Disconnecting two power valves. Will this work?

[Mostro]

Hey guys. In my quest to tame the volume on a AC30 build, I came across (on a Dumble schematic) the idea of disconnecting a pair of valves in a four valve power amp to get less power. He does it by switching a 10K R in the cathode, putting the valves into cutoff, so that pair won´t contribute to the amplification. Now, that is in a fixed bias amp, and mine is a cathode bias one. So I redrew the idea as I believe should be applied in my amp (the values are approximate):

[IMG:900:713]http://i68.photobucket.com/albums/i5/Morocotuco/Disconnect-two-valves_zps3d139cbb.jpg[/img]

What do you think? Will it work? The nice thing about this is you can switch "on the fly", the impedance doesn´t change so you don´t have to switch the speaker´s impedance switch or plug into another power out, the valves age equally, the voltages don´t change (at least I think so!)...

Are there problems that I don´t see?

Opinions welcome!

[d95err]

It should work. Just make sure the switch can handle full B+. Voltages will rise a bit, since the powersupply will draw a lot less current.

[Jana]

What's going to happen to the heater to cathode voltages?

[Jana]

With that capacitor in there, I also predict a big thump or pop when you flip the switch.

[Mostro]

d95, why should the switch be able to handle full B+? Will the voltage in the cathode rise that much? Or it´s just being cautious?

Now that I think of it, how much will the cathode voltage be? Hmmm...

Jana, I´ll investigate the cathode to heater max voltage for EL84´s. And, regarding the pop, I see how that can happen when connecting the cap to ground. Will it be a better idea to connect the cap across both resistors, I mean, to ground directly and only switch the R? Or, switch the resistors order ( I mean, the bigger one on top and the smaller one at the bottom with the cap permanently to ground)

Questions, questions.

EDIT: OK, the max cathode to heater voltage in EL84´s is 100V. But if I have the center tap of the heaters grounded at the non switched valves, that buys me about 10V more. So, I should no exceed, say, 90-95V at the cathodes when putting them into cutoff. OK. Let´s say that each valve draws 40 mA through the cathode with the 100R and that drops 8V. If I up that to 10K, I will get, ehh... B+? Maybe the R should be smaller. 1K? Will 1K get me 80V?

[Phil_S]

What are you really after here? If you cut it from 30W to 15W, at 15W it will still seem like it gives 90% of the 30W volume. To get to half volume from 30W you need to go down to 3W!

All the considerations when you lift two tubes -- change in Rk, output impedance/speaker matching, B+ rising, ligher heater load, etc., IMO, it isn't worth the effort!

OTOH, if you are chasing the technical aspect of this, please, go right ahead. I'm not trying to rain on anyone's parade!

[Stevem]

If the amp in question has two speakers and either driver can handle near the full wattage of the amp then just disconnect one driver and you will have way less SPL level that it sounds like your shooting for.
The impedance mismatch will cut another 1/3rd off of the wattage.

[Mostro]

Phil, I´m trying to get power amp distortion at lower volumes. I have a VVR installed, I thought this could be a complement to that. No single solution will get me significant reduction in volume, maybe both will get me closer to bedroom power amp distortion, so to speak. The VVR alone I can reduce to 30% voltage before it starts to sound awful.

What I like about this is that the output impedance doesn´t change, the rise in B+ I pressume will be small (I mean manageable). The heaters load shouldn´t change...

I know 15W vs. 30W is perceptually small, but I was hoping that this, combined with the VVR will be more effective in lowering perceived volume.

Also, I have the amp in pieces all over the bench, so I figured now´s the time to try whatever I want just for the fun of it and to learn more, and then arrive at the final schem and finish it.

[Jana]

If you are trying to get down to bedroom levels, this isn't the approach. Bedroom levels *only* need about 1/4 watt of power. That is one quarter of a watt. I think you would be better off building the amp in all its glory and then building an isolation cabinet. The benefit of that is that you can use it with more than one amp (not at the same time, of course).

[Jana]

Food for thought.

http://home.earthlink.net/~dnitzer/4Haa ... cibel.html


Depending on your speakers, they have a sensitivity in the range of 97 to 100 db. That is measured with 1 watt of power at 1 meter distance.

