Presenting the 74W Tung Sten preamp

[Fancy Lime]

Hi folks,

I have had a few threads here recently pestering you all for advice on "amp-like" distortion, "noise-less" preamps and such. Finally, these discussions have borne fruit. May I present to you: the ₇₄W preamp for guitar and bass. In case you're wondering, that is the element symbol for Tungsten, not a wattage.




Design objectives
I was trying to emulate the typical "low gain amp at full volume" sound popular in stoner rock and doom metal. Notorious amp popular in that genre include the Orange OR120, Matamp GT120, Sunn Model-T, and Ampeg V4 or SVT, preferably from the 70's. All of these have a few things in common that set them apart from the Marshall/Fender crowd. Most obviously, they all have James Tonestacks, which means they do not duffer the permanent mid cut of Marshalls and Fenders. They also seem to all have rather more generously sized power supplies, giving them lots of dynamic and less sag. Lastly their preamps are not particularly gainy. If you want to get them to distort properly, you have to play them way open. That, in turn, means that a significant part of the distortion happens after the tone stack, in the phase splitter or poweramp (in some of these the tone stack is even before the first clipping stage in the preamp). It also means that it is popular to hit the front with a treble booster, lest it gets too gloomy. So these were the things I tried to emulate while keeping the design reasonably simple. More importantly than design complexity, I did not want the resultant device to have more controls than necessary and wanted the controls to be intuitive, which I often find tricky to balance with keeping the total number of controls low because it means that controls cannot be too interactive. I also wanted the thing to work from 9V to 18V single supplies, and not be excessively noisy.

Design rationale
Input buffer
Not strictly necessary but I wanted to have the option to add a buffered dry out when necessary. This also provides a convenient place to add channel switching, if one were to build this as part of a multi-channel preamp.

Main gain stage
Fairly standard stuff here, basically a gain stage with built-in treble booster. The Presence and Gain controls are interactive. Presence does nothing when Gain is at minimum, at higher Gain settings, it boosts highs with a corner frequency of around 3kHz. When gain is set higher than noon, the opamp slowly starts to clip and C5 starts cutting the highs at 7kHz. When gain is at max, the high cut starts at 700Hz, which can be partially counteracted with the Presence. If Presence is wide open at max Gain, you get quite some clipping from the opamp, which is rather mid-boosted. Opamp choice makes a difference here. I used the NJM2068 (aka JRC2068) because it gave me a bit more clean headroom and nicer clipping than a NE5532. A TLC2262 will probably sound nice too but will be more noisy. I would not bother to try a TL072, as it clips early and real nasty. I find the two controls work nicely together and are fairly intuitive despite the (on paper) complex frequency shaping.

Bass control
Essentially half a Baxandall tone stack. The bass control is placed before the main clipping stages, so that the sound can be tightened or boomified (that's not a word but you know what I mean) pre clipping. I find this an important aspect of tone shaping for the sound I was looking for. At the highest gains, bass boosting looses some efficiency but not as much as I feared originally. And how often do I want an extremely distorted sound to be very bassy? Exactly. For clean and slightly dirty sounds, bass boost effectiveness is not really influenced by the placement. Implementing this stage with a CMOS inverter makes it sound nice even when the stage itself clips. Yeah!

Clipping stages
I love the sound of CMOS-clipping in the morning. The first of the two stages has a Boost switch to go from gain=1 (well, -1 actually but here we only care about the absolute value) to open loop gain. I experimented with a second gain control pot here, but found a switch together with the existing Gain control in the gain stage completely sufficient. With Boost engaged, the distortion with Gain at about 9 o'clock picks up around the same point where it left off with Boost disengaged and gain at 3 o'clock. Wehn building this (currently it is still on the bread board), I will make the Gain pot push-pull and activate the boost with the Gain pulled. The second clipping stage has gain=1. I like having a "gainless" CMOS stage behind the gainy clipping stage because it rounds off the clipping some more due to the non-linearity of the inverters. At a gain=0, the second stage is driven into non-linearity exactly when the first one is as well, which essentially just doubles the effect. That makes for a nice smooth transition from clean to distorted. I experimented with more stages but for anything above 2 total stages I cannot hear any more difference.

