Preamp distortion - faster transition from clean to distorted

[NSB_Chris]

For fun, I have built several different amp circuits from tweed Fender to Gibson to Marshall etc. but always return to JMP2204/JCM800 territory. I just love it. Like many of us mortals, we live in a dB level that does not allow for proper power stage yumminess and have to rely on the preamp. It is good the way it is, but I can't help tweaking!

I am interested in thoughts on sharpening the transition from clean to dirty. A lot of what I read on that comes from the power section with negative feedback etc. (something I will never get). On high gain circuits like SLO, they drive the preamp hard at the beginning with a stage that has a high plate voltages with 220k or so plate resistor. Does this add to the sharpness of the transition from clean-to-dirty because the driving stage has a large gain range? I did build a version of Rob's

I am not trying to build a super high gain amp, just achieve a little sharper transition in the preamp. Hoping to get some discussion. I will then either build a new am or modify one of the 2204 versions I have now. I have been pouring over Rob's website pages and google searches but would benefit from some infusion of experience!

 

[2L man]

Now when you know how different amps sound different input their stage values to this loadline calculator and you "see" how signal behave. Then input headroom and green anode sweep and estimated 2nd and 3rd distortions appear and it is easier to see is anode sweep linear=clean or unlinear=distorted and sometimes clipped like "cold clipper" stage.

Universal Loadline Calculator

This calculator allows you to trace the anode characteristics graph, and to compute and tune loadline and operating point of various vacuum tubes.

www.vtadiy.com

 

[owlexifry]

On high gain circuits like SLO, they drive the preamp hard at the beginning with a stage that has a high plate voltages with 220k or so plate resistor. Does this add to the sharpness of the transition from clean-to-dirty because the driving stage has a large gain range?

absolutely recommend trying this exact mod first (220K plate resistor on 1st stage). it's the only one i really condone.

i have gone through quite a through trials and tribulations with the 2204 circuit lol. I'm a big fan of it. also a huge fan of the SLO preamp, after doing rob's SLO40 6L6.

i've tried all sorts of mods, including values similar to the ceriatone chupacabra and yeti (jose arrendondo type mods i guess). wasn't a huge fan.
i have found the more i drifted from the stock values, the less pleasant and harsher it became, no matter much i tried to snub out harsh high frequencies with all sorts of methods (V1 plate snubber, PI snubber, NFB snubber etc.).
i feel the stock circuit is just about pretty much bang on.
sometimes all it needs is a good goosing on the front end with the right overdrive.

- adding a cathode bypass cap (0.68uf/1uf) to the third stage (V2a? 820R cathode resistor) - is nice a gain boost.

- you could also try the 'jose clipper/MV mod', adds a lot of saturation and compression, some love it, and it also cuts the output volume massively.
(ive done this mod ...but i just prefer to blast the front end with an overdrive)

(this inserts between cathode follower and tonestack - MV is moved in front of the tonestack)

- many people talk about clipping that 1000pf/1nf bright cap on the preamp/gain pot. in my opinion this 'bright' cap is critical to achieving a crisp overdrive tone. sounds muddy without it.

- increasing NFB is a good move. tightens up the drive and makes it punchier. i would recommend trying some different NFB values (on my 2204 i have a variable NFB pot)

- if you haven't done a depth / resonance control mod, do it.

- also this cathode follower bootstrap mod looks quite nifty and i really wanna try it out:

-

 

[2L man]

Perhaps you should test clipping diodes like Marshall Silver Jubilee and few other amps use. Or use LEDs or combination of both and silicon diode clip other half more abrupt and LED less abrupt. Perhaps 3 or 4 silicon diodes in series are needed so that LED begin softer clipping bit earlier and then silicon diodes clip comes after but harder.

 

[NSB_Chris]

absolutely recommend trying this exact mod first (220K plate resistor on 1st stage). it's the only one i really condone.

i have gone through quite a through trials and tribulations with the 2204 circuit lol. I'm a big fan of it. also a huge fan of the SLO preamp, after doing rob's SLO40 6L6.

Thanks. I may try 220k plate on one of my existing 2204 amps first. Several have two nearly identical channels, so I can test the effect.

 

[NSB_Chris]

Perhaps you should test clipping diodes like Marshall Silver Jubilee and few other amps use. Or use LEDs or combination of both and silicon diode clip other half more abrupt and LED less abrupt. Perhaps 3 or 4 silicon diodes in series are needed so that LED begin softer clipping bit earlier and then silicon diodes clip comes after but harder.

