Marshall 2203/2204 Master Volume Mod - Debunked?

Marshall Amp Discussion

Moderators: pompeiisneaks, Colossal

Post Reply
User avatar
Littlewyan
Posts: 1911
Joined: Thu Sep 12, 2013 6:50 pm
Location: UK

Marshall 2203/2204 Master Volume Mod - Debunked?

Post by Littlewyan »

I read in a few places online about a mod that 'fixes' the Master Volume circuit that Marshall used in the 2203/2204 amps. Basically the tone gets thin with low Master Volume settings and Kevin O Conner said in his book that this is because the Master Volume forms a filter with the PI Input Cap which cuts bass. The fix is to simply put a 100K resistor between the MV Wiper and the PI Input Cap.

Now, before I implemented this mod I wanted to be able to measure the difference. So I got out my Spectrum Analyzer, stuck a signal generator on the input pushing a 100Hz signal into the amp, put the Pre Amp Gain on full to get some harmonics going, measured the amp and I found the following.....

Master Volume Setting - 2
1Khz - -56.45dBu
.05Khz -39dBu

Master Volume Setting - 5
1Khz - -40dBu
.05Khz -28.6dBu

Master Volume Setting - 8
1Khz - -34.9dBu
.05Khz - -28.3dBu

The frequencies were measured on the Master Volume Wiper. With the MV on 2 we can see that the 1Khz signal is roughly 17dB lower than the .05Khz signal. With the MV on 5 its now only 12dB lower and with it on 8 its only 6dB lower. So what we're seeing is quite the opposite, the MV is cutting high end at lower settings.

I installed the resistor anyway, measured again and got exactly the same readings. I did a listening test as well but I couldn't really hear any difference. I should have recorded them but didn't have that much time.

Personally I think this is normal behaviour for a volume pot that hasn't got a bright cap. Also its interesting to note that the power amp only just starts to break up with the MV on 5. When the power amp breaks up, additional harmonics will be generated and will fatten up the sound. Thats my theory anyway.

So what do you guys think?
User avatar
Littlewyan
Posts: 1911
Joined: Thu Sep 12, 2013 6:50 pm
Location: UK

Re: Marshall 2203/2204 Master Volume Mod - Debunked?

Post by Littlewyan »

Some screenshots of the spectrum analyzer readings.
You do not have the required permissions to view the files attached to this post.
strelok
Posts: 276
Joined: Wed Mar 26, 2014 5:46 pm

Re: Marshall 2203/2204 Master Volume Mod - Debunked?

Post by strelok »

My theory has always been that the PI distortion is really what adds a lot of the fatness to the Marshall sound. I have a Metro '68 100 watt I built from a kit. I modded it to have the 2203 style pre-amp and PPIMV. So its got 2 masters one pre-PI and one post.

If you dime the pre-PI master and use the PPIMV to control for volume you get a really fat sound and a lot of gain. Very similar to early Van Halen. I typically won't set the gain any higher than about 6 as it starts to get really fuzzy past that point and not really necessary for the sound I want.

If you do the opposite and dime the PPIMV and set the volume with the pre-PI control its more aiken to that typical 80s rock JCM800 sound as you might expect. Brighter, thinner, more like Def Leppard. The overdrive is also significantly reduced since the PI is now running clean.

Both configurations sound great IMO, though I tend to favor the sound of the PI distortion. Very thick and mean.

That's my theory as to whats going on anyway.
User avatar
Littlewyan
Posts: 1911
Joined: Thu Sep 12, 2013 6:50 pm
Location: UK

Re: Marshall 2203/2204 Master Volume Mod - Debunked?

Post by Littlewyan »

Ah yes, I keep forgetting about the PI being pushed into distortion in these amps.
pdf64
Posts: 2688
Joined: Sat Mar 12, 2011 9:23 pm
Location: Staffordshire, UK

Re: Marshall 2203/2204 Master Volume Mod - Debunked?

Post by pdf64 »

I think that the most significant things that will likely cause the tone to change according to the master setting are
1 - parasitic capacitances between vol control terminal nodes; even without an actual bright cap, there will be a few pF between things. The effects of these will be a treble boost at low settings and treble cut at higher settings.
2 - interaction between the miller effect of the LTP input and the effective source impedance feeding the grid. At very low settings, the source impedance will be very low, hence little treble roll off.
At higher settings, the source impedance will increase significantly, reaching a max of >500k with a 1M pot; that will cause significant roll off with the miller effect capacitance, though the local and global NFB will affect it.

I can't see that a bass cut 'as such' will be caused.

I can see that a high value series resistor between wiper and LTP input will help to make the '2' effect more consistent of the master control range.

Lower value pots help to reduce the '1' effect, hence 1M pots aren't used on hifi etc, rather 100k or lower.
But with a Marshall, that would load the tone stack too much.

I suggest to another way to assess these effects would be to use a square wave ~200-400Hz input, perhaps best to inject it as late as possible (eg at the DCCF 1st grid), and adjust the tone stack for best squareness at the top of the master.
Then monitor the wave at the power amp output, presence at min, resistive load.
As a starting point, avoid overloading any stage.
Treble boost will cause a leading edge spike, treble cut will cause a leading edge rounding off.
User avatar
Littlewyan
Posts: 1911
Joined: Thu Sep 12, 2013 6:50 pm
Location: UK

Re: Marshall 2203/2204 Master Volume Mod - Debunked?

Post by Littlewyan »

Would a lower source impedance not change the cutoff frequency of the PI Input Cap? The same way lower grid leak resistors change the cut off frequency of the PI Coupling Caps?

Ill test it again when I get a chance.
Post Reply