Half-power operation

[benoit]

I'm reading through Basic Electricity (the Navy reprint from the 70s available at Barnes & Noble) and just had a thought which I wanted to run past people who know this stuff in their sleep.

Is it true that pulling two tubes provides half-power operation simply by halving the transconductance in the power section (effectively doubling resistance), which by virtue of P = I^2 * R must mean half-power? Voltage in this case would be the AC voltage fed to the PT grids, yes? Since that voltage is controlled by the same volume knob and fed from the input source, is "fixed" in that the same voltage is seen on each grid at any given instant, and assuming each tube is perfectly matched and biased exactly the same (which is fine, right, as long as one realizes these "ideal" conditions don't really exist?), the change of 4 to two should halve power (I suppose also assuming the power-supply to be a perfect blackbox that adjusts itself accordingly to provide the proper voltage).

If you made it all the way through that without the desire to wring the long-sentence-itis out of my neck, I tip my hat to you. So am I thinking about this correctly?

[Firestorm]

Yes. Sort of. But the remaining tubes will now see the load as being one-half as big as before and will attempt to develop their power into the smaller load. Depending on the tubes and the power available, they may or may not be able to do this. So this could result in slightly higher power output, or more likely, lower power output. The bandwidth would also be reduced, and, in this case, most likely shifted to lower frequencies.
Are you trying to accomplish something specific? Or is this an academic question?

[benoit]

Yes. Sort of. But the remaining tubes will now see the load as being one-half as big as before and will attempt to develop their power into the smaller load. Depending on the tubes and the power available, they may or may not be able to do this. So this could result in slightly higher power output, or more likely, lower power output. The bandwidth would also be reduced, and, in this case, most likely shifted to lower frequencies.
Are you trying to accomplish something specific? Or is this an academic question?

Very much an academic question. I realize that what I said simplifies a lot and ignores a lot, I was focusing specifically on power.

[Firestorm]

To get to the simplest answer you have to control all other variables besides the number of tubes. Output tubes develop their highest rated power into a specific load. This varies depending on plate voltage, screen voltage, type of bias used and tube configuration (triode, tetrode, pentode, g2 and g3 connections). If you keep all of the other variables the same (including adjusting the load for the halved tube complement) pulling two tubes will generally result in half power. There's one power transformer wrinkle, however. Pulling two output tubes will halve the current being pulled through the PT, so depending on the build (of the PT) the secondary voltage may rise, increasing potential output power. It's hard to control for that.

[briane]

we should also mention, that halving the power does not mean half as loud.

As I recall its more like 95%, or 70% as loud, or somewhere in there (anyone? I cant recall the correct number). Ussually half power is a 3-6 decibel drop. not all that much difference, eh?

[benoit]

we should also mention, that halving the power does not mean half as loud.

As I recall its more like 95%, or 70% as loud, or somewhere in there (anyone? I cant recall the correct number). Ussually half power is a 3-6 decibel drop. not all that much difference, eh?

Right. Half as loud would be 1/10 power, I believe. As I said before this is purely academic. I have heard plenty of amps with half-power switches, and wouldn't bother with them myself unless someone really wanted one.

[Andy Le Blanc]

the best example is a fender twin
the old trick was to pull half the tubes... one from each side
this creates a mismatch between the load and tubes
which gives an apparent drop in volume
the right way is to half the tubes and double the load
in a twin you would run only one speaker...
nobody did this.......but it wasnt too far out of the park to hurt the amp
the difference between 100 and 50 watts is only 3db
but thats alot to the ear

[Firestorm]

the best example is a fender twin
the old trick was to pull half the tubes... one from each side
this creates a mismatch between the load and tubes
which gives an apparent drop in volume
the right way is to half the tubes and double the load
in a twin you would run only one speaker...
nobody did this.......but it wasnt too far out of the park to hurt the amp
the difference between 100 and 50 watts is only 3db
but thats alot to the ear

