impedance mismatch changing to 6V6 without changing the OT and speakers configuration that is designed for 6L6GC.

Yes, a less efficient tube will give you less output power, on a lower voltage. But usually not a big difference.
But, it's not terribly noticeable to human ears.

thanks

The "best" methods are:

A. change the output configuration from pentode to triode. Easy to do, and lowers the output power without substantially changing the sound of the amp.
Changes (typically)
60 watt amp to 13- 18 watts.
100 watt amp (typically) to 34 watts, etc...
You can crank up the amp, with less loudness. Problem solved.

B. Power Attenuator.
Crank up the amp and reduce the output power.
Use a GOOD attenuator, that matches the amp and the speaker impedance.
*you CANNOT plug "any" attenuator into "any" amp and speaker. The impedance MUST be carefully matched. Amplifier CAN be damaged by a mismatch!!!

C. Combine both methods above.

D. USE a less efficient speaker - lowers the loudness substantially.

Took a Fender Supersonic 60, and changed it to 13 watts, by making the output triode.
This is the BEST sounding Supersonic, I have ever tried, period.
Perfect CLUB LEVEL volume, the customer LOVES it.

Took a Marshall Superlead, and changed it to 34 Watts, triode.
combines it with a power attenuator.
WAY better "Club Level" loudness...
The customer LOVES it.

What I have found, what the customers have found:
AMPS are too loud. They are overkill for most situations. Lower power is WAY better...

Unless you are playing in stadiums. And 99.9% of you are NOT playing stadiums.

Lower power amp is better, for 99% of all guitar players.
Concentrate on Tone Quality, NOT loudness and Power.

[Alan0354]

Thanks for the info, I found the schematic of Supersonic 60

http://support.fender.com/schematics...atic_Rev-A.pdf

But I don't see triode mode. Do I just tide the screen to the plate? I looked up Marshall JCM900 100W Dual Reverb, it just has a switch to tide the screen grid to plate of each tube. Is this safe?

Thanks

[R.G.]

1) I put in an adjustable bias and I like the sound when I lower the grid bias from -47V to -44V or even to -42V. This I run the tube a lot hotter. How does this affect the output power. That is originally it's a 40W amp, my guess is the power is lower a little by running the amp hotter.

Your guess is correct, although the difference is small.

For a given tube, there is a maximum voltage it can hold off without arcing over, and a minimum voltage it can pull down to at its maximum current. The difference between these two voltages is the maximum voltage change the tube can possibly do. The maximum current point is all the current it can pull. Long ago, someone realized that these points define a "load line" of the maximum power the tube can control, based on the limits of the tube itself. Using a lower maximum current point means you get less power because the tube isn't pulling max current; this corresponds to a plate load that's too high an impedance. Using a lower voltage likewise produces less power out for that tube.

There are three limits for a power tube - well, any active electronic device, but we're talking about tubes - the maximum voltage it can withstand without breaking down, the maximum current it can conduct without damage, and the maximum power it can dissipate without melting.

A single-ended tube amp can only use the min-to-max load line for making power, and for audio to avoid gross distortion, this must be class A; the single tube must be conducting at least a little bit, all the time. Class A requires that the idle current be set halfway between full current and zero current to avoid distortion, so the power generated in the tube is 1/2 the max current it can do times the full power supply voltage. This setup has the power dissipation as a limit, because you need to use max current, and you reach the max power product of current times power supply voltage before you reach the voltage maximum for the tube. Class A single tube is dissipation limited.

Push-pull allows you to use two tubes to make power. Originally, this was done in class A, and it effectively doubled the Class A power for a given tube. If you think about it, you can use the same power supply voltage as for single-tube class A, and get twice the power out, because two tubes can get rid of twice the idle-current heat as one tube.

Someone figured out that if you didn't make both tubes conduct all the time, you'd be able to get even more power. Class B was born. The idea was that one tube pulled the positive going side of the signal, and the other tube pulled the negative-going side. For the same power supply voltage and tube maximum voltage, you got even more power. Each tube only conducts a half-wave rectified version of the signal, so the idle dissipation can be nearly zero, and tube heating only depends on how big the signal is.

