Marshall Mods - What are your Favourite Mods?

[Helmholtz]

I probably wouldn't change the power amp.

So no PPIMV - or grid signal clamping to avoid grid conduction here?

[Helmholtz]

It is totally irrelevant what a tube was developed for or where it was first used.

No tube was ever developed for nice distortion in a guitar amp.

"Audio tube" just means that it's well suited for low distortion sound reproduction at audio frequencies- as opposed to RF applications.

It's all in the tube data.

[pdf64]

Regarding the claim in the now deleted post, my understanding was that the Soviet Union purchased valve production equipment and training from RCA pre WWII. Hence the similarities in types. https://en.m.wikipedia.org/wiki/Russ...e_designations

[Chuck H]

So no PPIMV - or grid signal clamping to avoid grid conduction here?

No on the grid clamping because... The tube I was tethered to for a long time was the el84. I had a whole bunch of them so that's what I built my amps with. Crossover distortion and phase error with these tubes sounds less like 'Bwah' (note bloom/swirl?) but more like 'Bwizzz' (can of bees). El34's and other big bottles behave more pleasantly to my ears WRT this particular effect and I don't see the need to mitigate it. Maybe the el84's don't do it as well because they have a higher mu? I don't really know, but crossover notch and phase error in a hard driven PP el84 amp just sounds almost staticky and broken. Don't get me wrong. I've learned to love el84's for their own merits and intent to continue focusing on designs that include them.

The phase error thing I mention is an overlooked part of the tone of some amps. Depending on design it's more or less pronounced with clipping that there's a change in wave form symmetry between attack and lower signal decay shifting as being pronounced from one side to the other. I believe this to be an element of the bloom and swirl that some Marshalls are credited with. More modern, high gain amps mitigate a lot of this effect for a flatter, more solid/stiff kind of attack and decay. I like it both ways (depending) so how I design a particular amp just depends on the specific goal for that project. I usually shoot for good clipped wave for symmetry with an input of 100mV. This allows the overall tone to 'still sound like your guitar' and then any differences between extreme high signal and extreme low signal amounts to a useful musical dynamic. I find that if there's too much asymmetry at 100mV input the clipped tone sounds a little like a cocked wah pedal and you lose the character of the instrument plugged in. And...

I just never liked the sound of PPIMV's. I like a good high gain preamp and I like power tube clipping. Any time I've experimented with PPIMV's and I have to hear the PI cold end clipping I've never liked it much. So for my part I think an amp should make it's distortion in the preamp and then maybe clip the power tubes a little for that refined sound OR include the power tubes in the clipping circuits so we can hear what THEY do. Much of this opinion has to do with amps that use NFB and the oddities that happen when placing any control within that loop. I also don't care much for the VOX high cut circuit or the Tiny Terror tone control for the same reasons. They just don't seem to perform with predictable and gratifying results. The PPIMV falls into the same category for me. Really ugly wave forms and they sound like it too. Since it's all subjective and PPIMV's are popular I'm not going to say anything like it's "wrong". But I will say it's not my preference.

[Chuck H]

No tube was ever developed for nice distortion in a guitar amp.

This^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

So just by designing/refining for this end game we're moving into a black art. Alchemy sometimes. And it might be worth noting that what we (guitarists and their audiences) have learned to like may have something to do with conditioning. The two things, what we like to hear and what we design for, have evolved together. And it started with small tube amps being cranked up to distortion levels to get the necessary volume for larger venues. And the genre where this was accepted was popular music. Eventually leading to the highly distorted guitar sounds in much of todays "popular music". Which, much to the chagrin of head bangers everywhere, includes metal, grunge, indie, etc. In fact the waning popularity of these genres in favor of hip hop, house, 'pop', alt, etc. and the lack of heavy guitar bias in the latter is a concern for me. Rock will never die, of course. But let's face it guys, it IS getting older. And kids these days...Ahhh.

[Greg_L]

Well I don't know that anyone is expressing "opinions". What I see is observations and facts. My own experience with 6550's is similar enough to yours that I wouldn't say we're in disagreement. All I pointed out was that the 6550 tube, as implemented by Marshall, probably never got a fair shake as an overdrive tube because of circuit incompatibilities. I think that the overwhelming opinion that the el34's sound better in Marshall amps is correct (with metal tones perhaps an exception?). But that doesn't mean 6550's couldn't be made to sound good also in a circuit more idealized for that goal.

EDIT: Well I just spent some time looking at the data sheets and what I see is that the way 6550's were implemented in the Marshall amps they were simply not driven hard enough to clip as much as el34's do in those amps. So if a few changes had been made in that regard maybe guitarists would have a different opinion about the 6550's

I have a 1973 JMP Super Lead with 6550s. It was a model slated for the US, so it came that way. I bought it with a half-hearted conversion to EL34s. I put it back to 6550s with the proper bias and bias resistor mods to run 6550s and it sounds freaking fantastic.

[Helmholtz]

The phase error thing I mention is an overlooked part of the tone of some amps...
.

First of all thanks a lot for your comprehensive reply and valuable insights.
You largely confirm my own thoughts and observations.

I think with "phase error" you mean the increasing duty cycle asymmetry of the grid signals at large drive.
Zollner discusses reasons and consequences in his book.
Apart from causing a kind of crossover distortion, it can be expected to generate low frequency artefacts and IM effects (including interaction with the OT).
Interesting that you think it might be beneficial to the "Marshall sound".

Agree on PPIMVs here as stated in other threads.

