Seeking more clean power from this design

[Atmospheric]

I've got this homebrew Princeton-oid thang than rawnster helped me build last Fall.

I basically use it as a pedal platform for absurdly low stage volume situations. However, the amp just breaks up way too soon for me (about 1/2 up on the volume, after that it sounds pretty ratty). It's also puts out around 1/2 the volume than the BF Princeton I used to own. That's just a bit too quiet to be useful.

I noticed that if I do turn the volume up to around 7.5 on the volume control (a known sweet spot for most Fenders), that the amp sounds much better: mids are fuller, speaker is working right, etc. What this tells me is that I think the power section is operating pretty darned well. I just need some more headroom out of the preamp and PI so that I can get it to 7.5 on the volume control without it falling apart on me.

So rawnster and I have been kicking around the idea of monkeying with the dropping resistors.

I measured the following plate voltages with the current 27K and 4K5 dropping resistors:

V1=220V, V2=222V, V3/V4=~439V

My question is, how high can I safely go with the V1/V2 plate voltages? ~220V doesn't seem particularly low.

Of course, feel free to point out anything obvious I may be missing.

[diagrammatiks]

are you reading the voltage at the b+ node or at the pin?

[Atmospheric]

all plate voltages measured at tube pins.

Also, power tubes biased at ~22ma.

[diagrammatiks]

You need the voltage at the node as well.

Well, knowing the size of the resistor and the voltage at the pin you can work backwards...

but the pin voltage only tells you half the story.

From looking at the schematic really quick...

you've got a cathodyne phase inverter.

depending on the design those are known to sound a bit ratty when they break up.

Have you thought about converting it to a long tailed pair?

Additionally, 6v6s need a bit of drive to really get going...more then a pair of el84s at the very least.

Do you remember the rationale behind the 3.3m/150k to ground voltage divider before the phase inverter?

That's almost 95 percent signal attenuation.

[surfsup]

The signal attenuation i believe is a separate issue/thought, yes? One thing i noticed is all the preamp stages are biased exactly the same. And the bias is set for some decent gain so one might consider biasing one of those more centered in addition to possibly increasing the b+ supply for more headroom?

[Milkmansound]

maybe you can pull that 3M3 and 10p out and replace it with a .001 poly al la trainwreck?

[David Root]

There's already a .02 behind the 3M3/10pF. I'd take out the voltage divider ie 3M3/10pF/150Kand replace it (even temporarily) with a pot. Maybe a 250K or a 500K. May have to reduce the .02 with a larger pot.

If you find a spot you like, measure the voltage divider in the pot at that pot setting, then you can pull the pot and replace it with fixed resistors, or change it to an internal trimpot or a panel mounted pot (which is what I'd do).

[Jana]

Remove the bypass cap on V1b.

Or

Take the wiper from the volume pot and go directly to pin 2 on V2a (totally bypassing everything related to V1b).

You are asking for more headroom--what I think you really want is less gain in the pre-amp so you can turn the volume knob up higher.

[Firestorm]

I don't think there's much to be gained from messing with the preamps; it doesn't look like any of the first three triodes is in danger of clipping.

I don't see any grid stops on the 6V6s. You should experiment with that; something bigger than stock Fender 1K5s.

You should also have a large grid stop on the cathodyne itself. Merlin has a whole discussion of this in the preamp book and on his site. Try 470K.

You can also play with dropping the grid loads on the 6V6s. I always liked the way 150Ks sounded on 6V6s that were running over 420V.

[statorvane]

Do you remember the rationale behind the 3.3m/150k to ground voltage divider before the phase inverter?

That is a leftover from the reverb portion of the Princeton. Since this schematic doesn't have one, it doesn't make sense to kill the signal, then amplify it again. That can only add noise, which may be part of the problem.

[wicker]

Try this:
http://www.freewebs.com/valvewizard/cathodyne.html

[David Root]

I had forgotten about that fix for the cathodyne PI. That should do it if that is the problem.

[Firestorm]

all plate voltages measured at tube pins.

Also, power tubes biased at ~22ma.

