Ground scheme 4 stage Marshall?

[alfi27]

Hi guys,

I'm 95% done wiring up my modified Marshall JCM800, which ended up being a total rebuild. Even redid the preamp heater wires, as apparently Marshall hired a chimpanzee that wasn't able to twist wires around the late 70s/early 80s...

Ground schemes have probably been discussed a ton on here, but I can't seem to wrap my head around it. The ground bus has been moved from the pots and is soldered to two ground lugs attached to the circuit board (just through two holes in the board, the lugs are not grounded to the chassis) to avoid any ground loops. What I have done so far/plan to do:

- 32 uf (for extra tube/gain stage) and 50+50uf preamp filter caps sharing one star ground
- star ground close to input connected directly to input jack, and also directly to the start of the ground bus which will connect to every ground connection in the preamp and bias caps
- 100+100uf and 50+50uf power amp filter caps grounded separately on one star ground each

It's the two first ones I'm not sure about, because in my head you shouldn't connect too many things to one star ground... And also, the actual ground lug for the preamp filters wouldn't be able to fit more than maximum one more wire anyway. Which is obviously fixable, but still.

If you could help me out that would be amazing, as I really want to get this amp up and running tomorrow... To be fair I haven't actually tried it with this ground scheme, and for all I know it works perfectly fine. It's more that I want to make the amp totally bulletproof/failsafe so I can put it back together and not open it up again until the filter caps need changing in 20 years (yes, I installed an external bias pot )

Cheers!

[Phil_S]

I don't know how much attention this will get because you are right, ground schemes have been discussed ad nauseum. I'm not quite sure that I follow what you did. It sure would help to see a hi-res picture or pictures. Ground schemes are the stuff folks will disagree about for all sorts of reasons, some of which hold no water and others that do. I have no idea whether my ideas hold water or not, but what I do works for me.

Here's the basic idea. Think of each gain stage beginning at its input, typically the grid of the preamp tube, but the first stage includes the input jack. Each gain stage ends at the cathode. All the grounds for a given stage, including its filter cap get tied onto the ground buss in one spot. I call it a node. It's like a starpoint. Who cares what we call it anyway? So the V1 node will be all the way to the right (random choice of right/left for discussion.) V2 node will be a little further left towards the power stage. V3 node will is next further to the left, etc., all the way to the PI node. That is the left end of the ground buss on the board. The power supply, power tubes, on the left side of the board, related filters for screens and plates, power tube cathode, output transformer and anything else I forgot about in the power stage goes directly to a bolt through the chassis. The other thing I do is run a wire from the PI end of the ground buss to that same single ground for the power stage. So, you might have a lot of stuff on that bolt. You can use some ring lugs to make it work. I make it so there is one single ground point on the chassis.

Others will tell you to take the ground buss to the chassis as close to the input jack as you can, so that power and preamp are grounded to the chassis about as far apart as possible. I believe them when they say it works and is the thing to do. It's just that I've had good results with what I do. So, there it is in a nutshell. You'll have to decide for yourself.

As an aside, look at the Trainwreck pictures. Ken Fischer, who everyone agrees was an amp genius, put all the filter caps in a single bank and didn't do anything like what I described. That worked well for him. So, go figure. I think the answer is that there is more than correct solution. If one doesn't work, then try another. Or search the internet for Randall Aiken's paper on the topic and after reading that, you'll be even more confused.

Just do something. Fire it up. See how it is. If it's quiet, you did it right. Good luck.

[alfi27]

Hi Phil,

Thanks for one of the best posts I have read on here ever! What you said is both quite settling and a tiny bit annoying, haha! You see, I had all of these self constructed myths in my head, like "filter caps are really finicky and if you "overload" one star ground point and somehow compromise their signal flow they might get damaged or something"... And another one is that "a really good ground scheme will help the amp keep noise low in situations like when it receives poor quality/poorly filtered power." I will certainly post gut shots, in here if it doesn't work and probably in a new thread if it does, as I'm fairly proud of my work considering I'm quite the novice

The fact is, there was nothing wrong with the original ground scheme, and it was with that scheme I originally had the thought "man this amp is quiet for such a high gain amp with no DC heaters!" I guess I got a bit caught up in that whole bulletproofing, or we could call it "German overengineering" mindset. Which is because the circuit I've done is from a Suhr amp, and man he can really build a bulletproof amp that is probably as close to military grade as you can get. Literally everything is "by the book".

