Half Power switch - best way?

[Mike Sulzer]

I'm all for using a combination resistor/inductance, but as someone noted, it becomes unstable and causes high frequency oscillation. Unless someone has a suggestion to make it more stable?

Chuck gave his solution: two coils in series oriented to pickup with the opposite polarity, a kind of hum bucker, but a signal bucker instead. And do not put it near the guitar or in the amplifier chassis. Also magnetic shielding (mu metal or whatever) is another possibility.

[wizard333]

or in the amplifier chassis.

Well that's the rub, trying to find a way to put it in the chassis and not have external boxes involved.

Maybe the dual secondary tap for OT is best. Hard to implement that on an existing amp though, and probably tricky finding a PT that has close to what an existing amp wants to see and also 3 secondary taps, one for preamp, and two for OT.

[wizard333]

Hmm thinking out loud,

Instead of multiple taps, have a 2nd power rail after the first, with new dropping resistor(s) for OT and screens, switch which point both draw from. Have to be a big resistor though and it would kick off some heat, other than that though....

[mozz]

Derive a 2nd +B voltage that is 0.707 times the main +B and switch between the two. When you use the 2nd voltage you'll have half the power. E.g. If you got 500 VDC, something around 350VDC will give you half the power.

Easier if the amp is self biased.

Yes, choke input filter.

[g1]

Instead of multiple taps, have a 2nd power rail after the first, with new dropping resistor(s) for OT and screens, switch which point both draw from. Have to be a big resistor though and it would kick off some heat, other than that though....

This seems to be a variation of the concept in post #14. Is there some drawback of the Fender circuit?

[wizard333]

This seems to be a variation of the concept in post #14. Is there some drawback of the Fender circuit?

Well, good question. I can't answer because I can't get that picture to load up right. I'll see if I can locate a schematic and take a look.

[wizard333]

Ok downloaded the whole schematic and took a gander. Interesting idea; Questions:

1) Effect of having the post-choke voltage not feeding the pre-amp? Less filtering of ripple obviously, other? Seems to me another filter cap in there wouldn't hurt in that situation.

2) Anyone heard one of these recently and can comment on how effective this actually was at power reduction, and any tonal considerations? I don't think I've actually listened to one in 15 years or so, and I don't recall having experimented with that switch.

[Rock Mills]

Lots of interesting ideas here. I'm pondering building a smallish (<30lb) combo amp that that covers the bases from accompanying acoustic instruments to being heard with a drummer involved -- for both guitar and bass. Gotta have some balls for bass, but.... Have consider most of these options -- drop B+ to one or more sections, fiddle cathode bias, kill tubes, speaker dummies..... All change the operating conditions, distortion characteristics, etc. of a tube amp. And as pointed out, lowering the volume changes your perception of tonal balance. I think what we're looking for an 'audio-taper' attenuator that compensates for the human ear so our tubes can weep/warble/scream without making folks ears bleed.

Tall order. It seems to me it's all about sound pressure. Let the amp do/see what it wants to do, and control it's presentation.

I'm thinking I'll go with two dissimilar speakers, 10" and 6"5, in series. Phase switch on one will cancel a portion of the other. The amp sees little or no difference. Overall volume *should* (haven't tested) be less than half with the midrange scooped. More or less what we're looking for.

Generalizing for any pair of 8 ohm speakers, put 8 ohms in series with one, 16 ohms in parallel with that resistor and the speaker and flop the phase. Same nominal ohms, but the speaker only sees 1/4 the power. That portion is out of phase with the other speaker, so it cancels that part of it. Essentially, you go from 2 x 12" to 3/4 x 12" in air-moving capacity. The resistors need to be large enough to dissipate about 3/8 of the amp's wattage. (My guesstimate is that anything over 10W should do for a 40W 2 x 12" combo).

Could get fancy and try to make up for the human ear by filtering some highs around the out of phase speaker. Haven't pursued that (yet). I figure the out of phase 6.5" should rob mostly mid-range from the 12". The lows should be OK. For a matched pair, you might want cut the lows to the out-of-phase speaker with a choke as well.

