Marshall JCM800 4210 Channel Switcher - modification - advice sought.

[walkman]

Back in 2007 i modified my Marshall JCM800 4210 Channel Switcher 50 watt combo and have been happy with it though new it could be refined.

Recently I changed the main power filter caps with new German F&T capacitors and got keen about working on the amp again.

Here is the original thread i started back then Marshall JCM800 4210 Channel Switcher

I've attached the original Marshall schematic, a modified schematic with the changes made in 2007 and preamp schematic of the proposed changes that I'd think could work well for the amp.

Essentially the clean channel has been left alone for now, all the work is the signal flow through the gain channel and removal of the clipping diodes. The main difference between the 2007 mods and what I'd like to do now is that the gain channel now feeds off the first triode in the clean channel and that the signals mix later on than before. Also the order of the FX loop and reverb is changed and the reverb mix is a virtual earth mixer.

Ive used Merlin's designing preamps book in the section about different amp designs for inspiration for the values of the inter stage resistors, capacitors and the cathode caps and resistors.

Please if you'd like to make some comments regarding the proposed changes that would be useful. Ive taken a bit of time considering things rather than jumping straight in with the soldering iron.

The next step will be to consider how it will all work on the PCB and whether the design idea will work with the physical layout of the amplifier.

edit: I've added another modification to the virtual earth idea, perhaps the fxs loop is better after the reverb, or maybe even return the reverb to other side of the PI

Happy New Year
Walkman

[Chuck H]

I think there must be some mistakes in your schematic. It appears to have the two channels operating parallel full time, the cascade preamp is disabled and so is any channel switching? The proposed change only looks different in implementation but not much in function.

I would think it's best to reverberate the effected signal rather than effect the reverberated signal in a perfect world. Some amps can't really be done this way though. Yours would be one (without more extensive modification).

Your proposed mixer looks cattywhompus to me. You appear to be amplifying the two incoming dry signals and staying at unity gain for that stage with the reverb. Which is the exact opposite of what that stage would normally do. I would expect a weak reverb like this. Also, There is too much signal at the loop send for most effects and the loop operates within the VE feedback loop. I've tried a similar FX loop design and it was unstable.

Remember, you asked for comments and opinions.

[walkman]

Hi Chuck

Yes there is switching as can be seen in the original schematic.

And thanks for your comments I'd rather make a mess of a drawing than a mess of the amp.

So do you think the cascaded preamp would work with the mixing occurring before the fx loop as set out in the original.

The current mod works well but can get a bit woolly when the gain and volume are both up.

Will redraw with the switching.

[Chuck H]

The current mod works well but can get a bit woolly when the gain and volume are both up.

You can't have it both ways IMHE unless you use an attenuator with an amp specifically intended to clip the power tubes. If a MV amp is designed to sound "right" with the MV in use then something has to give when the power tubes start to clip! It SHOULD sound different under these conditions. If it's not loud enough you can always bump the volume of the lower volume setting with a mic or if you prefer the "woolly" tone, but it's too loud you can use an attenuator or less eficcient speakers. The only way you can have it both ways with a MV amp is to design it big enough that you'll never be tempted to clip the power tubes (like 100W) and then design it so the power tubes only barely compress with the MV at full volume. That's pretty much what the modern uber gain designs do.

[walkman]

Hi chuck

No there's no problem with volume. Actually the master volume works quite well in this amp. It is post phase inverter. I was wondering if the current mod overdrives the reverb mixer or the PI too much. The speaker is also quite efficient. The tone when fully cranked is defiantly more Hendrix than metal but I was hoping to get a little more crunch or a little better note definition. It's great for grunge rock or heavy rock but lacks some crunch.

[Chuck H]

The proposed mods would certainly drive the PI a bit harder because the VE mixer values have been jostled from unity gain so that the mixer stage will now offer some amplification. As to crunch... In the "as is" schematic I can see that the second gain stage for the boost channel doesn't have a cathode bypass cap. If adding upper end gain like that is too harsh there's a neat trick I penciled that I want to try on my next build... Use a .1uf bypass cap on that cathode. Then counter the added brightness with a 470p cap across the plate resistor. You may need to add a resistor (probably 4.7k to 22k) in series with the plate cap. The goal would be to get that stage to sound as if you'd done nothing to it even though there is now a small bypass cap and a plate resistor bypass circuit. Then you can put a switchable 1uf cap across that cathode to achieve a mid boost, instead of the usual mid and treble boost you would typically get from a cathode bypass cap. I thought it was a clever idea. Maybe it can help you.