Gain

[wizard333]

I said you couldn't make a Twin/whatever sound like a SLO by any other means than copying the SLO circuit.

If you mean sound exactly like a SLO, maybe; but you can't make X sound like Y unless you copy Y exactly in in any case. You CAN get that kind if gain from 3 stages, WITHOUT using PI or power amp clipping, and have it sound quite good.

the number of gain stages is a very good indicator of how much distortion one can get from the amp (and amp alone). Beat the strings as hard as you like, you won't make a Twin/Deluxe/whatever sound like a SLO.

Your statement was about distortion, not tonal character. Set a SLO up to play clean, and you can beat the strings all you want and it won't sound like a Twin. That has nothing to do with distortion. And?

There is also a level of distortion that becomes un-usable squeeling, hyper-compressed fuzz. You can get as much usable distortion from a 3 stage design, have it sound good, and even push it well into"too much" territory.

I see that theory quoted all the time though, that somehow you "can't" get searing high gain from 3 stages, usually bandied about when discussing Marshall amps. It's hogwash though.

[Chuck H]

Where did you get that notion from? Lots of preamp distortion + PI and power amp distortion = a muddy, uncontrollable mess. There's a reason why the master volume is positioned between the preamp/tone stack and the PI/power amp, and these amps are usually 50/100 watts output. It's because more/less, *all* of the distortion is intended to come from the preamp.

That response was in reference to getting modern sensitivity levels from amps with only three preamp gain stages. I've done "ok" with only three stages. But to get the sheer sensitivity and harmonic content as a modern, uber gain, five cascade stage am from an amp with only three preamp stages (and have it sound good) my experience has been that you need to clip the power tubes and PI. Results may vary.

[Chuck H]

You can get as much usable distortion from a 3 stage design, have it sound good, and even push it well into"too much" territory.

I see that theory quoted all the time though, that somehow you "can't" get searing high gain from 3 stages, usually bandied about when discussing Marshall amps. It's hogwash though.

You seem to have found the secret recipe. I've certainly tried to get the modern sound from three stages with limited success. I think the Fender Prosonic is as close as anyone has ever gotten to making a three stage preamp that can do modern tones well. And for modern tones it's not quite as good as an SLO. But that's just MHO. Unless your planning on patenting or marketing this design, would you mind sharing?

[exclamationmark]

Honestly, you can get some pretty insane distortion levels out of a 3 stage pre-amp but you have make some sacrifices or employ dirty tricks like lowering the B+ of the last stage. I've gotten comparable amounts of of distortion by fully byassing every stage and biasing most of the stages hot in a 3 stage circuit, but it (IMO) doesn't sound great for high-gain oriented genres of music. Most 4 stage pre-amps dump some of their gain, which says something about how much distortion you can actually get from a 3 stage circuit.

Counter-intuitively, more stages can also give you a more dynamic sort of distortion. Each stage you add can potentially modify the duty cycle of the signal, depending on how they're biased and coupled together - I've had pre-amps that start off clipping by having a low duty cycle square wave output, only to have it grow really fat at moderate drive levels, and then somehow shrink again at even higher drive levels. Combined with the fact that you can dump some of the signal and clip the last stages more gracefully, it makes 4 stage amps an attractive option for 'high gain' amplifiers.

[cyclone]

Hey guys, I lurk and learn from all of you, appreciate all the help over the many years. Nobody has mentioned interstage attenuation or signal bounding, you can only put so much voltage on a preamp tube grid without an ass sounding square wave coming out of it. I get tons of gain and metal crunch with definition that cleans up with the guitar volume knob with three stages, four if you count the Vox type PI with a gain of 30 Carefully managing the interstage gain with a cold clipping third stage, then some attenuation before the PI makes all the difference. The point at which the PI signal output starts to clip (at an appropriate voltage level for the power tube grid of choice) vs the power stage is the key. Once you get the interstage coupling attenuation values to not drive stages too hard, you can adjust preamp and PI plate voltages to get different distortion " flavors" . I can't tell the difference between brands (or colors) of caps or even preamp tube brands, really, but I can damn sure hear the difference in sound and feel between when I change preamp plate voltages by 10v once I've got the gain structure where I like it. And speakers. Start with a cab or speaker that you like and design backwards. I haven't built any amps with power amp NFB yet, but I do use a PPI MV and a cut control, which is definitely variable NFB. One more recommendation; buy an O scope, you can get a tektronix on ebay for 100$ If you don't have one, you are a hack, sorry.

