Marshallness, or an amp’s guide on how not to be tweedy

Steve’s guide to classic amp sounds 2: The Marshall sound

The Marshall overdrive sound is instantly recognisable. If ‘crunch’ is the word, then this is the crunch of broken glass underfoot. Anything further removed from the Isley Brothers’ ‘Summer Breeze’ kind of overdrive (ok it’s a fuzz pedal but it is smooth) – or the Carlos Santana type of overdrive (Mesa Boogie/Dumble, very American overdrive) – is impossible to imagine. Marshallness is toppy, aggressive, harsh… and the instantly recognisable core sound of rock – specifically, British rock.

Marshall overdrive was certainly available from the very first amps Ken Bran put together in the back of Jim Marshall’s drum shop in 1962 (though he may not have meant you to turn them up to 10 to obtain it). And it hasn’t changed that much. From the JTM45, through the plexis and JMPs, the Master Volumes, the JCM800 and 900s through to the DSLs and TSLs, that mouthful-of-ground-glass sound remains essentially the same. Marshall have been pretty good at synthesising it in their solid state amps too (if you see a 5005 ‘Lead 12’ at the car boot, buy it – Billy Gibbons recorded some classic ZZ Top numbers on this 12-watt solid state Marshall, and it sounds, well yes, just like a Marshall).

So if the Marshall sound was there from the start, we should be able to see it being shaped in those early JTM45 circuits, and follow its essential elements down the years.

You may know that the essentials of the classic Marshall circuit are very close to those of the 1959 ‘tweed’ Fender Bassman. The 1950s Tweed Fenders tend to go into overdrive sooner than the 60s Fender classic ‘blackface’ models, and in that respect to resemble Marshalls. However, tweed overdrive is softer and looser than ballsy Marshall crunch. An understanding of the reasons for the differences beween tweed Fender overdrive, and the overdrive sound of the basically very similar Marshall amps, will help us to understand the components of the Marshall sound.

The Marshall/tweed Fender Bassman circuit is more complex and sophisticated than that of the basic AC30 circuit I wrote about last time. The basic AC30 preamp has just one gain stage and a volume control. The Marshall/Bassman circuit, however, goes like this:

input gain stage > volume control/attenuation > gain stage > EQ driver > three-knob EQ

As you can see there are two extra stages in there before the EQ, which on these amps is the full treble-middle-bass array (the presence control is elsewhere, in the negative feedback circuit). The volume control comes before these stages and is perhaps better described as a gain control, as it also controls the degree to which those ‘extra’ stages are overdriven. The reason for the first of the two, the extra gain stage, is that a three-knob passive EQ saps loads of gain out of a preamp. The second, the driver stage that also comes before the EQ, is a ‘cathode follower’. It has no intrinsic gain, but serves to make the signal low-impedance and high-current, the more efficiently to drive the EQ. It isn’t actually that necessary though, it’s an electronic luxury. Fender didn’t bother using it in the blackface preamps they went on to design (and they were ok!).

By the way, I have a sneaky feeling that a cathode-follower stage in itself changes the tone for the better somehow. The Vibro-King has a cathode follower as its first stage; it drives the complicated valve reverb in that amp, but an experience I once had when building a Vibro-King preamp circuit into another amp suggested to me that the cathode-follower stage actually brings something good to the sound – perhaps in the way it drives the next stage. One fine day when I’m not too busy failing to fix your amp I will try an experiment…

Anyhow, back to the story. Remember that we haven’t yet said what makes the difference between the Fender tweed Bassman and the Marshall; we’re still looking at the shared nature of their very similar circuits.

The EQ is the last stage in the Bassman/Marshall preamp. It comes after preamp overdrive has already happened. Modern overdrive preamp designs all do it this way around, but many earlier amps made the mistake of putting the tone-shaping controls earlier in the preamp. My beloved Rivera-era Fender Concert falls into this category – and that is the reason why its overdrive only sounds good at certain settings of the gain controls. The right way is to create the overdrive, then have an EQ stage to shape it. Back in 1959, as in so many other things, Leo Fender got it right.

You do in fact need to do something about the signal EQ before you overdrive anything, too. And there Marshall diverges crucially from Fender… I will come to this…

Ok let’s look at the differences between the Marshall circuit and its American predecessor. The first one is that Fender used a slightly lower gain preamp valve – a 12AY7 – in the first stage. Marshall plugged in an 12AX7/ECC83 – the high-gain type that most guitar preamps use in most positions. This gave much more gain pushing right through the rest of the preamp, and it means that the preamp goes into overdrive sooner. Whether this was deliberate or not is a moot point – quite possibly it wasn’t, because just after the volume control Marshall also introduced a larger value series resistor, which cut down the signal a little (though not removing anywhere near as much gain as that high-gain ECC83 valve had introduced). Much more crucially in producing the Marshall sound, though, on the ‘lead’ preamps they parallelled this resistor with a capacitor designed to let higher frequencies through. Here is where the pre-overdrive EQ shaping that I was talking about, happens. In a little parallel box that brakes the signal with a resistor, and shapes it with a cap that lets higher frequencies through.