1/4 watt of power will cut that by 6 db so you are still at 91 to 94 db.

With a 100db speaker,

30 watts of power will be 115 db.

15 watts will be 112 db.

How's that isolation cabinet coming along?

[Firestorm]

Phil, I´m trying to get power amp distortion at lower volumes. I have a VVR installed, I thought this could be a complement to that. No single solution will get me significant reduction in volume, maybe both will get me closer to bedroom power amp distortion, so to speak. The VVR alone I can reduce to 30% voltage before it starts to sound awful.

What kind of master volume are you using?

[pops]

Two things i can think of to do:

Ground the input of two grids, shouldn't take much of a switch and will not pop.
Take two tubes out and unhook one speaker.

[Mostro]

(I knew I shouldn´t have used the word "bedroom")

First of all, thanks all for the answers.

My last answer was kind of in a hurry, had to leave to help a student with his guitar buy.

What I´m trying to do is, try to get the lowest possible power out of this AC30 type amp to be able to crank it and get power tubes distortion, while retaining the possibility of using it at full power for clean sounds (that this amp makes beautifully). The operative words here are "the lowest possible power"

I know I won´t be able to get it to 3, 2, 1, or 1/2 watts, or play it beside a sleeping baby. I´m not really aiming for that. I just want to see how much I can reduce the power with the methods explained earlier. So I´ll have an amp that is quite powerful, can make nice clean sounds, can make very nice distorted sounds but at a deafening volume, aaaaaand can make nice distorted sounds at a somewhat lower level, that could be just 0,1 dB lower, that´s OK, I´m not aiming to reinvent the wheel, just have some fun with it. Since the amp is in pieces on my bench, now´s the time for trying every idea I can come up with, see if they work or not, discard the ones that don´t and eventually consolidate the schem and finish it. I´m just a hobby builder, but love to learn all I can. Maybe this "disconnect two valves" idea won´t work, that´s OK, i´ll try it and learn from the experience.

I know about speaker sensibility, iso cabs, SPL, dB´s and so on. I don´t want an iso cab. I actually have a little twin LM386 amp that, when dimed, makes quite a nice sound through a decent cab and speaker and at a manageable volume, I think it puts out about 1 watt.

And just to set the record straight, I don´t play in my bedroom!



So, beyond the fact that this idea maybe won´t give me a significant volume reduction, I want to try it. So, what do you think about this:

d95, why should the switch be able to handle full B+? Will the voltage in the cathode rise that much? Or it´s just being cautious?

Now that I think of it, how much will the cathode voltage be? Hmmm...

Jana, I´ll investigate the cathode to heater max voltage for EL84´s. And, regarding the pop, I see how that can happen when connecting the cap to ground. Will it be a better idea to connect the cap across both resistors, I mean, to ground directly and only switch the R? Or, switch the resistors order ( I mean, the bigger one on top and the smaller one at the bottom with the cap permanently to ground)

Questions, questions.

EDIT: OK, the max cathode to heater voltage in EL84´s is 100V. But if I have the center tap of the heaters grounded at the non switched valves, that buys me about 10V more. So, I should no exceed, say, 90-95V at the cathodes when putting them into cutoff. OK. Let´s say that each valve draws 40 mA through the cathode with the 100R and that drops 8V. If I up that to 10K, I will get, ehh... B+? Maybe the R should be smaller. 1K? Will 1K get me 80V?

Firestorm, I´m using a PPIMV,I´m scaling the PA only.

[Mostro]

Two things i can think of to do:

Ground the input of two grids, shouldn't take much of a switch and will not pop.
Take two tubes out and unhook one speaker.

Nice idea to ground two grids...i´ll try it!

The "take two tubes and change the speaker load" idea will of course work, but my idea has the advantage of being able to do it without turning off the amp, unplugging speakers, taking out the tubes...

By the way, I´m pretty sure I´ll make a 1x12" cab for this, I have a Red Fang that should fit the amp soundwise. So no chance of unplugging one speaker.

[Jana]

Well, if you're up for experimenting, do as Pops says and ground two of the grids.

You could also switch the tubes from pentode to triode mode by switching the screen grid resistors from the supply node to the plates.

Those two things should drop you down to about 5 watts (two EL84s in triode mode).