[Chuck H]

That's great. And My MO isn't dissimilar. But for the purposes of this discussion the PI is strictly off limits. The PI can't be clipping. Neither can the power tubes. We have three preamp stages to work with. That's it. If the circuit will support an EQ, as it must to do modern high gain tones, then the EQ must be at the end of the chain followed by a master volume. A cold clipping third stage would make that very difficult. Including the PI in the clipping chain makes three clipping stages into four stages. So that's disqualified. As would be the analogous power tube clipping that should be evident if the PI is already driven into clipping.

...Unless we use a post PI MV. But then it's still not just three clipping stages.

I want to see the three preamp stage amp schem that does modern gain tones WITHOUT clipping the PI or power tubes.

[Enzo]

Good to decide what gain means. Among technicians we can discuss it. In the world of kids with amps, gain is just a generic term for distortion. I have had kids come to me with as PV 5150 asking me to give it "more gain." Really? More gain than a 5150?


I see minimalaist6s counting tubes. Sure you can get large amounts of gain from few stages. The thing is with multiple stages and level reduction between stages, you can overdrive a stage, that tends to be asymmetrical. Each stage inverts, so if we clip a stage then knock the level back down, we can then take the asymmetric signal and run it through another stage which then clips the other way. That is interesting in that we now have evened the sides up a bit.

You may or may not like it, but it is different from the sound of fewer stages.

[wizard333]

would you mind sharing?

I do market it, so yes unfortunately I'd mind sharing. But there is no rocket science. I don't use a cold stage. I don't goof around with lowering the plate voltage of any stage. I don't rely on power section clipping. I wouldn't mind sharing a sound clip, or a number of them since I don't do one trick ponies in a way where 1 clip goes "here's what the amp sounds like". I like amps that you can take to a Metallica cover gig or a Larry Carlton cover gig, same amp. I will say that you can't do that with a standard tone stack, or without a couple push/pulls.

A few things I've found: You don't have to use a cathode follower to sound like you are. You don't have to alternately clip to get asymmetrical clipping as Enzo describes, although with more stages that becomes more necessary as you need to throw away more gain. You also can't put bunny ears on a duck and expect it to be a bunny, which I've seen a lot of people do.

I'll see if I can get some clips up soon.

[Steve Conner]

I've built a couple of tube preamps with switches to select the number of gain stages in use. 2 is fairly clean, 3 breaks up and crunches, and for the modern high-gain tone I use 4.

Maybe you can do the high-gain tone with 3, but the extra stage gives the freedom to dump a bit of gain here and there for tone shaping.

[cyclone]

Dang Chuck, who says the PI doesn't count? Voltage gain and tone shaping takes place there, even if its a gain of 1.

More stages, say one more, three versus two sounds cleaner through the dynamic voltage range of a guitar players output. My two stage amps sound rawer than my identical amps with a cold stage, the three stages sound more liquid. The two stage amps have more dynamics and more spank, not quite as much squeeze. Better for clean players, heavy string, hard pickers. People think the two stage amps have more "gain" but they don't.

I love you Chuck, you are a hardass seat of the pants guy like me and I respect that. Build it and see. I've learned a ton of stuff from you from your work and posts over the years, thanks.

[Enzo]

You don't have to alternately clip to get asymmetrical clipping as Enzo describes, although with more stages that becomes more necessary as you need to throw away more gain.

Actually that is not what I said. You can use alternate stages to make asymmetric clipping more symmetrical. You may or may not want to do that, but it was not a means to asymmetry.

[Steve Conner]

I agree. It is possible to squeeze the same gain out of, say, 2 stages as a different design would get from 3. I think the design with fewer stages will always sound more dynamic and "rawer".

The transition into clipping is more sudden, and the duty cycle modulation from grid current is greater, because you have fewer stages each overdriven by a greater amount. If you have lots of stages, and only overdrive each one a little, you get that spongy, liquid, "sustains for ever" sound that was so popular in the 80s. If you're not careful, you also get a lot of noise.

In metal amps, it is common to do some pretty aggressive EQ immediately before the power amp. The archetype is the Marshall with its tone stack placed immediately before the PI, giving a hefty midrange scoop even with all knobs set to 5. Mesa took it further with a graphic EQ.

