The Watering Hole

It's no secret that the first amp brought into the limelight, created by Jim Marshall, was a copy of the Fender Bassman, specifically the 5F6-A chassis.   There are some interesting, and subtle, differences, but we'll save that comparison for a later time.   Let's just take a quick look at what Marshall has to offer.    When the first JTM45's were released Jim/Ken were futzing around with component values so there are a lot of versions of this thing, most of which no one really talks about.  What they ended up with, though, after a bit of work, is a topology that really didn't change much, ever!   The circuit in the JTM45 ended up in the JCM800, the JCM900, JCM9000, etc.   Rather than change a lot of stuff in the basic design, they just added features and renamed the amps.

I'm going to spend most of my time on the JCM800, chassis 2203/4 just cuz it's my favorite... actually, for another reason, it's got ALL the Marshall "stuff" in it.  Here's a link with the Marshall "standard preamp" design.   I have included a color-coded version of the Master Volume preamp, here, for review.  

http://www.drtube.com/schematics/marshall/jcm800pr.gif

So, onward, let's dig through this thing.   The standard Marshall preamp has two inputs a "High" and a "Low" input.   The "high" input adds two things:  1) An extra gain stage and 2) A "high pass" filter... so "higher gain" and fewer low frequencies.  I'm going to focus on the "High" input as this is "the sound of Marshall" that we're all used to.  The other input, most folks don't use it, much.  

So, plug the guitar into the "High" input and you see the 1M input impedance - as per the Fender.   This will not load the guitar.  The 68K grid-block resistor is there to protect the tube, as per Fender, too.  This is standard design.    Tube V1B is the first tube the signal hits when a guitar is plugged into the "High".  This stage, in the Marshall, is a bit lower gain than the first stage in the Fender.   WHAT???? In the Fender, the Cathode resistor was 1.5K ohms.   Remember, we discussed, that the lower the value of the resistor, in this position, the smaller the signal has to be in order to get a large output and the higher current that is going through the tube.  Higher current = "warmer".  Lower current = "colder" or more "sterile" and "brittle".   Interestingly Marshall starts out with a more "brittle" stage than the Fender amp does.  

Another difference, in this stage is the fact that across that 2.7K resistor is a .68uF capacitor.  This is a MUCH smaller value (36 times smaller) than the 25uF capacitor used in the Fender amp.   Using the 1/(2 x pi x R x C) equation we find this circuit provides a high-pass filter with a shelf centered at 86Hz.  This is the first difference between the Fender and the Marshall.   In the Fender, stage 1 is flat response from 8Hz to 20kHz.  In the Marshall we start the "flat" response at 86Hz.  

The output of this stage is coupled through the .022uF capacitor (used ONLY for blocking the DC plate voltage from V1).   The signal hits the next part of the Marshall equation which is the parallel combination of a 470K resistor and 470pF capacitor.    These two components, together, form a frequency selective network which blocks pretty much everything below 720Hz (1 / 2 * pi * R * C....    1/(2 * 3.1415 * 470000 * 470E-12 - where 470E-12 is 470 x 10 to the minus 12 power).  

This is, essentially, a high pass filter, but it forms a slightly different slope than the standard high pass filter.   Marshall describes this as a rising slope of 10dB per decade.    A "decade" is a 10x multiplication in frequency.  So, if the slope upward starts at 720Hz, at 7200Hz the output would be 10dB higher.  This 720Hz shelf is THE "sound of Marshall".  It's all right here, folks.  

Let's view these two stages a little more closely.   Unlike the Fender first stage, this Marshall stage has less gain and can take a MUCH larger input signal swing WITHOUT distortion than what the Fender's first stage can.   WOW!  So, NO distortion would occur in this stage, normally.   The signal is then "filtered" or "shelved" at 86Hz by the feedback network connected to the cathode, it's then pushed to a SECOND filter (still clean) that shelves AGAIN at 720Hz.  This is, then, "picked up" by the "Gain" control and fed into the second stage. Because the signal DID NOT pass through a tone stack, the input signal level, to this second stage will still be higher than the input signal level at the 2nd stage of a Fender amp!    

OK... moving onward...   V1a, second stage tube... if you notice there's a 10K resistor at the cathode of this stage and NO capacitor.  This stage is FLAT frequency response AND lower gain.   The input signal can, again, swing quite a bit without distortion, but at this stage some distortion will be generated here.  Because of tube charactaristics there will be slightly more distortion on the negative direction of the signal (which WAS positive feeding the first stage - each stage "inverts" the signal 180 degrees) and less distortion generated on the positive direction of the signal.   It's been filtered one time, already, at 86 and 720Hz and THEN the distortion starts getting generated.    As per the first stage, this stage has an even HIGHER value resistor in the cathode position.  This biases this stage even "colder" and, in reality, makes it sound even "thinner".  

At the output of this stage, through another .022uF capacitor and ANOTHER 470K/470pF parallel circuit the signal will pass.  AGAIN with the 720Hz filter.  So the partially distorted signal, from the last stage, is fed through another 720Hz filter and the low end is, again, removed/filtered out.   This feeds the third, and final GAIN stage in the preamp, V2a.  This stage has a very low value resistor in the cathode portion (820 ohms).  Again, there is no cathode bypass capacitor which means the stage gain, here, is flat.  A couple of things, to note, here, too.  First, the small resistor in the cathode position biases this stage "hot" and "warms up" the signal.  Secondarily, the small value of resistor makes this stage EXTREMELY sensitive to input signal variations (small signals produce HUGE changes) and thus is significantly MUCH MORE PRONE to distorting than stages 1 and 2.   The signal has been filtered at 720Hz twice, prior to this so it's pretty "clean" in the low frequency region.  