I once worked on a Twin for a jazz player who had replaced the two 12"s with a single 8-ohm JBL 15. I put in a switch that would lift the cathodes of two tubes so he could run, properly, into the 8-ohm load, or switch them back on to use with an added extension cabinet. Just for jollies, we toggled the switch back and forth while he was playing into the 8-ohm JBL. ZERO difference in sound or volume; just one of those situations where the mismatch exactly halves the power anyway.
I would submit, though, that while 3dB (going between 50 and 100 watts) might make a difference in trying to be heard over a loud drummer (and even there I bet it's more SPL than watts), if you're listening to the amp alone, 100 watts is just barely louder than 50 watts.

[muchxs]

In my experience a small amp with big transformers "thinks" it's a big amp. What that means is that if you attempt to throttle back a big amp with expedient mods you'll lose very little volume.

[Andy Le Blanc]

I like that solution.... but it raises some questions
the comparison you decribe seems to be between
a fairly well matched load operating a pair of 6l6
VS a mis matched load for a quartet
you might very well get what you have described
ie: no audible change
it is a matter of horse power....... I love 100 watts when I actually need it

[muchxs]

it is a matter of horse power....... I love 100 watts when I actually need it

Actually measure it sometime.

Years and years ago I had one of those cheap-o Radio Shack LED power displays. I plugged all sorts of "really loud" amps into it and I was surprised to find that most of the time I using less than a watt.

One of my associates continuously monitors output power on his fancy hi-fi rig. His power display is switchable, it has a 1 watt setting, a 10 watt setting and a 100 watt setting. 1 watt is enough to shake bric-a-brac off the shelves. I expect the full 100 watts with the meters pinned would result in a smoking hole in the ground where his living room used to be!

[Andy Le Blanc]

I tried to set my bench afire once .....
setting the bias on a twin with a twenty watt dummy load
I was in pinch..... power......heat
on the other side of the coin......
I have little stereo I built SE useing 6v6 triode....
2.5 watts ..... awesome
the apparent loudness and sound stage was more than
enough for the application

which is better...... a smaller amp
or a bigger amp that can be (has to be ) compromised for less watts

[UR12]

the best example is a fender twin
the old trick was to pull half the tubes... one from each side
this creates a mismatch between the load and tubes
which gives an apparent drop in volume
the right way is to half the tubes and double the load
in a twin you would run only one speaker...
nobody did this.......but it wasnt too far out of the park to hurt the amp
the difference between 100 and 50 watts is only 3db
but thats alot to the ear

I once worked on a Twin for a jazz player who had replaced the two 12"s with a single 8-ohm JBL 15. I put in a switch that would lift the cathodes of two tubes so he could run, properly, into the 8-ohm load, or switch them back on to use with an added extension cabinet. Just for jollies, we toggled the switch back and forth while he was playing into the 8-ohm JBL. ZERO difference in sound or volume; just one of those situations where the mismatch exactly halves the power anyway.
I would submit, though, that while 3dB (going between 50 and 100 watts) might make a difference in trying to be heard over a loud drummer (and even there I bet it's more SPL than watts), if you're listening to the amp alone, 100 watts is just barely louder than 50 watts.

I think you guys may be confusing some terms. A 3db change in spl is what the human ear percieves as a doubling of the loudnes of sound. If you increase something 3db it should sound twice as loud. If you drop 3db then the sound sounds 1/2 as loud. This has nothing to do with watts.

In terms of power, (Watts) in order to make something sound twice as loud (or increase it by 3db SPL) you have to increase the power by a factor of 10. In other words if you are playing a 1 watt amp and you want to make it sound twice as loud you need to up the power to ten watts. If you want to make the 10 watts twice as loud then you have to turn it up to 100 watts.. This is why when you drop two tubes in a 100 watt amp it doesn't sound 1/2 as loud. You would have to drop it down to 10 watts for it to sound 1/2 as loud. This is also the reason a 36 watt Express sounds damn near as loud as a 50watt Marshall.

