The Cerberus project

[shane]

One thing you may want to watch is that the grid of the stage before the cathode follower will need a reference to ground while the relays change over, otherwise you may well get a loud pop while they switch.

Cheers
Shane

[romberg]

One thing you may want to watch is that the grid of the stage before the cathode follower will need a reference to ground while the relays change over, otherwise you may well get a loud pop while they switch.

Thanks! You are the second person to advise me that there may be pops with the switching. What would be a good cure? Would adding 4.7M resistors from the pole of each relay to ground be a good way to keep a ground reference while switching?

Part of me likes the idea of another gain stage entering the fray with the fanfare of a loud *POP*. But I bet my bandmates may think otherwise .

Mike

[shane]

If any of the poles already have a permanent connection to ground they should be OK its just that stage that doesn't.
So yes a 4M7 to ground off that grid would probably do the trick.

Cheers
Shane

[romberg]

Been measuring (and re-measuring) things and now have the basic dimensions for the new holes in the chassis (a watts tube audio jcm-50 kit). Still need to decide on exactly where to put the footswitch jack and rotary selector. I'm debating putting it way over by the power switches. Once that's decided, I can order faceplates.

The circuit board needs some more tweaks to move things closer to the new tube socket. And I gotta figure out exactly where to put the relay boards (either by the tubes or by the pots).

Mike
[/img]

[Smokebreak]

Lookin good! You can put the relay footswitch on the back panel. I had a Lot of depth in my chassis, and the relays were an afterthought, but I had them between the board and pots(Colossal's boards come complete with standoffs) on the chassis.
More fuel for the fire.. You could switch out the normal master to run it like a non-MV amp when in 1959 mode. Push pull pot? I'm working on something similar, and need the panel space, so I'm thinking about putting the parallel input stages on a dual gang, then internally jumpering on a switch. Naylors do something similar but with series stages.

Edit: I would put the channel switching rotary in a spot that allows for the shortest runs. Could get tricky over by the power switches.

[romberg]

Lookin good! You can put the relay footswitch on the back panel.

Yea. That idea is starting to become more appealing (see below)

I had a Lot of depth in my chassis, and the relays were an afterthought, but I had them between the board and pots(Colossal's boards come complete with standoffs) on the chassis.

That is now what I'm planing. There just flat out ain't room behind. Unless I did something goofy like mount them on the back wall. Thanks for the info that they can work over by the pots. I've never done switching. So, I was concerned about the relay currents radiating into the tone stack, etc.

More fuel for the fire.. You could switch out the normal master to run it like a non-MV amp when in 1959 mode. Push pull pot? I'm working on something similar, and need the panel space, so I'm thinking about putting the parallel input stages on a dual gang, then internally jumpering on a switch. Naylors do something similar but with series stages.

I'd thought about a switchable master volume. Of course since I'm gonna fit a pre PI master volume and a post PI master volume, I can just crank the pre up when in plexi mode to take it out of the circuit and drive the PI hard. Homebelly's comment got me thinking about balancing the volumes a bit. But I think I'd have to switch between two post phase inverter mvs to do that. And... I think I have enough switching for now .

Edit: I would put the channel switching rotary in a spot that allows for the shortest runs. Could get tricky over by the power switches.

I've sketched out a first draft of the preamp layout. And I'm now thinking of keeping the rotary switch by the input but moving the foot switch jack either over by the power switches or to the back of the amp. Mounting it where I got it now won't let me move the relays any farther over to the right (which I think may work better).

Anyway, here is the current working draft. Still waiting on a transformer. So, I still have time to tweak.

Mike

[Smokebreak]

Right on man, these Marshall circuits sure are fun.
Since you're goin all out and putting a clicker on the floor, switchable normal masters linked to trigger to your switching is really nice. They just go in parallel off the T pot, PI common, wipers to switch. You can use it to match channel levels, or use it as a serious solo boost. It's about the best boost pedal you can make for $1.50 , and only 1 more knob

[romberg]

Well, I think I've nailed down a location for the relays and finished a layout. I may have to shield a couple of runs to the grids of V2 and V3. And I think I got the grounding more or less like Merlin describes in his book. I'm still a bit unsure about where to ground the relay boards. So, I'm just running a wire back to the "big" ground the center taps connect to.

