Dropping B+ substantially

[gui_tarzan]

So I've got a Pro Reverb UL SF amp that the guy wants to put 6V6GTs in but the first stage B+ is 538v. I have new TungSol 6V6GTs and from what I see the plate voltage should not be higher than 400v or so. The original OT was roasted, I think because someone changed the funky output jacks and that screwed up the impedance, and all I have at the moment in stock is a 45w Heyboer non-UL OT. It actually sounds nice but won't handle the 6L6 power output in this circuit. So we chance roasting the new OT. The reason for using this particular OT is that I have one on hand and he doesn't want to spend the money on another UL transformer and have the same LOUD amp. I suggested fixing it, selling it and buying a different amp but he wants to try this, knowing the caveats.

He said he's not going to play loud, and I believe him, he really seems to have a good handle on this. All he uses these days is an acoustic so the goal is to lower the power output substantially. Half the battle was the OT which we solved, but the high voltage is really complicating things. The 2nd position of the power rail is 440v so there is quite a drop between the 1st and 2nd stage to begin with.

I've considered using a voltage divider to knock it down to the 450v range using a large wirewound 2k and a 470 ohm 5w resistor, and running the normal OT power to this voltage divider and dropping the 2nd stage just a tad, 420-425v. The first stage runs through a pair of 220uf/300v caps in series and the grids would be run off the 2nd stage along with the PI.

This might be a no-brainer to you guys but that's why I'm asking before I try this - will this work reliably? At all? Should I discard this idea completely? Can it be tweaked to operate the way we want? Putting more money into a DR PT isn't in the budget, but I don't want to blow something up if we try this. It's really out there in left field in my mind but I have to ask or I won't know.

[Enzo]

First, did you write that 538v on the drawing? It doesn't look like the Fender writing. And if you measured that, did you do it with all the original tube types installed? Circuit voltages only have merit when the circuit is complete. If the power tubes are out, for example, the voltage in the B+ will rise substantially.

Just in general, power is not loudness, so when you say you want to reduce power a lot, I think you really mean you want to reduce loudness. As an acoustic guy, chances are he won't crank the amp looking for power tube distortion anyway. perhaps consider reducing the signal going into your phase inverter.

I wouldn't be trying to make a voltage divider at my B+ like that, besides I believe you drew it incorrectly. First consider that 2k 10w resistor will have the full 538v across it. I suspect that would drag the voltage down just by loading, but at 538v, 2k draws about a quarter amp, 269ma, and unless I slipped a decimal, it would then be trying to dissipate 144 watts. Hot isn't the word. I THINK you meant to draw the 470 ohm below the 2k not above it, on the B+ line to the OT. What you have won't be a voltage divider. If you make that change, then your divider would try to knock 100v off, but considered in the circuit, remember the circuit load is in parallel with that 2k. That changes your ratio. Plus even at only 440v, that poor 2k will still try to dissipate 100 watts.

Others have discussed power scaling with MOSFETs, that would probably be a better way than heating up resistors.

Aha, I just looked at your graphic, you drew that as a divider, you just put the 470 in the wrong spot on the schematic. All that watt stuff still applies.

If you look, right now the screens grids and the PI B+ have a 1k resistor between them. You don't want to connect the screens and the PI to the same node, keep them separate.

[The Dude]

I have the same question as Enzo. Is the 538 volts measured with tubes in and amp running? If not, further discussion is a bit premature. The voltage will certainly lower when the amp is running.

The only other thing I have to add at this point is that I don't see where the 2K/10W (additional resistor to ground) in your diagram is a good idea. You're only making the power supply run harder with no benefit to the amp. Unless creating heat instead of sound is your goal, all you should need to do is add a choke or resistor in the B+ supply. The rest of the voltage divider is the output circuit itself.

There are plenty of ways to do this- chokes, resistors , zener diodes, mosfet regulators, etc., but first lets see what the voltage is under load.

[Chuck H]

As above... Voltage, loaded? moving on...

Why in the world does a guy who plays clean want a lower power amp? All you need to do is turn it down. Maybe show him the volume knob and explain it to him. Ok, that was snide. Sorry in advance, But seriously, with the amp running correctly and capable of proper adjustment a bigger amp is better than a smaller amp for acoustic guitar. I've played many coffee houses happily plugged into a 400W P.A.! No one ran away screaming and holding their ears (Not because of the volume anyway ) I just don't understand where the problem is if the guy is willing to work with that amp, as in lug it around, then size and weight aren't the issue. So what's the problem with a bit more power than needed for clean, reproduction sort of amplification? From the description so far I don't think "he really seems to have a good handle on this".

Rant over.

[Mick Bailey]

A single power MOSFET bolted to the chassis and a three components solderd onto the legs is (IMHO) the best way to drop B+. Cheap, reliable and easy. A low-power Zener or two sets the drop.