Never heard a Silver Jubilee. I would be concerned that the diode clipping would be too "fizzy". For that effect, maybe pedals in front would do the trick.

 

[owlexifry]

Thanks. I may try 220k plate on one of my existing 2204 amps first. Several have two nearly identical channels, so I can test the effect.

you'll love it.
(also make sure it's at least 1W for least noise)

 

[mrriggs]

If you want clipping without the fizz you can use tube diodes.

Better yet, add local negative feedback to it to get that sharper output stage transition you're looking for.

http://www.valvewizard.co.uk/localfeedback.html

I'd start with a closed loop gain about half of the open loop gain. Pick up the signal from the clipping diodes and feed it back to the grid with the appropriate feedback resistor. You'd need another cap in line since the clipping diodes are at a DC potential.

 

[2L man]

I briefly did test a balance adjust for Phase Inverter outputs and it has potential to even better clean the amp output because the use of typical 82k and 100k LTP anode resistors are close but not always perfectly optimized. When making power tube drives intentionally different increase sweet sounding 2nd harmonic distortion to whole frequency range and louder played output lows can make OT saturate which can sound interesting as well.

On cathodyne PI typical 56k resistor can be replaced using 47k resistor and 10k potentiometer.

Using potentiometer on current path is not recommended although wiper failure won't completely kill the sound so this is still an idea if someone could sophisticate it? I recall seeing PI balance on HiFi amps so thats where I will look correct way to make it next time.

 

[Lynxtrap]

Thanks. I may try 220k plate on one of my existing 2204 amps first. Several have two nearly identical channels, so I can test the effect.

220K on the plate of the first triode will give you a bit more gain out of it. Triodes downstreams might go into overdrive at lower gain pot settings, but I can't see how it would change the way they transition between clean and overdrive.

 

[CirrusBand]

Just literally throwing this out there with the bare minimum of understanding...

Negative feedback seems like it causes a quicker change from clean to distorted, it's said that lack of negative feedback makes that a smoother transition...

What if you had local negative feedback around a preamp stage, but diode clamped it so the negative feedback signal couldn't go above a certain level? So it stays clean until a certain level, then breaks down quicker. Like I say, just throwing it out there. Might be no better than diode clipping the output of that preamp stage, might have terrible unforseen consequences.

 

[NSB_Chris]

If you want clipping without the fizz you can use tube diodes.

I remembered seeing that but it took me some time to find it:

Hollow-State Hard-Clipper

I've got boxes of old TV-tubes and find it fun to come up with uses for them. There's like two dozen 6AL5 dual-diodes in there which got me thinking about building a hard-clipping distortion stage using tube diodes. I've also got this old Bogen PA amp chassis that I picked up at a flea market...

www.tdpri.com

Searching with "diode" and "distortion" search results are dominated by pedal design posts.
That was a killer amp experiment thread and I remember it took a while to get my head around all the things you were doing in that! Some time I will try it.

 

[radiocaster]

Although I'm not really sure what you're talking about and what you want, I think you may also be interested in cathode bypass caps.

 

[dan40]

I would be concerned that the diode clipping would be too "fizzy". For that effect, maybe pedals in front would do the trick.

It actually sounds pretty good but the diodes do add a bit of compression. It's the same type of clipping that many OD pedals use. You can use lower voltage diodes (12v) for a massive volume drop and compression but I normally use 20-24v diodes because the volume drop isn't as extreme and the clipping sounds a bit more open with less compression. Transistors can also be used in the same way. Friedman amps use the MPSA06 transistors in place of the diodes because they offer a slightly different sound. They are very easy to solder onto your MV pot for a quick listen if you ever decide to try them. Adding a switch to take them in and out of the circuit would be even better.

 

[andrewRneumann]

I agree with those that mentioned local NFB as a way to sculpt the transition to overdrive. The local NFB will be harder to overdrive, but when it does, it should clip harder.

I have two other suggestions that might be worth your while. For stages that are biased warm and clip the top of the input signal by grid-current, try a large value grid stopper. The idea being that you want to lose voltage as soon as the first grid current electron flows.

On the converse, for stages that are cold biased and clip the tops of the output signal, my suggestion would be to reduce the plate resistor value. This steepens the load line and should theoretically speed the transition to cut-off. You would probably need to adjust the cathode resistor to keep the bias properly cold.

Good luck, hope it helps.