If you do this, you can't use the PI and power amp as part of your distortion-generating scheme, because the distortion will undo the EQ, leading to what Mike called a "muddy uncontrollable mess".

[defaced]

If you mean sound exactly like a SLO, maybe; but you can't make X sound like Y unless you copy Y exactly in in any case. You CAN get that kind if gain from 3 stages, WITHOUT using PI or power amp clipping, and have it sound quite good.



Your statement was about distortion, not tonal character. Set a SLO up to play clean, and you can beat the strings all you want and it won't sound like a Twin. That has nothing to do with distortion. And?

There is also a level of distortion that becomes un-usable squeeling, hyper-compressed fuzz. You can get as much usable distortion from a 3 stage design, have it sound good, and even push it well into"too much" territory.

I see that theory quoted all the time though, that somehow you "can't" get searing high gain from 3 stages, usually bandied about when discussing Marshall amps. It's hogwash though.

Let's do this again:

"that somehow you "can't" get searing high gain from 3 stages"

I did not say that. I said:

"the number of gain stages is a very good indicator of how much distortion one can get from the amp (and amp alone)." I fully understand that there are different roads that lead to mecca, however unless I want to spend copious amounts of time writing a novel on the topic (and in the process writing about topics I know nothing about, like trying to get 4 gain stage like distortion out of 3 stages in a tube amp), some level of containment of the subject and generalizations need to be made. That "very good indicator" qualifier falls under the heading of generalizing.

Also, this year was my first opportunity to play a Plexi and a JCM800 (I'm a new school metal head, those amps aren't at the top of my list of amps to own/build), so I'm going to go out on a limb and say I probably don't fall into the category of people spewing BS about the capabilities of Marshall amps.

I'm also all too familiar with the "too much gain" problem. I have to finish boxing up my SLO preamp so I can go about getting this project under control (which incidentally, sounds pretty modern with only 3 stages, but they're MOSFETs so I didn't think they'd be worth mentioning, but with 4 stages sounds very farty and blaty - too much gain): Turkey Day SHOOTOUT Competition

Clips of your amp would be cool though.

[defaced]

That response was in reference to getting modern sensitivity levels from amps with only three preamp gain stages. I've done "ok" with only three stages. But to get the sheer sensitivity and harmonic content as a modern, uber gain, five cascade stage am from an amp with only three preamp stages (and have it sound good) my experience has been that you need to clip the power tubes and PI. Results may vary.

Ah, gotcha. Sorry about the confusion.

[R.G.]

@trem:
In the more general real world, outside tube guitar amps, gain is not equal to distortion. Gain is one means to get distortion. There are other ways.

Substituting "gain" for "distortion" is a linguistic distortion, much like how we got the word "curry". If you limit yourself to talking about only standard-format tube guitar amps, there is some reason to equate the two, and to think in terms of the number of tube stages as being related. But they're not really the same thing.

As an example, RF front ends may have gains in the 100db range. A single NPN transistor may be set up in circuits to give gains on the order of 2000. That much gain would, if not limited somehow, produce signals of upwards of 200V peak.

I suspect that guitarists go to "gain" as the word they use to tell techies what they want, confusing what they want (distortion and "sustain") with how they get it. In my view, the previous discussion brings a lot of this out.

If you limit yourself to getting distortion with 12AX7 triodes used as amplifier stages, then yes, the discussion of how to get more distortion ("gain") is plausible. If you want that to be controllable, you need to limit the frequency responses at the bottom/bass and top/treble to avoid massive gating of the treble by bass components ("mud"), shrieking resonances in the treble. For enough gain, you have to start taking into account the response spectrum of the speaker, enclosure and sometimes the room, as these can all make for a one-note feedback as gain gets really high.

The frequency and phase response of the entire guitar->amp->gain->distortion->speaker->guitar loop leads to some frequencies being suppressed, some enhanced. Without a way to measure what's going on, the discussions inevitably descend into anecdotes.

In fact, there are several almost-orthogonal things going on, and they all contribute to the final sound. EQ before distortion sets what frequencies get distorted most, EQ after distortion determines what harmonic products gets emphasized or reduced. Loop resonances give "voicing" and feedback sensitivities.

In a way, saying "gain" when what's being talked about is the qualities of distortion is a very human thing to do in the face of a vocabulary that doesn't have enough descriptive words.