Any non-switching method involves dissipating heat and the chassis makes the best heatsink - just remember to isolate the tab depending on the type of device. For more info see RG's 'Mosfet Follies'

Here's a pic I saved from somewhere showing how compact this can be made:

[gui_tarzan]

Yes, the 538v is with all tubes in place. That's my number on the schem, Enzo, the original schem shows +500v so I attribute the 38v difference to modern line voltage. I like the idea of the MOSFET, trying to use large resistors in this to knock down the voltage just doesn't make me feel all fuzzy knowing how much heat they will dissipate.

Chuck, I get your point. It's something to consider.

Mick, I have read about those and I think it is the better way to go to do what I want to do, but after all is said and done I think this is probably not the right way to handle this. After I slept on it I think I'm going to toss out the voltage divider idea.

[patlaw]

trying to use large resistors in this to knock down the voltage just doesn't make me feel all fuzzy knowing how much heat they will dissipate.

The MOSFET trick will dissipate the same amount of heat. Ohm's Law.

[Chuck H]

Another option would be to lose OT replacement idea and go with a PT replacement instead! No reason the existing OT couldn't be used with a pair of 6V6's. I think it has a primary of 4200 (UL taps not considered). You could try it into double the intended secondary load for a primary of 8400 on a pair of 6V6's (just unplug one speaker, also good for power reduction). The PT OTOH is simply unusable for a pair of 6V6's, but it's just a typical three secondary PT (bias, HV, filaments) so finding a more suitable unit should be easy. Save the replacement OT for another project and replace the PT instead.

Or, use the MOSFET dropper and don't trouble with replacing any transformers. You don't need to use the UL taps just because they're there either. So if you want it wired more like a typical guitar amp you can do that and just omit the screen taps. You'd need to get a little creative with the HV rail supply for the screens but I don't think any real modification should be necessary.

[gui_tarzan]

I agree, and after carefully considering all you guys have said I'm not going to try this on this amp. One of the reasons I throw out these outrageous ideas to you guys is because; 1) I know you'll be honest with me if they're bad ideas and rein me back in when the idea is bad, and 2) tell me what the RIGHT solution is. I think the best thing for this amp is not to try to convert it to 6V6s.

[gui_tarzan]

Another option would be to lose OT replacement idea and go with a PT replacement instead! No reason the existing OT couldn't be used with a pair of 6V6's. I think it has a primary of 4200 (UL taps not considered).

That's not a bad idea Chuck, but as I said in the original post the OT is shot. It does not put out anything at all and after removing it from the chassis you can see on the bottom where it cooked.

[Chuck H]

That's not a bad idea Chuck, but as I said in the original post the OT is shot. It does not put out anything at all and after removing it from the chassis you can see on the bottom where it cooked.

Ah. I read past that to the question. Ok then. I'd go with the MOSFET unless the amp will just get repaired and replaced. Not a very collectible amp anyway so the mod is attractive. No one it beating down the door for a UL Pro Reverb.

EDIT: Just looked on that auction site. UL PR's seem to go between $450 and $650. Not as bad as I thought, but not great when you consider the cost to repair it and then subtract that from the funds available for a different amp.

[potatofarmer]

You want to significantly drop the power and voltage without creating tons of heat? The PT has a center-tapped secondary - convert it from a bridge rectifier to a FWCT.

Actually, you know what's even easier? Just look at whatever reissue Twin had the "1/4 power switch". I'm pretty sure it just moves the plate and screen supplies to the middle of the stacked caps. It's a DPDT switch, and the other pole adjusts the bias voltage for the reduced plate voltage.

Found it, it's the "red knob Twin" - http://ampwares.com/schematics/the_twin.pdf

So just copy that, but don't include the high-power switch. Tweak the bias and preamp supplies as necessary. Done.

[Chuck H]

You want to significantly drop the power and voltage without creating tons of heat? The PT has a center-tapped secondary - convert it from a bridge rectifier to a FWCT.

Brilliant! Don't know how that got past me and others.

EDIT: It occurred to me that there may well be TOO MUCH voltage reduction by changing rectifier arrangements. Probably something around 250Vp in this case.

Actually, you know what's even easier? Just look at whatever reissue Twin had the "1/4 power switch". I'm pretty sure it just moves the plate and screen supplies to the middle of the stacked caps. It's a DPDT switch, and the other pole adjusts the bias voltage for the reduced plate voltage.

I proposed that sort of circuit here and took some heat for it. I know it'll cause uneven ware to each cap, but so what! I can't remember what else may need to be considered about this circuit. Another good call IMHO.

EDIT: It also occurred to me that this would produce about the same voltage as the scenario above. So how about a three totem arrangement with the HV taken from one node down.?. All this assumes some modification to the power supply.

[überfuzz]

I've once grounded the output jack in a new build, but luckily double checked it before I vent live. I guess your amp had something similar done to it.

About the transformers. I think it's a wonky idea. Tell him to get a new stock OT, or a new amp with a 6V6 power section. It's probably just some wizkid that has chimed wisdom over him '6V6 sounds way better'... If you play acoustic guitar!?

[Chuck H]

I want to second uber's sentiment as I iterated in post #4. If it were MY amp and I was going to play mostly acoustic guitar through it I would just fix it. I only moved on with other options because it addressed the actual question. Which, while it seems to be what the amp owner wants, isn't a good solution IMHO.