The effect of going all the way to Class B is that you can use a much, much higher power supply voltage because you're no longer burning up the tubes with half-maximum current all the time.

But the audio quality was ugly, because the tubes did not hand off the audio signal well in the middle, at zero voltage - it had cross-over distortion. Enter Class AB. Class AB turns both tubes on a little bit at idle. This smooths out the ugliness at crossover. However, the amount of overlap increases idle distortion.

All of that blather is preparatory to saying that there is a continuous range of bias conditions from half-maximum-current for class A to zero idle current for true class B. In between are various degrees of Class AB. You must use a smaller power supply voltage for Class A to keep from burning your tubes out. As you accept smaller idle currents by biasing the tubes' idle currents ever more toward zero, you can increase the B+ voltage and get more power out without burning out your tubes from idle current heating.

And now we come to your actual question. For the same B+, what does increasing the tube idle current do?

You are correct - increasing the conduction overlap does reduce the amount of signal power that gets out as output power, but it is really a small reduction. It also increases the idle-conditions power tube heating, perhaps a lot. Trivial power reduction, maybe significant heating problems.

2) I am putting in a Zener circuit and a bypass switch so I can either run the amp like original or lower the +B by 40V from 440 to 400V. At the same time, I have the grid bias at -44V when switched to +B=440V, then lower to -42V when the +B is switched to 400V. If I use 6L6GC alone, How much have I decrease the output power from 40W when I lower the voltage?

Takes a lot of math to work that out, but probably a trivial amount, possibly not easily audible. You do reduce the heating in the power tubes; this power now comes out in the zener.

3) I mainly do the switchable lowering the voltage so I can change to 6V6GT. Will I be lowering the output power to 20W or so? The reason I asked is because even I lower the +B to Deluxe spec and use 6V6GT, but the power and output transformer are still the Pro Reverb and be able to provide the power and output power. Will changing the +B and changing the power tubes lower the output down to 20W? The advantage of this is I lower the power by creating an impedance mismatch changing to 6V6 without changing the OT and speakers configuration that is designed for 6L6GC.

Again, probably a lot of math to figure out how much lower. It's simpler to measure it on a dummy load. It's not going to be a huge reduction in audible loudness, though. You generally need a ten-to-one reduction in amplifier power to get a perception of "half as loud". This is a wobbly rule-of-thumb based on human hearing research, and varies with frequency, etc., but is a useful across-the-board generality.

[pdf64]

'triode mode. Do I just tide the screen to the plate?'
Yes, preferably via a screen grid resistor.
My experience is that triode mode doesn't sound good, rather flat and blah.
Pete

[Alan0354]

'triode mode. Do I just tide the screen to the plate?'
Yes, preferably via a screen grid resistor.
My experience is that triode mode doesn't sound good, rather flat and blah.
Pete

That's what I think too, my JCM900 has that feature, I don't like it.

[J M Fahey]

Triodes have no teeth in a guitar amp.

[frus]

agree, horrible

[Chuck H]

Let's keep in mind that a substantial part of the power loss in triode mode (WRT guitar amps) is that the new operating conditions aren't compatible with the OT primary impedance by a long shot. Now, I don't know that matching the load to triode operated tubes would make for an improvement in the tone, but I've never tried it. So this may be worth experimenting with. Of course then you wouldn't get the same level of power reduction either. So maybe a moot point. With every one bashing the tone of triode operated pentodes this is part of the picture so I mentioned it. I too have found that a triode operated guitar amp with no other modifications sounds dull.

Nothing is ever going to be a panacea. I'll second the suggestion of a good attenuator and agree that it's not going to be perfect. But I think it's the least bad solution.

