The Raven - CKD

[shoggoth]

Thought about it some more, because this is such an interesting and impractical amp. I like complexity.

a. Zeners have a reputation for being noisy. Since variations (avalanche diodes) are used as random number generators, I can see that.

b. What's your plate supply to the fifth stage? 125v with the regulator was the idea right? So the 12AU7 was biased at -8v with respect to the cathode. The zeners were keeping your signal at +-6.2v or so, and prevented the clipping you see now w/o them, by clipping earlier.

So thinking "aloud" as it were, I suggest:

1. Use different diodes. Use silicon diodes instead, with identical resistors between them and ground that will get you attenuated to +-6v or so. It should be more "musical" as well with the resistors softening the clipping. It can be asymmetrical clipper by using two different values of resistors.

Hmm if you go symmetrical, you could just use one resistor and save a little space.

2. If you want to drill another hole and have spare heater current, you could use a small-signal dual diode tube instead of a pair silicon diodes

3. Maybe increase attenuation (resistor between coupling cap & 100k/bright cap junction) so that at very lowest levels of gain from previous stage, signal comes through clean (that is, its amplitude fits within the +-6v you are aiming to clip at). Gives you more control over tone w/ the gain knob on the prior stage

[shoggoth]

Whoops I am a doofus. A resistor biasing the diodes will of course act as part of a voltage divider, and thus the clipping voltage will vary with the signal strength. So it will be softer clipping, but it will almost always be clipping regardless of signal voltage. You're already "almost always clipping" because of the signal amplitude from the 4th stage w/ the zeners, but it's probably going to be a different tone to the one you already know & like.

You'd need a giant string of regular diodes to replace the zeners while reducing noise. Maybe worth it, maybe not.

Anyhow, I'd still go with idea #3, at least temporarily, as it will let you put a clean signal through that last stage for debugging purposes.

[Meat&Beer]

Thanks guys! I'll try out these ideas in the order they were posted, and report back the results later tonight.

The purpose of those diodes, as far as I can explain in paraphrase;

The "designer" or this amp said here, try these in that spot there. We're hoping for no audible difference, these are just to help ensure stage 5 tube is happy, and within its limits/operating range. They were never meant to add gain, as I'm sure you can all tell, we've plenty on tap already!

I just keep picking on this area because the scope now additionally says so, and I get most reaction here.

Thanks again, all the best!

Chris

[Meat&Beer]

Well, we're getting there it seems.

Increasing attenuation in a couple different ways/areas is getting rid of that gross over-clipping sound, and trying a thing or two here or there while watching the scope is definitely very beneficial.

It's a lot of tweaking and adjusting, un-tweaking and re-tweaking elsewhere type of thing. I'm trying to find a sweet spot, so to speak, and the right amount of one thing versus over-compensating type of idea.

The 4th stage actually did get a bit of a decrease in the cathode R, down from 10k to approx 7k. You called it, Shoggoth! (It want's to be 7.5, or so I'm told, but all I had handy was that 10. I tacked in a 22k across it.)

Thanks very much for the assistance, hopefully soon I'll be able to post a quick little sound clip on here for all to check out! As soon as I can please my own very picky and discerning ears.

[Meat&Beer]

It really seems that this problem is blocking distortion... I've attached an up to date schem as to what is in there. (Minus my attempts at voltage dividers) Can anyone sort of give me a lesson here in how to calculate what they should be for each stage?

I tried matching the grid stopper and grid leak, as seen on so many schematics to form the divider. This is just guess work, I know. It helps, but there's still remnants of the undesired fizzyness, and at this point I'm just attenuating the signal.

There's 12ax7's, 12au7's, 310 plate volts here, 125v there, 100k and 330k plate resistors... (Don't sing the Old MacDonald song!)

Before this, I attached a 10k Linear pot in place of the cathode resistor, thinking it was maybe asymmetrical clipping that I was hearing and didn't like. Adjusted them while watching the scope, fired it up, still sounds like fizzcrap.

Increasing the cathode R also helps attenuate the problem. But, then the tube is getting thrown out of bias happiness. Not sure how big of an issue that is, aside from probably shortening its life or something.

Any help/guidance is greatly appreciated guys, thanks!

*Edit* Do I have to take all the stuff on a webpage like this into my brain? So much math! Damn you math!

http://www.ampbooks.com/home/tutorials/lesson-004/