I forgot the most obvious fix: if the output tubes are closely matched, try rolling them back to 18mA at idle.

[M Fowler]

Use a solid state rectifier plugin and change the V1a cathode bypass cap to 1uf for less gain. Use all 12AX7 tubes and swap 6V6 for 6L6.

Mark

[Atmospheric]

First, thanks one and all for the responses.

You need the voltage at the node as well. Well, knowing the size of the resistor and the voltage at the pin you can work backwards... but the pin voltage only tells you half the story.

Unfortunately, I had already removed the dropping resistors before I posted, so I can't take that measurement without reinstalling them. I bought a bunch of different values. rawnster mentioned that he found another version AA964 schematic that shows two 18Ks in there. So I'm planning to try that next, take some measurements, see how it sounds. I will be sure to also measure the B1 values from now on. I've got two versions of AA964 schematics. I will post those as well.

Have you thought about converting it to a long tailed pair?

Here's a little more background on what I intended this amp to be. I previously owned a BF Princeton (non-verb) that I was forced to sell. I had always regretted doing that. So I basically wanted THAT particular amp but without trem. Just a minimalist signal path. I always dug the sweetness of that old Princeton, especially compared to DRs. I figured that was probably due in part to the cathodyne PI. So I'd like to keep that for now. I may consider a long tailed pair later on if I can't get the cathodyne to work satisfactorily.

Do you remember the rationale behind the 3.3m/150k to ground voltage divider before the phase inverter?

That is a leftover from the reverb portion of the Princeton. Since this schematic doesn't have one, it doesn't make sense to kill the signal, then amplify it again. That can only add noise, which may be part of the problem.

The entire V1b stage was suggested by someone rawnster knows who has a lot of experience with Fenders. I'm inclined to take Jana's suggestion and just jump around it to see how things sound without it. Hopefully, rawnster will jump in at some point. We have already tweaked the V1b stage and added an external pot to control gain into the PI. It doesn't do *that* much.

I don't think there's much to be gained from messing with the preamps; it doesn't look like any of the first three triodes is in danger of clipping.

I don't see any grid stops on the 6V6s. You should experiment with that; something bigger than stock Fender 1K5s.

You should also have a large grid stop on the cathodyne itself. Merlin has a whole discussion of this in the preamp book and on his site. Try 470K.

You can also play with dropping the grid loads on the 6V6s. I always liked the way 150Ks sounded on 6V6s that were running over 420V.

I think I need to acquire a bit more knowledge before I can use what you are suggesting. I understand AC/DC theory decently well (I used to be an electronics tech in an R&D lab ), but this vacuum tube stuff is still a bit strange to me. I'm not sure I know what "grid stops" are. By grid loads, I assume you mean swapping out the 220ks with 150ks. Would this result in less signal being appiled to the power tubes?

Take the wiper from the volume pot and go directly to pin 2 on V2a (totally bypassing everything related to V1b).

Once I get the amp back up running, this will be the next thing I try. It's an obvious suggestion, but thanks for making it.

I forgot the most obvious fix: if the output tubes are closely matched, try rolling them back to 18mA at idle.

Another obvious suggestion. Thanks for suggesting that.

However, again I may be completely wrong, but I think the power section is right where I want it. If I crank the amp up, things sound really good except that there is too much clipping if that makes any sense. It doesn't sound like power tube clipping, more like mush somewhere in the preamp. But it's probably worth revisiting what the amp sounds like running a bit cooler at idle.

Use a solid state rectifier plugin and change the V1a cathode bypass cap to 1uf for less gain. Use all 12AX7 tubes and swap 6V6 for 6L6. Mark

Hey Mark. This is ToneMonkey from the old Harmony Central forum. Thanks for chiming in.

Both V1 and V2 are 12AX7s.

I have already tried 6L6s in there. I did get a bit more headroom but not as much as one might expect.

With the voltages running as high as they are (505V across the first filter cap on Standby), I'm reluctant to go the SS recto route. I'd probably to have to get a new PT. In which case, I might as well just build another amp (which I may do at some point anyway). Even if I build another amp, I would still want to get this one functioning a bit better.