So this resistor/capacitor filter that follows the volume control in the classic Marshall circuit has the effect of cutting back the bass and boosting the treble. I would guess that this was probably a response to guitarist’s demands that the amp help them ‘cut through’ and be heard. Well you know what guitarists are like (…oh I forgot, you are one).

Anyhow what it did in my view was to shape the Marshall overdrive sound. Toppy, harsh, aggressive. What Marshalls do via this filter and the extra gain in the first stage is to drive lots of high-end frequencies through the first three stages, then let the player shape the sound with the TMB knobs and the presence control. Not that I imagine they intended to do that – in 1962 nobody was designing amps to create distortion. As I wrote, I expect they were going for a treble boost so guitarists could rise above the general racket with a solo – same thing Vox were after when they put a treble boost in front of the simple AC30 preamp. And with the same probably unintended result – not searing clean solos but screaming treble-y overdrive when the amp was turned up higher than it was ever designed to be.

You know how bass controls on Marshalls are. Kind of pointless. Doesn’t really matter much what you do with them. That’s because the filter I’m talking about cuts back the bass so much that there’s little left for the bass control to do (it comes last in the chain, remember) . As I said, this may well have been the result of the designer trying to help guitarists ‘cut through’, but what it also did was to create an overdrive sound that people liked to hear. Overdriven bass does not sound good, it’s muddy and, well, farty. Overdriven middles with the treble rolled off gives you ‘woman tone’, and brings that jasmine into bloom. Open up the treble and drive it and you’re listening to the sound of rock, the Marshall sound.

A little later, in the early 70s, Marshall cottoned on to this essential misuse of their amplifiers and, after listening to their customers, introduced a master volume control to tame output levels whilst allowing some overdrive, and also tacked on an extra gain stage at the front end of the preamp. This too was followed by a resistor/cap filter, rolling off the muddy bass before overdrive, and shaping what was still the same Marshall sound, with the same essential components. At the heart of the DSL/TSL preamps is the same basic circuit.

So the Fender/Marshall differences that make all the difference are (a) a high-gain valve in the Marshall stage one, and (b) a filter after the volume control that cuts the bass and boosts the treble. Now all the Marshall was waiting for was for Eric Clapton to come along and turn it up to 10.

But there’s a couple more crucial differences. For one, the Marshall has a bit more negative feedback than its ol’ pappy the tweed Fender. This actually cuts down the gain a little (but the Marshall still has plenty to spare thanks to its ECC83 in stage one). It also produces a sound that is little more controlled, a little less ‘natural’ maybe. This adds the final touch, keeping the Marshall’s inherent wildness in check just enough. Turn a Bassman up full and you will often get some slightly odd effects on loud notes. With an ECC83 in stage one a Bassman becomes almost untameable, and these ugly overtones, which are caused by too much gain causing blocking distortion and/or internal oscillation, become intrusive. I would suggest that Marshall found this to be the case after putting in that ECC83, and introduced a touch more negative feedback, and a slightly bigger series resistor after the volume control, to tame things a wee bit. Then they added the cut-through-the-band cap to give guitarists’ fragile egos a bit of a support. And they were nearly there.

One last difference. The Fender has four 10″ alnico speakers in an open back cabinet. Sweet and airy. Marshall really wanted a 2 x 12″ cab for reasons of economy, and they were used to making closed back units because they had started out making bass cabs in response to bass players’ understandable grumbles about how the likes of ‘bass’ amps with 4 x 10″ open back cabs (can you think of one?) were not helping them compete with guitarists. But speakers in those days couldn’t handle much more than 15 watts, and Marshall kept burning their little voice coils out in 2 x 12″s when they turned up these new 40-plus watt amps they’d been building. And so, in response to some very simple maths, the closed-back 4 x 12″ cab was born, right back at the start of Marshall history, the little box was stacked on top of the big box, it sounded like rock music, it looked like rock music, and the circle was closed.

2015 footnote:

Had a read through this and realised I’ve noticed a couple of other things about the early Marshalls. One is the ‘bright cap’ – a cap across the volume control that lets tops through at lower volumes. The old Bassman had a big value there – 1000pF – and so do some of the early plexis. Coupled with the extra gain in the first stage in the Marshall this produces an amp that feels almost uncontrollable with the volume knob, so much high frequency signal comes through. Of course once it’s on 10 there’s no difference. If you have a plexi-circuit amp that has a lower value bright cap, try a 1000pF (=0.001 uF). If you like wild, free and wicked like all the bad girls do, you’ll enjoy it. The Bassman doesn’t feel the same at all, though that cap does brighten the Bright channel very significantly, and probably this is due to lower gain and more compression.

Compression is the other thing. The old Marshalls tended to have more capacitance in the power supply, meaning that they didn’t compress so much and had more dynamics – dig in and they respond with more power rather than by compressing via power supply ‘sag’. The extra capacitance gives the amp the resources to supply the extra power a loud note demands. Another reason – perhaps the main one on reflection – why the old Marshalls sound ‘harder’ than the ’59 Bassman. Whisper it, but the Brit transformers might have added to the effect.

Ah well, happy days! Haven’t written owt on these blog pages for yonks, must make summat else up.