50w Plexi mods

[Greg_L]

Okay, lots of tedious tiny micro shit today. Today's progress was good, but I'm concerned about noise with how I'm having to route everything. The general rule is keep things short. I left things a little long. LOL. I want to have slack to move things around if need be. We'll see. That push/pull pot is in cramped quarters. I can see this being way way easier on a full size 100 watter, but this is a 50, and it's a smallbox. No matter, I am the man.

So this right here is hopefully the last bit of PCB work for his mod. Replaced the V2a cathode resistor (in the shadow of the giant yellow thing) with a JCM 800 value, which is also an early Plexi value, and added a .68uf bypass cap (the giant yellow thing) to tighten things up and maybe bump the gain a hair across all modes. Both the JCM 800 mode and stock Plexi mode pass through this part of the amp. The .68uf was not present on this amp as stock, but there was a spot for it on the PCB. Weird. Anyway, the .68uf is optional. I can remove it later or maybe even put it on a switch.



Now to connect everything to this fucking switch. I mounted it in the lower hole just to make the connections. The cap piggybacking the resistor is called a treble peaker and the little cap on the pot is the bright cap for the JCM 800 mode. These are partly responsible for controlling lows so they don't get amplified later and muck things up. Pretty much everything about the "Marshall sound" is controlling the low end and letting the mids and highs get gnarly. I used stock values. 470pf/470r for the treble peaker, 1000pf for the bright cap.



This is everything connected to the switch. Hopefully this is right!



Now, the jacks back in place and that wirey octopus switch tucked underneath it all.




This is what the front will look like. Amp is upside down. The pot replaced the Ch II high input and it will control turning on the cascaded gain and be the gain pot for that mode. The Ch 1 inputs are stock Plexi.



And that's fucking enough for today. What a tedious stressful task that was. Tracing wires, going back and forth from the amp to the schematic, making wires, ugh. I'm done for today.

[Tadpui]

Nice going dude! I've only done paint-by-number soldering projects, but this is a whole other level. Hope it turns out awesome.

[rayc]

Impressed again.

[Greg_L]

Update - slight failure. Just as I suspected, I'm getting oscillations on the high gain side. I've got to do something better about the wiring. The Plexi part works as it should. The cascade side works with the gain turned down, but once I turn it up the signal disappears and comes back with some god-awful squealing. It's gotta be parasitic oscillation. I suspect it's all those long loopy wires crammed into the corner. I'm gonna have to shorten and straighten the wiring. Blah.

[rayc]

Fun....

[Greg_L]

Fun....

Lol. Yeah I'm rerouting the inputs so the pot with all of it's sensitive wiring is right up close and personal to where the leads need to go. So what's gonna happen is, my plan anyway, the bright channel Plexi inputs will be moved to the Ch 2 holes. They'll work the same, they'll just be moved over. The JCM 800 pot and low input will be moved from the Ch 2 holes to the Ch 1 holes, but flipped. The low input will be in the top hole (which will never get used anyway), and the gain pot/switch will be in the lower hole. This will give me clearance and proximity to shorten all of the wires and make every connection much more direct and separated. Hopefully this works.

[Minerman]

Update - slight failure. Just as I suspected, I'm getting oscillations on the high gain side. I've got to do something better about the wiring. The Plexi part works as it should. The cascade side works with the gain turned down, but once I turn it up the signal disappears and comes back with some god-awful squealing. It's gotta be parasitic oscillation. I suspect it's all those long loopy wires crammed into the corner. I'm gonna have to shorten and straighten the wiring. Blah.

Well dude, the way I see it is you're halfway there, & as long as you don't blow the amp up like I'd do it's just a matter of time before you get it how you want it...

I'm rerouting the inputs so the pot with all of it's sensitive wiring is right up close and personal to where the leads need to go. So what's gonna happen is, my plan anyway, the bright channel Plexi inputs will be moved to the Ch 2 holes. They'll work the same, they'll just be moved over. The JCM 800 pot and low input will be moved from the Ch 2 holes to the Ch 1 holes, but flipped. The low input will be in the top hole (which will never get used anyway), and the gain pot/switch will be in the lower hole. This will give me clearance and proximity to shorten all of the wires and make every connection much more direct and separated. Hopefully this works.

Way over my head, but it sounds like you've figured out how to fix it, good deal man...Keep us updated...

[Greg_L]

Ok so the re-route of the wiring is done. Looks clean and tidy - short runs, no extraneous crisscrossing, visually this seems like a much better way to do it.




The long black shielded grid wire is super duper sensitive, and I need to dig up a 68k resistor to put at the end where it connects to the tube socket. I'm gonna try it without the resistor first, but I think it might need it. It helps tame high end noise and RF that gets into the guitar signal. Any noise that gets in that wire gets amplified because that wire goes straight to a gain stage. There are two unused 68k "grid stopper" resistors on the PCB that used to go with the normal channel. They served the purpose I need them for with the original layout. So I can yank one of those, or maybe route this grid wire back through the PCB for a stock look. Maybe. Those are my options.

[Greg_L]

Update: Got the wiring finished. I ended up stealing an unused 68k from the board to use on the tube socket for the grid stopper.






And......IT'S ALIVE!!!!!!!!!!



