wreck-spec Princeton Reverb, sounds "hard" .. idea

[krash]

OK, I didn't build a t-wreck clone from scratch. I have a Princeton Reverb (silverface) that has been basically converted to express preamp (2.5K V1B cathode res, .002/150K coupling network, 10K V3A cathode res, .1uF PI input cap, t-wreck tone stack) but with basically the stock Princeton output stage & PI (split-load, fixed bias, 6V6's). I am working on a schematic and I can post it if it helps.

Last week it was bright sounding with Fender tone stack and I changed it to wreck-spec and that improved the brightness a lot, now it has tremendous range. But the attack is kind of "hard" sounding. What causes this? Is this a normal t-wreck feature? Maybe it's the cold biasing of the preamp tubes? I did put a VOX-style post-PI "cut" control (.01uF/250K pot) in and it helps a little bit, but still the attack is kind of "hard". Can't really describe it any other way.

This "hard" sound goes away when I play the amp through my Weber MASS, fwiw.

I'm not trying to do a wreck clone, but just working this little amp over to get lead guitar tone, so the wreck-spec preamp is a part of the experimentation phase. It's almost there, but I think I'm still tuning. It's a whole different animal than before. Any advice on what causes the "hard" sound?

[krash]

here's a schematic of how it is at present (excluding tremolo & power supply). The reverb has been disconnected/removed. The feedback is the stock/unmodified feedback but I'm thinking of moving the feedback to the cathode of the PI, maybe having it come into the last 'wreck stage is not good.

thoughts?

[krash]

hit post too soon. here's the schemo.

[doctord02]

Whats up with the 10k/47k split cathode tying the NFB to the third stage? The feedback loop should tie to the tail of the PI. That third stage should be nothing but a 10k going to ground... Thats the clipping stage... My guess is that you'll need to change the Princeton PI to something more like the Kelly '90 or A1a schematic. I also suspect the stock Princeton PI wont sound like the LTP in the wreck...

edit - I re-read and see you were questioning the NFB connection to begin with...

[krash]

that's 10K/ 47 OHM. That's the factory Fender part (47 ohm) and I put in the 10K (in place of 1.5K Fender part). This is just the regular feedback connection point in a Princeton Reverb. Converting that to a clipping stage makes the feedback sound mucho bad.

So you say move the feedback to the PI but if I do, it works like a master vol... radically attenuates the signal with any value feedback resistor below about 250K and doesn't do much of anything with anything above 250K... dunno why. For the moment I've just disconnected the NFB but would like to find a way to make use of it to give a bit more stability to the power stage (bottom end control mostly -- gets flabby with no NFB).

[doctord02]

The stock Princeton PI isnt a long tail pair like the wreck circuit uses, so you either have to change the PI or do without the NFB. I kinda like the sound without the NFB connection in a hybrid hodge-podge amp I built... It's very loose sounding/feeling...

Changing the PI to a long tail pair will be closer to the wreck, but a major change from the Princeton... Take a look at a BF Bassman or BF Deluxe and see how different their PI circuit is configured... They use both halves of the tube to split the signal, whereas the Princeton uses a single half.

[krash]

Thanks, and yes I have considered changing to a long-tailed PI as well. I am not using the reverb in the amp so I have an extra half-a-tube there but this is a MAJOR change do the amp, and ideally, the whole PI would be in one tube rather than split across two as it would be if I used the reverb recovery triode for the other half of the LTPI.

The amp is kind of finicky and real bright. I don't know. I just think maybe I don't like it. I like a lot of things about it, but I am irritated by other things about it. It does about two things super duper well but does not really do much else. I guess I have some more tuning to do before it's really done, but I just don't know where to start. Going to LTPI will make return to PR spec very very difficult.