So I'm thinking if I just redo the original ground scheme, and the chassis ground is coming from the start of the ground bus only instead of through every pot, it certainly shouldn't be less quiet than it was. I love the fact that if it's quiet it's good, and can just be left alone. I'll let you know how it goes, I'm bringing the amp on the plane with me in a few hours so hopefully it will survive to tell the tale if my new ground scheme works or not

[Stevem]

Hey, don't pick on Chimpanzee's there as one was our first Astronaut back in 61" lol!

[pompeiisneaks]

also I'd say 'well twisted wires' are over stated in importance for heaters. I just built a marshall dual with no twists in my heater wires, just kept them next to one another and routed them in/out carefully away from the signal wires and they have induced no hum that I can hear.

Maybe I'm going deaf

~Phil

[Phil_S]

What you said is both quite settling and a tiny bit annoying...

So I'm thinking if I just redo the original ground schem...

LOL, I don't know that much to have so much of an effect. If it ain't broke, don't fix it. I don't think moving the buss off the pots is much of a change. I've never soldered to the pots. I figure my soldering skills aren't quite that good and that I'll burn the pot. I tend to agree, you are overthinking this. It might be more critical in a very high gain amp, not in a Marshall circuit.

[alfi27]

Yeah, the thing is that the Suhr amp has DC heaters, and I was willing to go to extreme measures of nitpicking and perfecting every little thing to not having to wire that up... Good news is, the grounding scheme is totally quiet and I honestly struggle to understand why Suhr bothers doing DC heaters in this circuit. It has potential for loads of gain as well, so it's almost weird that it's so quiet.

Another thing though, I'm a bit tired of starting a new thread for every question I have so here goes: I prefer the 5k presence pot even in a high gain circuit like this, so I wired that up, and to copy the Suhr amp I removed the NFB resistor from the board and put it between the impedance switch and the resonance pot.

First things first, what does this movement achieve by itself (with a resonance pot naturally)? When I did this with the 5k pot there was DC on the resonance pot, and the low end very big/uncontrolled/flubby. It also sounded a lot harsher even with the presence at 0. That's one of the two changes I did after the last time I played it and loved it (the other being a cap switch, same value different brand), so I'll just reverse engineer it until I'm happy again, but I struggle to understand the theory behind moving the NFB resistor... I'm assuming the NFB resistor between presence and resonance will take DC off the resonance pot, and possibly break some kind of correlation between the pres and res caps?

If you need a schematic to visualise it or whatever, look at the Ceriatone Yeti or Chupacabra. The NFB circuit is the same except what I just said about the placement!

[Reeltarded]

The res and pres are both on the front? Fly the wire to the imp switch where it will lay on the head box shield. That wire is fantastically reactive. You might have a problem with the NFB circuit at thise pots if you changed it and it got crazy.

What are your resonance values? 1M and 4700p are only useful in most of my amps to around 3 with the bass somewhat moderated. Typical settings for me are Bass 3-4 Mid 5-7 Treb 4-6 Pres 4-6 Res 2-3. I only use the resonance to widen the Bass q and the B control for level.

The res control should scratch just like the old style pres. There is expected DC on those pots.

[alfi27]

Sorry, forgot all about this thread! Resonance is on the back (5600pf cap), replaced one of the speaker jacks and presence is on the front. I might decrease NFB (increase the value of the resistor) and get rid of the resonance knob. My other Marshall has the 100k/4ohm NFB and I dig it a lot, no need for a resonance knob with that one and it chuggs like crazy even with "just" 3 gain stages! After decreasing NFB and also increasing the preamp voltage to tighten it up a bit, I think it will be perfect. I agree that the bass knob has to be backed down a bit when dialing up the resonance, I like it around 3-4.