High frequency attenuation at the power section

[anson]

I'm looking for ways to attenuate high frequencies after the LTP PI, or preferrably, at the power tubes themselves. The reason is to see what it sounds like when the out put section is very overdriven but not allowed to introduce the resultant high-order harmonics associated with the clipping. I have a hight-cut filter before the PI, but still when I crank the amp it can still get a bit fizzy--which is good and all but I'm curious what it would be like if it didnt, but still had lower and mid frequencies going through.

My amp is 4xel34 btw.

So one idea is to to put a small cap between the gridsof the push-pull halves, but it the past I tried it and wasn't really stoked with the effect.

Another idea, which may be not realistic but I want to see what you all think, is to put a very small cap between the plate and the screen of each tube. Would that make the high frequencies amplify as if the amp is in triode mode with the associated lower gain, while making the lower frequencies amplify as if they are in pentode mode due to the high reactance of the cap to low frequencies?

I'm thinking this because it seems like if you shorted the plate and the screen, it would be triode, but if you put a very large resistor between them, then it would be pentode, and putting a component (such as a cap or maby an inductor?) that has a varaible reactance to frequencies, it might do both.

Anson

[txstrat]

The cap between the grids of the power tubes. Have you done it like VOX did. They have a pot in series with the cap (cut control) That way you could dial in (or out) the amount you like.
just my 0.02$

[anson]

The cap between the grids of the power tubes. Have you done it like VOX did. They have a pot in series with the cap (cut control) That way you could dial in (or out) the amount you like.

Hmm. I thik I'll have to give that a try and see what it's like.

[Enzo]

And another option is to do it on the speaker side, like a crossover.

[yankeerob]

Ever thought about the OT's role in all of this?

[pdf64]

Conjunctive filter (cap across OT primary)?
Or different tubes - el34 can be fizzy compared to 6L6.

[yankeerob]

I was thinking about the load (OT) and how it affects freq. response... but yer why not and fair comment re EL34/6L6... especially new production EL34's...

[Hardtailed]

Many ways to skin a cat

A conjunctive filter is one (small cap accross the secondary, sometimes with a resistor in series)

The Cut control on the Vox is pretty useful.

On my personal amp, a Marshally build, since I loved the tone with the NFB disconnected but found it too bright, I went with a Mesa style Presence control (can't use a real Presence without NFB anyway) which is the same concept as the Cut control, except it's pre-PI. I love it, I can now make my amp sound really thick and meaty and emphasis the highs more in the preamp for a tighter distortion, without getting the typical ear-splitting highs of a cranked JCM800.

If you're overdriving the PI, the Vox cut control will work better of course.

You'd be surprised how little highs the power tubes are adding once you cut them in the preamp like that.

You can also use bigger grid stop resistors on the power tubes

Lots of options!

Oh, and you can use different speakers. I have a G12T-75 and a G12H30 in my cab. The highs on the H30 are so much more pleasant.

[Merlinb]

Many ways to skin a cat
Lots of options!

Yet another option is a feedback-free presence control, which can be achieved by varying the amount of decoupling on the 'unused' grid of the LTP. For example, put a 1M variable resistor in series with the usual 0.1uF cap, plus a 470pF cap in parallel with the whole lot.

[anson]

Thanks for all your advice.
The vox 30 schem that I looked at used a 220k variable resistor with a series .0047 cap between the output tube grids. So basically the high cut circuit, if i'm understaning it right, consists of the output Z of the PI, together with a capacitance that is 2x the cap used?

As for the conjunctive filter... what approx. values should i use there? If it's accross the OT primaries, then would the high cut circuit consist of the el34 plate resistance and the conjunctive filter used?

[Koreth]

Another thing to try would be to put some tiny caps across the plate resistors on the phase inverter. Mesa does this in some of their amps with small value caps around 100pF or so. That gives a -3dB frequency of over 10kHz depending on the value of the plate resistors in the PI, which makes them more useful for keeping supersonic and RF frequencies snubbed so as to prevent oscillation. However, higher value caps could bring the cutoff frequency down into the highs or high mids, or wherever you please. Reducing the gain at higher frequencies should reduce the distortion in the PI and power tubes at those frequencies, and tame the brightness a bit.

[anson]

Thanks for the ideas. I tried a 250k ohm resistor in series with a .0047u cap accross the grids, and it worked really good, pretty much what I wanted. Also, I tried using a dual gang 250k pot to seperately shunt the highs to ground from the grids through caps, also sounded really good. But that was at low volume. Now I need to take it to the rehearsal and crank it and see how it really goes!!

[mikeboone]

And another option is to do it on the speaker side, like a crossover.

I use Eminence Red Fangs now and they are a nice dark speaker with great focus and will kill the shrill...

[anson]

So I configured my amp so that one p-p pair of the el-34 grids are shunted to ground through a 2.2n cap in series with a 250k pot configured as variable resistor, while the grids of the other pair of el-34's is shunted through ground through a seperate 2.2n cap and 250k pot. The two pots are dual ganged. I tested it out at rehearsal at cranked volumes, and it sounds real good, pretty much giving the effect I want. Gives me more control of some of the highs. Also I noticed that when the caps are shunted directly to ground (variable resistors set to 0 ohms), the high freqencies of courese drop off, but then the presence control tends to behave more like a "mid boost" rather than a "high boost" control. I like it.
I don't hear any audible oscilations, but I don't have a scope so can't really say what's happenning at higher frequencies.
Maby I'll change the 2.2n caps to 1n and see how that goes...