This is the configuration that's gonna stay. The Ch 2 top and bottom inputs are really Ch 1 inputs, and Ch 1 is now the low input on top, and Plexi/JCM push/pull gain control on the bottom. The guitar can just stay plugged into the top right input for all of it. I still had some oscillation though. I was just about to throw it through the window then I remembered something a guy told me - bypass cap across the plate resistor of V1b. I threw whatever high voltage cap I had on hand across V1b, and voila! It worked! I need to fine tune the value, but fuck it, right now it works and sounds pretty awesome! Now it sounds like I imagined it could sound. The Plexi side works perfectly, the cascaded "JCM" side works perfectly, it now has a traditional master volume and a post-phase inverter master volume, it's all good! Clips to come soon.

[Lt. Bob]

I had a 50 plexi that I used exactly like that .... no cab .... just sitting on the speaker naked in its glory.

One night a glass of water fell from someone else's gear that was right next to and above the Marshall pouring right across all the tubes while in full cry!

*pop* *pop* *pop* went the tubes!




Amp was fine though ... replaced all the tubes and it worked same as always.

[Greg_L]

Lol well mine won't be staying in the nude like this.

[rayc]

Nude but nice!
I look forward to hearing it - you little mod guru - good that you've written this into a thread. You should copy the text & store it - that little "bypass cap across the plate resistor of V1b" is obviously gold that needs a safe deposit box.

[Greg_L]

Nude but nice!
I look forward to hearing it - you little mod guru - good that you've written this into a thread. You should copy the text & store it - that little "bypass cap across the plate resistor of V1b" is obviously gold that needs a safe deposit box.

It's an oscillation-killer. I just don't know why. Some 2203s/2204s and higher gain amps have this little feature as stock. Some 2204s in particular use the shielded part of the grid cable connected to the plate pin on the tube socket as sort of a de facto capacitor. I'm not sure why it works yet, but it works. I think it's bleeding off the super high frequencies that overload the tube and you can't hear them anyway. I don't like the cap across the plate resistor method because I think I'm gonna have to try a bunch of different values to find the smallest one that works. I'm going to make another shielded grid cable and swap it in to see if that works.

[Farview]

I had a 50 plexi that I used exactly like that .... no cab .... just sitting on the speaker naked in its glory.

One night a glass of water fell from someone else's gear that was right next to and above the Marshall pouring right across all the tubes while in full cry!

*pop* *pop* *pop* went the tubes!




Amp was fine though ... replaced all the tubes and it worked same as always.

Who puts drinks on their equipment?!!! That's a disaster waiting to happen, especially if there is alcohol involved.

[Greg_L]

Who puts drinks on their equipment?!!! That's a disaster waiting to happen, especially if there is alcohol involved.

There's this one dude that calls himself "Billy Blades" and he builds cup holders into his head cabs.

[Lt. Bob]

It's an oscillation-killer. ..... I'm not sure why it works yet, but it works.

me either ... caps don't allow DC thru so maybe that has something to do with it

[Greg_L]

me either ... caps don't allow DC thru so maybe that has something to do with it

The plate wires carry both B+ DC to charge the plate and audio signal AC out of the tube. You have AC and DC in the plate wires at the same time. The coupling caps block the DC before the gain/volume pots or wherever the signal goes after that stage so you're left with AC only. I *think* what the plate bypass cap is doing is blocking very high supersonic dog hearing frequencies that can throw the tube out of linear mode. I've found multiple methods to achieve the same thing, but for me right now the cap across the plate resistor seems to be the easiest. Now what I'm gonna do is install smaller and smaller caps until I can find the smallest cap that will still block the oscillation. The relatively gigantic 8200pf that I have on there now is too big and shelving off too much high end. I'm gonna go all the way down to a 10pf and see what happens.

[Greg_L]

Well alright! She's all buttoned up!




So let me get the nerdy bit out of the way. I ended up with a 470pf cap across the V1b plate resistor. The minimum I could go with was 220pf, but my only 220pf cap was of unknown voltage capacity so I didn't wanna risk it. 220 did kill the oscillation though. 10pf was like it wasn't even there. 110pf still had oscillation, but higher up the gain dial. 220pf was the minimum that worked. My highest *known* voltage rated cap was the 470pf. I plugged all this shit into a plate bypass calculator and the difference in gain at 1khz between 220pf and 470pf was less than half a decibel, so the 470pf was gonna do the job without sacrificing anything. Win!

So that's that....unless something blows up. The Plexi side still sounds like a Plexi, like it always did, and the JCM cascaded side sounds like an angry Plexi. Awesome! The cascade is not "high gain" by today's standards, it's just more grunt than Plexi and very similar to my JCM 800. It might be a little tighter and "grindier" than the 800 due to the added bypass cap on the V2a cathode. Very punchy. Might be a little bright, but we'll see.

Clips soon.

[Lt. Bob]

dude ..... you're on the edge of being an amp builder ...... or at least a hot-rod mod shop.

A little more studying and you're there.

[Greg_L]

dude ..... you're on the edge of being an amp builder ...... or at least a hot-rod mod shop.

A little more studying and you're there.

Haha not quite but I learn a bunch more with every little project.

My next trick might be converting this old SS junk combo amp I have in the garage into something useful...like maybe a Champ or Deluxe or something. Maybe I can gut it and build something in the chassis, put a nice speaker in it, and have a decent combo. It's got a reverb tank already.