Shared preamp cathode trem intensity

[EFK]

Here's a trem circuit which I've worked into a Supro-type amp. It works quite well, at least imho (don't have a lot of trem experience nor much use for it, frankly), especially given that I only have 1/2 12ax7 to work with. This is a really basic unbalanced paraphrase inverter, so the only way I can see to tie the oscillator into the circuit is via a shared preamp cathode. There are a few preexisting amps using such a circuit, the Kalamazoo 2 being pretty much the same thing. Problem is, none of these amps using a single triode carry an intensity control *unless* they're working on a different inverter and this varying the power tube cathode bias. So, this is what it is, and I did manage to work an intensity control into it that works decently (I've heard more dramatic intensity controls). The value of pot I'm using is the most effective out of different values I've tried. The real problem, to my mind, is that there is a one-way interaction between the intensity control and the speed control: when I vary the intensity, the speed changes and so the speed control needs to be readjusted. I don't have enough experience with trem to know whether or not this is normal but it's bugging me.

Any term wizards out there have any ideas for perhaps a better way to create an intensity control, or at the least, a way to kill this interactivity?

(I know there are a couple of Airline amps such as the GIM9171 that do what I'm trying to do in a different way, but those circuits are vastly different than mine and honestly I don't really understand how the cathodes are tied together, so I can't even attempt to alter values).

I wonder if there is any way to nix the shared cathode, give the oscillator it's own cathode r, and then maybe use another coupler off the plate to somehow tie this into a different part of the circuit via a grid. ???

[EFK]

Am I that boring?

Seriously, here's something interesting; the circuit I'm trying to recreate was butchered, but it's possible that the intensity was controlled by putting the plate voltage across a 250K pot (the original intensity pot was 250K, and it definitely was not tied into the paraphase inverter). !!!!!!!!!!!! Now, my initial thought would be, that's nuts, but I was looking through schematics and it looks like the Gibson GA20t did the same thing if I am reading the schem correctly.

What are the dangers of putting high DC voltage across a pot? It shorting to the pot casing? Technically I guess the B+ could short to the chassis in many different locations, is a pot a higher risk?

Wouldn;t altering the plate voltage yield the same effect as altering the cathode voltage, i.e., the speed will change as the intensity is changed? I don;t have a GA20t so I can;t elaborate on it......

[EFK]

Yes, altering the plate voltage via a pot is exactly the thing in practice as altering the cathode voltage; same interactive side-effects, and not nearly as safe I imagine. *sigh*

[octal]

I'm hardly an expert on this stuff, but the 25 uf cap on the cathode of the tube being modulated seems like it will be "fighting" the trem. (By making it harder for the osc tube to pull up and down the voltage.)

Would you consider a mosfet buffer? You could give the oscillator its own fully bypassed cathode resistor, buffer the plate output with the mosfet and drive the cathode of the modulated tube. (Like a 5E9-A Fender tremolux, but subbing the mosfet for the Fender's cathode follower.) That should work well, but I don't know how you feel about letting silicooties into your amp....

I bet your idea of coupling the modulating signal from the plate of the osc tube into grid of the modulator tube would also work. I wonder if there would be more risk of thumping doing it this way? This is just intuitive speculation, nothing based on actual knowledge or theory!

Nathan

[tubeswell]

There are some schematics of different trem depth controls and insertion points in p22-24 of this attachment FWIW.

[EFK]

The 25uF cap is being shared by both the preamp tube being modulated and the oscillator tube; I *think* the oscillator tube needs it in a single-triode oscillator to make full gain, doesn't it? I know the trem effect is quite strong, not weak at all, so it doesn't seem to be a problem. As drawn, in my circuit, it also starts and stops instantly. I'm not sure about thump risk, but if I *don't* use a switch in the line b/t the two cathodes being shared, you can hear a little thump bleed-through when the volume is wound all the way up even with the footswitch disengaged. So, if I change it and tie it into a grid, I'm still going to retain the switched pot to completely kill it when not being used. As far as the mosfet thing, I don;t have any problem with it but you might as well be speaking an alien language to me hahahahahahaha!

Tubeswell that was a really interesting bit of text! His first example is very similar to what I'm doing, shared cathodes, and he notes the GA20t approach of altering the plate voltage w/ a pot. What I don;t understand is why there is no mention of the speed changing as the frequency is changed. So, first, to better understand the problem I have here, I need to understand WHY the speed is changing as the intensity pot is rotated. Why is altering the voltage through the tube changing the speed? I assume it's because the remainder of the oscillator circuit values are remaining static. I don't remember reading of anyone w. a GA20T mentioning this interactivity - have to research that.

[EFK]

How about if I split the oscillator off the preamp tube so it's a standalone oscillator. Then, using another cap off the plate and an intensity pot (like the majority of the trems out there), try to tie it into this very primitive paraphase? The two grid returns for the power tubes (270K/270K + 4.7K string) go to a common ground, so maybe I could lift that and tie in there. It would add to the grid resistance total but I think I'm ok in cathode bias. Maybe could do something like in Darr's illustration like adding another cap and resistor between the intensity pot and the grid R's to give it a stable ground reference and keep the R value from creeping up too high. Have to think about this.

[tubeswell]

I'm not sure what you are suggesting with the last suggestion. In terms of your depth pot altering the trem speed in the circuit in your first schematic, the pot is changing the bias of the oscillator, so I guess that would be causing it. What about just making a conventional CF stage following the LFO stage and running the depth pot from that? (See example)

[EFK]

I only have a 1/2 triode to work with here. If I could add another tube, I could do a "standard" 60's Supro trem or one of the Fender trems but I'm in a 5F4 chassis. For fixed bias the 6G3 is hard to beat - used that before, but this is cathode bias. Anyway - what I was hoping for, as per the way it was originally set up (first drawing above) would be a way to alter the "blending" of the two cathodes together without actually changing the bias, but that's not going to happen. So, I've split off the LFO, added another cap off the plate and now I'm experimenting w/ resistors and intensity pot to inject it somewhere else. I can do it into the grid of another tube or I can add it to the PI grid (which would be perfect as I could then have trem on both channels), but either way at the moment, I am getting a LOT of serious thump. It works, but too much thump and my output tubes are throbbing so I'm working on that.

[tubeswell]

If you only have one triode, you can get the same kind of result with a triode and a (high-voltage) mosfet source follower (in place of the 2nd triode). Just put a 100R-200R resistor in series with the gate and put a 12W zener between the source and the gate (cathode/banded end pointing to the gate)

[EFK]

You guys and these g** damned mosfets!!!!! hahahahahahaha just kidding. It's like an alien language to me. I have a hard enough time understanding plain old 50s and 60s circuits.

Looking at it from an old-fashioned perspective, it seems like thumping is a common problem. I've got the new setup functioning pretty well now, all through trial/error and a bunch of clip leads, but I'm trying to minimize the thump. I'm injecting the signal into the input grid for the PI, which gains me the advantage of having the trem on both channels, and I've got it setup so that there is no discernible volume/gain loss when the trem is engaged (I put the switch in the line to the PI input). Still woking with the thump, which doesn't increase with volume, so it's mostly only noticeable when you're not playing. I'm getting a +15/-15 volt swing on the line into the PI. Maybe this is too much, hence the thumping? Going to play around a little more.