[Firestorm]

I think you guys may be confusing some terms. A 3db change in spl is what the human ear percieves as a doubling of the loudnes of sound. If you increase something 3db it should sound twice as loud. If you drop 3db then the sound sounds 1/2 as loud. This has nothing to do with watts.

In terms of power, (Watts) in order to make something sound twice as loud (or increase it by 3db SPL) you have to increase the power by a factor of 10. In other words if you are playing a 1 watt amp and you want to make it sound twice as loud you need to up the power to ten watts. If you want to make the 10 watts twice as loud then you have to turn it up to 100 watts.. This is why when you drop two tubes in a 100 watt amp it doesn't sound 1/2 as loud. You would have to drop it down to 10 watts for it to sound 1/2 as loud. This is also the reason a 36 watt Express sounds damn near as loud as a 50watt Marshall.

The real confusion arises because a decibel is not a discrete unit of anything; it is a logarithmic relationship between two measurements in the same frame of reference: power or voltage or acoustics or signal strength, etc. For clarity, engineers will typically append suffix letters to clarify what is being measured (and thus what the reference level is), hence dBm, dBv, dBmV, etc. The relationships hold within the frame of reference, but don't translate to other frames. As Dana said, if we are measuring SPL, a 3dB (SPL) change corresponds to a doubling or halving of the SPL; if we are measuring power, then 3dBW corresponds to a doubling or halving of the power in watts. The problem is that a 3dB change in power does not result in a 3dB change in SPL. Maybe we should adopt our own measurement system: +3WML (way more louder) and -3WMQ (way more quieter).

[benoit]

I think you guys may be confusing some terms. A 3db change in spl is what the human ear percieves as a doubling of the loudnes of sound. If you increase something 3db it should sound twice as loud. If you drop 3db then the sound sounds 1/2 as loud. This has nothing to do with watts.

In terms of power, (Watts) in order to make something sound twice as loud (or increase it by 3db SPL) you have to increase the power by a factor of 10. In other words if you are playing a 1 watt amp and you want to make it sound twice as loud you need to up the power to ten watts. If you want to make the 10 watts twice as loud then you have to turn it up to 100 watts.. This is why when you drop two tubes in a 100 watt amp it doesn't sound 1/2 as loud. You would have to drop it down to 10 watts for it to sound 1/2 as loud. This is also the reason a 36 watt Express sounds damn near as loud as a 50watt Marshall.

Respectfully, that's not correct. To get a 3db SPL bump you double the power. This, however, is not, however, going to give you perceived double the loudness.

I quote from the Yamaha Sound Reinforcement Handbook:

"How do we perceive SPL? It turns out that a sound which is 3 dB higher in level than another is barely perceived to be louder; a sound which is 10 dB higher in level is perceived to be about twice as loud (loudness, by the way, is a subjective quantity, and is also greatly influenced by frequency and absolute sound level)."

Try a 3 dB boost or cut to the master of any mixer (or other audio device) and you'll see what I mean.

Like firestorm said, it does get confusing. Double the SPL (but not double the perceived volume) is a 6 dB change.

If you drop 3db then the sound sounds 1/2 as loud. This has nothing to do with watts.

In terms of power, (Watts) in order to make something sound twice as loud (or increase it by 3db SPL) you have to increase the power by a factor of 10. In other words if you are playing a 1 watt amp and you want to make it sound twice as loud you need to up the power to ten watts. If you want to make the 10 watts twice as loud then you have to turn it up to 100 watts.. This is why when you drop two tubes in a 100 watt amp it doesn't sound 1/2 as loud. You would have to drop it down to 10 watts for it to sound 1/2 as loud. This is also the reason a 36 watt Express sounds damn near as loud as a 50watt Marshall.

Although the perceived loudness thing is fairly subjective, it does, as you yourself go on to say, have a mathematically defined relationship with wattage. Your paragraph about wattage is dead on other than that it should read "10dB SPL" to correspond with a doubling or halving of perceived loudness.