[Alan0354]

I think changing to 6V6 either using JJ 6V6S without changing the +B leaving at 440V but lower the bias a little. OR lowering to 400V and bias to -36V might be the answer. You guys be the judge in the video:

This is a comparison between original SF to 6V6 at 440V and bias at -46V. Also, I set to 400V with bias at -40V and -36V. And finally I increase the treble from 5 to 6.5. The volume is at 5.5 to 6 and never been touched. Treble and bass at 5 except the last one. Bright switch off.

1) 6L6, 440V, -46V. Slightly cracking sound because the mic of the camera is distorted.

2) 6V6, 440V, -46V.

3) 6V6, 400V,- 42V.

4) 6V6, 400V, -36V.

5) 6V6, 400V, -36V, treble at 6.5.

6V6 440V - YouTube

I swear, the pair of 6V6GT have been sitting in the shed for 17 years new. When I first powered it up, they sounded really bad. But as time goes on, they seemed to sound better after burning in. Someone mentioned that it takes some burning in if a tube has been sitting there for years. This is a pair of Ruby 6V6, but I don't care whether I stress them. I just bought a pair of JJ 6V6S on ebay as they are designed for up to 500V.

[Alan0354]

Let's keep in mind that a substantial part of the power loss in triode mode (WRT guitar amps) is that the new operating conditions aren't compatible with the OT primary impedance by a long shot. Now, I don't know that matching the load to triode operated tubes would make for an improvement in the tone, but I've never tried it. So this may be worth experimenting with. Of course then you wouldn't get the same level of power reduction either. So maybe a moot point. With every one bashing the tone of triode operated pentodes this is part of the picture so I mentioned it. I too have found that a triode operated guitar amp with no other modifications sounds dull.

Nothing is ever going to be a panacea. I'll second the suggestion of a good attenuator and agree that it's not going to be perfect. But I think it's the least bad solution.

My whole purpose with this amp is to keep it as original as possible, any mode has to be able to switch out back to original. I buy a good quality amp to use as a reference so I can design my own and compare with a standard reference. So I try very hard not to get fancy with this one. Please listen to the video and comment on this. I think the ticket of preserving the original sound and lower the power is using the 6V6. I tried attenuator and substituting one speaker with resistor, they do change the sound. Particular the attenuator, it change the sound more even at only -4dB setting.

[Bruce / Mission Amps]

...
I swear, the pair of 6V6GT have been sitting in the shed for 17 years new. When I first powered it up, they sounded really bad. But as time goes on, they seemed to sound better after burning in. Someone mentioned that it takes some burning in if a tube has been sitting there for years. This is a pair of Ruby 6V6, but I don't care whether I stress them. I just bought a pair of JJ 6V6S on ebay as they are designed for up to 500V.

During that period, I think Ruby was relabeling inexpensive Chinese 6V6s but they were working with them to come up with a slightly better design that could handle +420vdc.
As to needing to be broken in after sitting for 17 years sitting in a barn....or what ever you were told, ... I doubt it.
It's probably more like they sounded poor-bad the day they were built and heating them up now has changed some unintended characteristic of the tube over your current time in use.
Mass produced electromechanical devices (especial early Chinese tubes) have wide ranges of suitability.

[Alan0354]

During that period, I think Ruby was relabeling inexpensive Chinese 6V6s but they were working with them to come up with a slightly better design that could handle +420vdc.
As to needing to be broken in after sitting for 17 years sitting in a barn....or what ever you were told, ... I doubt it.
It's probably more like they sounded poor-bad the day they were built and heating them up now has changed some unintended characteristic of the tube over your current time in use.
Mass produced electromechanical devices (especial early Chinese tubes) have wide ranges of suitability.

Thanks

[kg]

ime triode output stages really sound best with zero loop nfb.

[Alan0354]

I want to double check. I did the measurement using 6V6GT with 400V. At a 8 ohm load using the THD hotplate as dummy load and put 1KHz. I see the sine wave started to clip at 36V peak to peak. So the peak to peak current is 4.5A peak to peak.

V rms =(36/2)X0.707V=12.726Vrms.

I rms =(4.5/2)X0.707=1.59Arms.

So power =IV ( in rms)=12.726X1.59=20.24W.

Is this the right way to find output power?