Lowering my 50W Plexi clone efficiency

[Hardtailed]

2 years after building this amp (think Marshall 1987 with 2 EL34, cascaded preamp and no NFB), it is clear to me: I will never need that much headroom.
I run the amp so low most of the time, it might as well be solid state.

I'm not so much looking at lower power output as getting more "tube character" early on. Removing the NFB helped, but it's not enough.
Here's what I'm thinking:

- Add a sag resistor (since I don't have the filament supply for an actual tube recto)
- Modify to cathode-bias (maybe even add a switched for bypassed/unbypassed)
- Lower plate supply filtering (currently 100uF)

I'm guessing just the last 2 should give a much more reactive amp (I just want it to react more when I dig in). Cathode-bias is rather easy and will have the nice side effect of dropping the effective plate voltage.
Lower filtering… I just have to disconnect the 2nd half of the 50/50 can and it's done.

I was actually thinking about putting the extra 50uF cap ahead of the sag resistor. Bring the rectified B+ to the first one, connect the sag resistor in between and feed the plate supply and the choke from the 2nd one. Effectively that's an extra filtering node so an even cleaner supply.

What I'm unsure is what to shoot for as the sag resistor. I remember reading that 100ohm was the magical number in a SS-rectified 18W clone, but that's with a lower current draw. Yet, I don't see much more than a 10V drop happening from idle to full-tilt (on top of a 8-9V drop at idle, right now B+ is 487V). Anyone has experience with a sag resistor in a 50W-ish build?

Any other suggestions I could use to make this 50W more alive?
(I've tried attenuators and just don't gel with them… and I really want the sound of a pair of push-pull EL34 working in Pentode mode)

Didn't I read somewhere that JJ 6V6s can actually survive that kind of voltage? Plug my 16ohm cab into the 8ohm and I effectively get a 6800ohm primary Z, add the voltage lost across the sag and the cathode bias resistor and we're well under 500V too.

[txstrat]

JJ claims 500v for their 6V6s tubes. Look here: http://www.jj-electronic.com/pdf/6V6.pdf
What about a smaller OT for 6V6 type tubes? I mean in the 20w range (although I doubt it would be much less headroom).

[Steve Conner]

Remove the power tubes and connect the PI outputs straight to the OT. That will do it.

[Chuck H]

http://music-electronics-forum.com/t23924/

[teemuk]

You can get phase inverter overdrive (which by its characteristics is reasonably close to power tube overdrive originating in the final stage) with plenty of different setups: You can make the PI's "bias" resistor variable (e.g. Carlsbro TC60), you can vary PI's plate voltage with a manually adjustable regulator, you can fit in a post-PI master volume, you can fit in a 12AU7 tube to attenuate the PI's output (e.g. Seymour Duncan Convertible 100), etc. Even building a switching scheme that swaps the output stage to use only preamp tubes with much lower power output isn't that far-fetched idea; some modern amps like Traynor Darkhorse or Blackstar HT-5 do so already.

[Hardtailed]

You can get phase inverter overdrive (which by its characteristics is reasonably close to power tube overdrive originating in the final stage) with plenty of different setups: You can make the PI's "bias" resistor variable (e.g. Carlsbro TC60), you can vary PI's plate voltage with a manually adjustable regulator, you can fit in a post-PI master volume, you can fit in a 12AU7 tube to attenuate the PI's output (e.g. Seymour Duncan Convertible 100), etc. Even building a switching scheme that swaps the output stage to use only preamp tubes with much lower power output isn't that far-fetched idea; some modern amps like Traynor Darkhorse or Blackstar HT-5 do so already.

I don't really like PI drive... had a PPIMV and removed it. Went with a regular MV in the end.

I'm not looking for more gain, just want the power amp to be more dynamic, to sag a bit under load.
6V6s might not be the trick either cause then I'd end up with an oversized set of trannies compared to the power output...

[guitarmike2107]

My order of suggestions

Try big screen resistors
Try a Sag Resistor.. must be high wattage though.. work it out
Try reduce filter caps to 47uf

try one thing at a time, you may find a few small adjustments in each area get you where you want to be

[MWJB]

"6V6s might not be the trick either cause then I'd end up with an oversized set of trannies compared to the power output... "

You say that like it's a bad thing? ;-) As long as plate & heater voltages don't end up as excessive, due to reduced current draw, trannies can be pretty well as big as your back can stand.

I'm personally not a fan of the sag resistor, can make for a wheezy/fuzzy tone.

Power supply being "more dynamic" to me suggests that it WON'T sag under load. I guess you are looking for a greater envelope of sag & decay (to me this is less dynamic, more compressed) ? As well as the suggestions above, you could also consider...

Switching out a power tube, by lifting the "out of phase" (OOP) power tube's (closest to PT) coupling cap. Still have dc current at both ends of the OT primary, less worry about dc core saturation & magnetisation of the OT. Don't disturb the grid load resistors or the neg dc voltage. Expect pops whilst switching, do this in standby mode.