[Greg_L]

I love PPIMVs on NMV amps. I think they sound good if used properly and make old school fire breathing Marshalls usable on modern stages. I personally prefer a PPIMV to an attenuator, and if you can use both, well that's good too. A little PPIMV, a little attenuator, virtually unnoticeable. I'm sure we'd all love to be able to crank 100w stacks everywhere, but small club gigs don't allow it.

[Helmholtz]

I love PPIMVs on NMV amps.

Interesting.

Did you compare it to a pre-PI MV?

Did you try a good "reactive" attenuator?

[Greg_L]

Interesting.

Did you compare it to a pre-PI MV?

Did you try a good "reactive" attenuator?

Post-PIMV is definitely better than Pre-PIMV to me on the "vintage" Marshall and Fender circuits. Those preamps don't make enough gain for a Pre-PIMV MV to be useful for rock and roll. Try a Pre-PIMV on, say, an otherwise stock Marshall 1987 circuit. It's garbage.

I have a 1987 that can be a 2204 or a 1987 with a push/pull pot and it has both Pre and Post PIMVs. When in 1987 mode the Pre PIMV definitely hurts the drive and sound of the amp.

I've tried reactive attenuators. And resistive. My current attenuator is kind of a combination of both. They're ok. Everything has compromises. I prefer the PPIMV for my kind of stuff though. When I'm just dinking around at home I don't obsess over sound that much. I accept that playing huge amps at bedroom volumes has tone limitations. But when I'm recording or playing live, sound matters, and I prefer the PPIMV option over any other volume-lowering alternative. That's just me.

[Helmholtz]

Thanks for your contribution!

I might add that a "reactive" attenuator always should provide a combination of reactive and resistive load impedance just like a real speaker.

[Greg_L]

Thanks for your contribution!

I might add that a "reactive" attenuator always should provide a combination of reactive and resistive load impedance just like a real speaker.

My current attenuator is a Weber Mass 200. It has a small speaker assembly inside of it, minus the cone. Magnet, coil, etc. It also has a massive rheostat. The dummy speaker supposedly gives the amp something familiar to work with while the rheostat eats the wattage and turns everything down. It sounds pretty good for just knocking a little rage off the top of the output. It's transparent enough to use for some volume attenuation. But like any of them, it gets a little ugly when turning a 100w amp down to bedroom levels. And hot. That's where if I'm really being picky about things I'll use the attenuator in conjunction with the PPIMV. You can get the best of both while minimizing the worst of both.

[loudthud]

The PPIMV does one thing I'd like to explore. As you turn it down from max gain, it's lowering the gain inside the feedback loop of the power amp and raising the output impedance as seen by the speaker. Not all the way down to bedroom levels, just the first 6dB or so of volume lowering where the Presence control still has some effect. Has anyone noticed a change in tone, good or bad in amps that have both a PPIMV and MV at the output of a high gain preamp ?

[Chuck H]

I think with "phase error" you mean the increasing duty cycle asymmetry of the grid signals at large drive.
Zollner discusses reasons and consequences in his book.
Apart from causing a kind of crossover distortion, it can be expected to generate low frequency artefacts and IM effects (including interaction with the OT).
Interesting that you think it might be beneficial to the "Marshall sound".

Thanks. Yes, that's what I'm talking about. Not being schooled in electronics I thought it might be a consequence of phase changes when the impedance drops at an input grid associated with clipping. I should read Zollner's assessment, though I can't count on understanding all of it. Anyway... Vintage Marshall designs do it A LOT. I think it's because the preamps make a lot of gain between triodes when cranked up and there are no grid stops. I've noticed that grid stops can mitigate the affect sometimes. And if you look at more modern, high gain preamps you'll notice that it isn't just more gain stages. They also don't push as much between the cascaded stages and usually employ large grid stop resistors. Some of these amps have very little asymmetry shifting from attack to decay. So since we've all heard Marshalls that bloom and swirl I just assume that, since Marshalls are prone to this shifting symmetry behavior with clipping, that it is likely associated. Of course we all know there's no such thing as "mojo". There are no magic amps doing something other amps can't do. But learning just what the good ones are doing can be very nebulous in a clipping circuit. An awful lot of things are changing and reacting.

EDIT: Ah! The other reason I think the asymmetrical response with input level is significant and may be related to phase is that when the wave form is asymmetrical it sounds a little like a cocked wah pedal. And the Marshalls I've heard that do this shifty thing as the note is plucked and decays sound like this too, but with some clarity sneaking through subtly and differently from attack to decay. So I'm thinking it could be this shift in duty cycle symmetry that is at least part of what we're hearing. Listen to the solo in this Black Crows song. It's reported to have been played by the producer Brenden O'Brien. When they couldn't get a track they liked Brenden pulled out an old Marshall he liked, plugged in, cranked it up and played the solo you hear on the album. No pedals and just killing it with an amp he knew would deliver. See what you think. Solo starts at 2:06 and then a great outro at 2:43. This is an infamous Marshall tone on more than one forum. Though we haven't addressed it here much. And, FWIW, you can hear the low end woofing artifacts very clearly. And these seem to show up a lot in popular and revered guitar tones from Guess Who, BTO, Scorpions, AC/DC, etc. Coincidence?

[Helmholtz]

Yes, that's what I'm talking about.

Zollner calls it duty-cycle asymmetry, where duty-cycle means the time percentage of the signal cycle (or period) the power tube is conducting current.
So class A means 100% duty-cycle and class AB means between 50% and 100%.

With increasing signal the operating points and average plate voltages of the LTPI triodes shift asymmetrically.
The changing PI plate potentials charge the coupling caps and cause the grid bias to shift differently for both power tubes.
This is a dynamic effect involving relatively long time constants.
With beginning grid conduction the effect gets intensified (and eventually leads to blocking distortion).