Seems a rather low-tech way to ameliorate the problem. It's all in the speaker cabinet, no mod to the amp, nor major change to it's operating conditions.
Am I on to something, or am I off my rocker? Has anyone tried anything like this?

[wizard333]

Really don't think you'll like the way it sounds, and if it's a combo, you may run into mechanical issues. I saw a VOX AC30 with speakers someone wired out of phase almost shake itself apart under those operating conditions. With speaker phase corrected, it was fine.

[pdf64]

I think that the frequency cancellation node points will be dependent on the spacing between the speakers, and that a comb type filter effect may be created.
The degree of cancellation / depth of the node will be dependent on the relative sensitivities of the speakers, and also on their frequency/directivity characteristics, and position of the listener.
Hence I suspect that the idea won't achieve the desired effect.

[Chuck H]

Except that it isn't. Where the gain and master knobs are, how hard you hit the power amp even if you aren't pushing it to distort affect tone and response.

I get that. I was thinking more about the uber gain guys that almost always just use the power tubes as part of a reproduction amplifier for the preamp tone, rather than clipping them at all. Certainly there are some cases where even those amps and tones find a sweet spot with a little power tube clipping. It use to be more common than it is now, I think.

[g1]

Rock Mills: good points and welcome to the forum.
All my experience with out of phase speakers is that the lows are the first thing that gets lost. But that has been with equal size speakers.

[Rock Mills]

No question that two identical speakers out-of-phase in the same cabinet sounds like mud. Not useful except for measuring circuit behavior at full power into your actual reactive load without making your ears bleed. More realistic than a dummy load.

My thinking (perhaps wishful) is that having about 1/4 of the power out-of-phase should provide more reduction in volume than simply replacing a speaker with a resistor. Keeping the (anti-) speaker in the circuit would also preserve some of the speaker's dynamic reactance, so the amp should behave closer to normal than it would with a simple resistive load.

Will it sound the same, only quieter? Not likely. Filtering some lows (<500Hz) and highs (>5000Hz) out of the anti-speaker with an appropriate RC bandpass filter should help. The mids would essentially be scooped, approximating our ear's Fletcher-Munson curve.

With the anti at 1/4 power and the bass rolled off, I doubt it would develop the shakes you saw with the out-of-phase 12s. Am I on to something, or off my rocker? Should be doable with a couple of caps, couple of resistors and a DPDT switch. Time to rustle up some parts and give it a go.....

[Chuck H]

Hmm thinking out loud,

Instead of multiple taps, have a 2nd power rail after the first, with new dropping resistor(s) for OT and screens, switch which point both draw from. Have to be a big resistor though and it would kick off some heat, other than that though....

Well... Any resistance in the power supply prior to the plates, and to a lesser extent the screens, sufficient to reduce power by half is going to create tremendous amounts of sag, seriously affecting the tone and feel of the amp. But I did read this and think:

What if you used a dedicated filament transformer and then use the 120/240 primary of the HV transformer to change the voltage on the plates and screens. You would need to simultaneously switch the HV rail resistor feeding the preamp to maintain a similar tone. But since the preamp doesn't draw much current it wouldn't produce the same sag anomalies. This would also likely require the bias to be on a switch too. So that's a three pole switch. You can get that in a high power toggle.

[wizard333]

Interesting idea. If you switch to the 240 primary running on 120 from the wall, the PT will kick out half, but I'm not sure you could change the resistor feeding the preamp enough to make that work. Say in normal operation B+ is 440, your pre-amp sees something like 220 before plate resistors. Now you kick in the 240 primary and your B+ drops to 220. You really have no where to go but down to feed the pre-amp. Even with a bigger swing, that would be a very small resistor to maintain pre-amp plate voltages.

Plus you're adding a transformer, though a dedicated heater PT is a good idea anyway.

Probably not something I'd want to retrofit, but on a new design that would be interesting to try.