Cathode bias switch you have already thought of, easily installed on a DPDT to go from fixed to cathode...here's 2 extentions on that theme:

1) Switchable cathode bypass cap for the power tubes.
2) SPST shunting the out of phase tube's grid load resistor. Don't ground the grid load resistor, if you do you will kill your tubes in fixed mode.

In cathode bias, if you shunt the OOP tube grid load, with a cathode bypass cap on the power tube cathode resistor, you have a SE cathode biased amp (5-10W ouput) . Now switch out the cathode bypass cap and you have a cathode biased, self-split amp & probably half the W of regular cathode bias. A fixed bias to "self-split" switch can be easily fitted using a 3PDT switch, this also eliminates the risk of shorting the grid load resistor when in fixed bias mode. So just the 3PDT & a SPST on the cathode bypasscap will give you fixed bias, cathode biased self-split, cathode biased SE.

Also when trying to cut volume (dB), once you have cut the power, you might consider less speaker area/efficiency, a self split 8W amp running 500v B+ (inc. cathode voltage) and 350"sq. of speakerage, is still fairly loud.

[d95err]

When you tried attenuators, even if you didn't like the tone - how much attenuation did you need to get the poweramp more "dynamic"? Are we talking -3dB? -20dB?
The magnitude of attenuation would help to find out what type of methods to reduce efficiency or power are appropriate.

I recommend taking a look at Kevin O'Connors sag kits. They will enable you to dial in exactly the amount of sag you want for the amp. See Tube Amp Kits, Tube Amp Books, Tube Amplifiers by London Power.

[Hardtailed]

My order of suggestions

Try big screen resistors

I have the classic 1K right now, would 2K be enough to make an audible difference?

Maybe I'm not helping myself with the big fat choke though (it's a 7H, but it's rated for 125mA! My screens are a ridiculous 5v lower than the plates)

[Hardtailed]

"6V6s might not be the trick either cause then I'd end up with an oversized set of trannies compared to the power output... "

You say that like it's a bad thing? ;-) As long as plate & heater voltages don't end up as excessive, due to reduced current draw, trannies can be pretty well as big as your back can stand.

Hahaha, what I mean is that I like the sound of a tube amp that is struggling to keep up, like old tweed Fenders. If I swap 6V6s in there, the massive power supply will just become even stiffer.

And yes, I guess guitarists tend to use the wrong term... when I say dynamic, I mean an amp that actually compress a bit and therefore lower dynamics... It's about the amp reacting in a dynamic way, as opposed to an amp that let my dynamics through clearly.

[MWJB]

No it won't, because the 6V6s will distort earlier, losing shape/dynamics/definition earlier. Bigger PT will hold a constant voltage better, run cooler...I like to rely on less "skin of the teeth" methods to tweak sag/envelope...like bypass cap in cathode bias vs. no bypass cap, softer sounding power tubes like Shuguang/EH 6V6 vs. JJ6V6 (assuming the former take the voltage) etc. e.g. design it in, rather than be held to ransom by it, due to borderline parts/filtering.

[guitarmike2107]

I have had screen resistors up to 4.7k, not with EL34's though, they draw allot of screen current compared to 6l6 say. suck and see

I dont see how the physical size of the choke will be against you, some might comment to differ though.

[EFK]

Hi MWJB: excuse the rudimentary question, but are you saying that in a cathode bias amp I could either (1) lift the grid R on one power tube or (2) lift the grid end of one PI coupling cap and end up with a working single-ended amp? When you state OOP tube, do you mean the tube in-phase or out of phase with the PI input? Thanks!

[MWJB]

Hi EFK,

To switch from P-P cathode biased to SE, you can just ground out the power tube grid load resistor. You can't do this in fixed bias as you will eliminate the negative bias voltage at the tube grids. If you do this on a cathode biased amp, with no power tube bypass cap, the amp will still be push-pull, but self-split rather than PI driven.

For a fixed bias amp, you can lift one end of the OOP PI coupling cap to achieve the same effect. Do not disturb the negative dc feed to the power tube grids. Of course, you can also just pull out the power tube(s) on one side of the OT primary, some folks have concerns regarding OT magnetisation due to subsequent dc core saturation (a tube still drawing curent at each end of the primary should eliminate this), alternating which tube you pull out, from time to time, may help prevent this.

When I say "OOP tube" I mean the "bottom" power tube in the schematic, or the "second power tube" e.g. the tube that takes its feed from the inverted signal/triode B from the PI, next to the PI tube physically, in Fender. I thought that this was simpler that discussing a tube that might be in, or out of phase with the signal at earlier stages...because that changes with the number & type of stages, or which input you use. The "in-phase" tube is the one that would remain in a SE amp ("top tube/first power tube"/farthest from the rectifier/PT). But, yes, the OOP power tube would be in phase with the PI input signal...well, that's my reckoning on the subject & it makes sense to me ;-)