Treble control
Now this one was tricky. This is the second half of the Baxandall. I definitely wanted to have the Treble after the clipping stage, else it would loose too much efficiency for high cuts on highly distorted sounds. But this stage also needed to bring the overall level down by about 6dB so itself and the following stages would not clip. I could not find a good design tool for this (as most people seem to want their baxandall to have a out-of-band gain=1), so I wrote one myself. Here is the simulated response, x-axis in Hz, y-axis in dB (labeling of the pot setting is reversed and the simulation ignores C19, sorry):


Speaker sim / output buffer
Just a 2nd order Sallen-Key bandpass with near-Butterworth characteristic implemented with the leftover half of the NJM2068. Not strictly necessary as all preceding stages already limit the bandwidth to ~20Hz-7kHz. But having this here allows easy later changes to tighten the bass or make it resonant (to simulate an open or a vented cab) by changing the ratio of R22 to R23, or make it more or less present by changing C22/C23.

The rest should be self explanatory, I hope. If not, please ask. I must say, I am really happy with how this turned out. Really nice overall sound when played straight into a neutral headphone amp as well as when fed into a bass amp. Very nice break up at the verge of clipping with guitar as well as bass. Great overdrive sound, again with bass as well as guitar. And it definitely captures the stoner/doom/sludge sound I was after. Hurray! It also sounds very good for bluesy and hard-rocky stuff. I'll try to make some sound samples but I don't have a working recording interface at the moment and I'll be super busy for the next weeks, so this may be a while. Maybe I can manage to draw up a perf or vero schematic but I cannot promise that that will be any time soon either. If anyone else wants to build this thing and add some sound samples here, I'll promise my eternal gratitude

EDIT: A few words on the name: I wanted to keep the unofficial naming conventions for amps, namely 1-3 letters and some numbers. A chemical symbol seemed to fit the bill. And since this was supposed to be aimed at the heavy / stoner rock crowd, the symbol for tungsten seemed most appropriate. "Tung sten" is Swedish for "heavy stone" or "heavy rock". So the name of the preamp should be written "₇₄W" but spoken "tungsten".

Cheers and keep on rockin',
Andy

[garcho]

EDIT: A few words on the name: I wanted to keep the unofficial naming conventions for amps, namely 1-3 letters and some numbers. A chemical symbol seemed to fit the bill. And since this was supposed to be aimed at the heavy / stoner rock crowd, the symbol for tungsten seemed most appropriate. "Tung sten" is Swedish for "heavy stone" or "heavy rock". So the name of the preamp should be written "74W" but spoken "tungsten".

Nice work, nice schematic, nice plot, and nice post, well done. But seriously, calling it "74W" and expecting anyone to think anything but 74 Watts is a Quixotic, uphill battle. Just my 2 öre

[Fancy Lime]

...is a Quixotic, uphill battle.

But I like those.

Andy

[EBK]

But seriously, calling it "74W" and expecting anyone to think anything but 74 Watts is a Quixotic, uphill battle. Just my 2 öre

I worried that I might have reflexively punched him in the face if I were in the same room as him during that name reveal. I would've felt terrible and would have probably apologized non-stop for several days, of course--I'm really a very nice person, honestly! 

I do like the name now, though.   

Anyway, Andy, I'm glad you designed this, I'm glad it sounds good, and I'm glad no one got hurt, but just to be safe, please promise me that you'll never label a power jack ₂₃V.

[Fancy Lime]

Hey, I have been dealing with enough students trying to solve equations by confusing units with variable names. But I am open to renaming it the "Tung Sten". Or maybe "Tongue Stan"?

BTW, is "Quixotic" taken as a name for a booster? Or is that a bit too on the nose?

Andy

EDIT: There, see what you made me do? I changed the title of the original post...

[EBK]

No, don't rename it.  Like I said, I actually do like the name.

[Fancy Lime]

Apropos of nothing:

Dibs on the "₄₂Mo Fullrange Booster", the "₇₇Ir Treble Booster", the "₂He Octave" (unless that one already exists. I have a hunch it might), and the "SF6 Suboctaver".

Pedal name dibs is a thing, right? No? Damn!