Stage 3 is directly coupled to stage 4 which has no plate resistor and a 100K cathode resistor.  Stage 4 is what is known as a "Cathode Follower".  This stage has no gain.  The input signal is "coupled" directly to the output.  This stage is designed, primarily, to be an "impedance" converter, converting the high output impedance of the preceding stage to a low output impedance for this stage.   This allows almost the ENTIRE signal, generated by the stage previous to this, to be "dropped", or "fed into" the tone stack.   Secondarily, the tone stack is now part of the "feedback" portion of this tube stage, which affects the character of the tone stack, overall.   I'm almost out of characters, so I need to quit this thread and head into the next one...

BRB!

Dar

This is super cool.  Any knowledge about my baby, the Sovtek Mig 60?

Send me an email at kornowsd@zeecon.com and remind me to do that... I'll try to get 'r' done...

Dar

No prob.  I am asking in this thread because I think it is very similar to the JCM 800...

OK... I left off at the "Cathode Follower" in the last post... so, I'm starting with the final thoughts on that, here.   I noted, in parting, that the tone stack was in the part of the cathode follower that provides "feedback" for this stage.  This differs from most implementations where the tone stack is fed from the plate and not part of the feedback network.  This does have an effect on the overall "feel" of the amp, with most saying the tone stack is more "responsive" or "bouncy" when it's fed off the cathode of a cathode follower, rather than off the plate of a tube.  It is something to think about...

Now, Marshall amps have varied, a bit, over the years.   In almost every Marshall, however, the first stage tube has a 2.7K resistor bypassed by a .68uF capacitor.  I, honestly, don't think Marshall/Bran put this value in for distortion/voicing purposes.  In the early days of tubes, as you recall from the power supply section, it was hard to find capacitors large enough to supply the power amp sections with enough current to run wide open and the supplies "sagged".   Well, the other part of this equation was, that "ripple" that couldn't be gotten rid of also caused a lot of 60Hz hum.  The first stage in the preamp is most susceptable to this hum.   By choosing a capacitor that set the high pass filter well above 60Hz, Marshall mitigated this hum without compromising the guitar frequency range which started at about 84Hz (Low E) and went up from there.   Eventually, though, this intial 86Hz filter contributed to the "sound of Marshall" without them designing specifically for that purpose and it's stayed there ever since.

One thing I did not cover is that .001uF capacitor that goes from the top terminal of the Gain pot to the middle terminal of the Gain pot.  On most Fender amps, you'll see a 120pF or a 220pF capacitor in the same spot but with a switch labeled "Bright".   Marshall just plugged the thing in, and left it there.  No switch.   This capacitor is most effective when the gain is turned down toward "0".   The lower the gain, the more "effect" this cap has.  This is the SAME thing as the mods you make to the guitar volume pots by putting the capacitor across them. It allows retention of high frequencies as the volume is turned down.   This keeps the Marshall "crisp" and "clean" at low volumes.

Also, remember, that harmonics (either generated by the strings or through distortion) are all high frequencies!  At lower gain levels, the Marshall amps will still distort some amount.   By adding more high frequencies, at higher amplitudes into the following gain stages, you end up with the high frequencies being distorted more than the low frequencies, at lower gain settings.  This is what makes for that "Crunch" we're all used to when the Marshall gain control is set at 3-4 - it's "clean" yet "distorted" at the same darn time.    

As you turn up the gain, and the signal gets increasingly distorted, a larger and larger body of high frequency harmonics are being generated.   This, by default, reduces the impact of the fundamental and gets increasingly "shrill".   As the Gain control is turned up, on these amps, the capacitor is slowly "removed" from the circuit coupling less and less high frequency to the following stages.   This keeps the balance of highs/lows roughly the same all the way from low gain to high gain.   Another sort of "Marshall innovation".   With Fender amps you had to switch that in/out, all the time.  

Marshall also used a much higher value cap, here, which means it "extended" the reach further toward the low midrange when the gain control was turned down.  MUCH further than the Fender "Bright" control.   This, again, added to the "girth" we hear in the Marshall and the "glassy highs" we're so used to in the Fender amps.   Another minor, but important, differentiator between the two amps.

And that, my friends, is what creates the "sound of Marshall..."   A very clean first stage with a high-pass filter set to 86Hz.   A 720Hz interstage coupling filter with a 10dB per decade rise.   The .001uF capacitor across the volume control helps increase the gain of high frequencies at low Gain settings.  Despite the fact that the first stage is lower gain than the Fender, there is less loss between the first and second stage providing a larger signal at the second stage of the amp, causing a bit of distortion at this stage.   Both stage 1 and stage 2 of the Marshall run "colder" than the Fenders, which produces a more "brittle" and "dry" sound.  Over the years, the values, for the resistor, in stage 2, have varied with amp model, but the basic design has stayed the same.  

Another 720Hz filter shows up, further filtering the signal for care and feeding of stage 3 where the bulk of the distortion actually occurs, in this amp.  That 820 ohm resistor biases this stage "hot" and "warms it up" quite a bit.   Further, that small resistor makes this stage highly "sensitive" and significantly more likely to distort.   Again, the stage runs flat, in terms of frequency response.

Stage 3 feeds directly into a cathode follower.  Gain = "1" or "Unity Gain".   This is an "impedance" converter which feeds the tone stack... and, voila, we're done.

Enjoy...

Dar  

One more point to add, that I left out in the previous two posts... you can get an amp to sound INCREDIBLY like a Marshall, just by inserting interstage coupling networks (470pF / 470K) between the gain stages and that is WITHOUT changing the tone controls one whit!    It's just a thought... and I move on.  

Dar

Dar....you the man!  :)

Randy