I still have 1 unused relay board and some space sitting next to the post master volume. I'm thinking about switchable post phase inverter master volumes. But perhaps that would be too much?

Now just waiting on the power transformer to arrive...

Mike

[Colossal]

I still have 1 unused relay board and some space sitting next to the post master volume. I'm thinking about switchable post phase inverter master volumes. But perhaps that would be too much?

Never! That would actually be very useful. One for normal and one for lead boost.

[Smokebreak]

Worse than an unused triode in some circles I'm a fan of the switchable normal masters, but interested to see how PPIMV would switch out.

Oh and as I've learned from Colossal, 1 ground for the relays back home works great.

[romberg]

OK, I'll have to think about switching a bit more. The tough bit will be keeping the bias voltage on those power tubes at all times. My first guess is that it should be doable with one DPDT (relay). The realy and the pot sure would tuck in there nice

Mike

[romberg]

How goofy does a setup like this look? By my math the phase inverter should still see a load of 218k. And I've moved the 2.2M "safety" resistors to the pole side of the relay where they will now for sure have the job of keeping bias on the grids during the switching. Does this seem safe? Any alarm bells going off to any more experienced builders?

Also, the grid leak resistors will go to zero as the master(s) are turned down. I guess this is OK as most of the other master schematics I've seen do this. But, I'm not very knowledgeable about power tube operation.

Anyway I may throw this into the mix unless someone spots a reason why I should not.

Mike

[Colossal]

Anyway I may throw this into the mix unless someone spots a reason why I should not.

Mike

Hi Mike,

Placing the MV switching relays where they are shown will result in them carrying the bias voltage. I would place them in front of the master volume pots and route the output from the phase inverter inverting and non inverting outputs to their respective master volume pots when the relays are triggered. Placing the relays in front of the pots, they will only see the PI output AC signal as they are protected from DC by the coupling caps.

Also, any reason you are going with the 470k/470k grid leaks and then the 1MA/2M2 pots? The 1M pots will be more sensitive to noise than 250kA/dual and the usual 2M2 safety resistors (sans the 470k pair).

Just some thoughts.

[romberg]

Placing the MV switching relays where they are shown will result in them carrying the bias voltage. I would place them in front of the master volume pots and route the output from the phase inverter inverting and non inverting outputs to their respective master volume pots when the relays are triggered. Placing the relays in front of the pots, they will only see the PI output AC signal as they are protected from DC by the coupling caps.

Hmm... I think the relay would still be at the bias voltage even placed on the other side of the master volume pots as they won't block DC. Maybe you are thinking about adding in additional coupling caps? I've thrown together another small schematic which is what I think you are describing below. But I get confused as to what to do with the four center wipers.

What I've drawn just joins them together at the grid stoppers. But this will cause funky interactions between the controls (if one is at 0 then the other will have no effect). I think you would need an additional relay after the master volume pots. Or, I may just be missing something obvious .

Also, any reason you are going with the 470k/470k grid leaks and then the 1MA/2M2 pots? The 1M pots will be more sensitive to noise than 250kA/dual and the usual 2M2 safety resistors (sans the 470k pair).

It took Mouser eight weeks to get in the single 250k dual audio I have now. So, I was thinking of 1 megs (with the 470ks in parallel to get close to 220k). But maybe I should look harder to source some dual 500ks or 250ks.

Mike

[Colossal]

Maybe you are thinking about adding in additional coupling caps? I've thrown together another small schematic which is what I think you are describing below.

Mike,

Whoops, yes, that was the most important part: additional coupling caps!

This could be another approach:
https://tubeamparchive.com/download/file.php?id=32488

This has the advantage that the tubes always see bias voltage no matter what master volume is being selected, or if there is a failure of some kind.