Andy

[Fancy Lime]

No, don't rename it.  Like I said, I actually do like the name.

No worries, it's just a second name. The "74W" stays.

[EBK]

I'd like to see you make a germanium fuzz and call it ₂₃Es for экасилицій (ekasilitsiy)/ekasilicon.

[EBK]

Need to add one more comment here:

I really like your schematic style. 

[soggybag]

This looks pretty neat, and if it gets into the sounds of stoner doom I'm there with you!

I'm not getting the purpose of the two LEDs on the input, what"s going on here? From my amateur perspective if the input is more than the 2v drop of the diodes it clips the input. If I'm correct in this assumption is that useful?

Is there another effect somewhere that uses this? I'd like to hear it.

[Fancy Lime]

Need to add one more comment here:

I really like your schematic style. 

Thanks! I was not aware my schematics had a distinct style. In case someone is interested, I use the "gEDA Schematic Editor". That takes care of most of the look. Then there is just a bit of prettying up in Inkscape.

Andy

[Fancy Lime]

This looks pretty neat, and if it gets into the sounds of stoner doom I'm there with you!

I'm not getting the purpose of the two LEDs on the input, what"s going on here? From my amateur perspective if the input is more than the 2v drop of the diodes it clips the input. If I'm correct in this assumption is that useful?

Is there another effect somewhere that uses this? I'd like to hear it.

Hi Mitchell,

thanks! Yeah, these diodes are not strictly necessary. All they do most of the time is protect the JFET from excessive voltage spikes that should not occur at the input under normal circumstances anyway. The only situation, where they actually do matter, is if you hit the front end of the preamp with a booster that delivers more than +- 1.7V or so. In that case, you get additional clipping from the diodes and still get a fairly predictable results from the Gain and Presence pots without excessive clipping from the opamp. If the preamp is run at 18V, one might want to change these diodes to blue LEDs, or a pair of back-to-back Zeners with about 5V.

Andy

[EBK]

Need to add one more comment here:

I really like your schematic style. 

Thanks! I was not aware my schematics had a distinct style. In case someone is interested, I use the "gEDA Schematic Editor". That takes care of most of the look. Then there is just a bit of prettying up in Inkscape.

Andy

It's mostly the neatly labeled boxes and the overall flow/width, but the font choice is cool too.

[vigilante397]

So... can we hear it?

[Ben N]

Very nice! I love the approach, the modular schematic, the use of CMOS, and the split eq. I also suspect that this might be a an effective tool outside of the intended sonic territory.

[bool]

Wrong name.

You should rename it to "74W Tung Stone" preamp.
Spelled "Tung", not "Dung", he he he...

[duck_arse]

...is a Quixotic, uphill battle.

But I like those.

Andy

the Quix-exotic would have a tilt tone control? been dibs'ed yet?

[Fancy Lime]

I'd like to see you make a germanium fuzz and call it ₂₃Es for экасилицій (ekasilitsiy)/ekasilicon.

I have never bothered with Ge, so far. But with that name, it is, admittedly tempting. Especially because a fuzz is exactly what should go in front of the 74W.

Cheers,
Andy

[Fancy Lime]

Need to add one more comment here:

I really like your schematic style. 

Thanks! I was not aware my schematics had a distinct style. In case someone is interested, I use the "gEDA Schematic Editor". That takes care of most of the look. Then there is just a bit of prettying up in Inkscape.

Andy

It's mostly the neatly labeled boxes and the overall flow/width, but the font choice is cool too.

The organization in boxes is something I stole from Electrosmash. I always found that extremely useful for discussions about what does what. The "overall flow", as you call it, is also something I like in the Electrosmash schematics, so I guess mine resemble that a bit as well. I want to explain as clearly as I can to anyone who is interested what part does what and why it is designed like that and not some other way. Other people doing the same thing is what taught primarily me this whole business, and still does. I would like to contribute to educating the next generation with the same method that worked well for my own learning and Electrosmash makes a damn good job of that for many classic circuits. So I just shamelessly copy their method. 

The font is called Gismonda FG. It is free and can be downloaded here:
https://www.dafont.com/gismonda-fg.font

Cheers,
Andy