Question about DC filament supply (Madison)

[gmoon]

I picked up a Madison Divinity head recently (for $100, and the seller threw in 4 extra preamp tubes  ). I'm not a metal-head, but couldn't resist the amp at that price. I know controversy swirled around these amps in their time, and the company was short-lived.

It has a couple problems--the last owner pooched the bias trimpot on one side, breaking and partially popping it off the board. I've replaced it.

Anyway, here's my problem: the amp has a DC filament supply for two of the preamp tubes, but 3 of the 4 filter caps are shot (this amp could only be 6 years old or so, as they only made amps from ~2005 to 2010):





The filament supply circuit looks like this (From a schematic the Madison was partly based on):





Note C71 and C73. If I compare this to the JCM 2000 DSL DC filament supply, the caps connected to the chassis GND with opposing polarity, unlike the schematic above. Since the reference to GND is set by the two 100 ohm resistors, both poles of the rectified supply can't be positive relative to ground (right?)

However, the PCB as wired conforms to the Marshall schematic, and ignores the parts mask on the PCB itself (one of the caps is reversed, C71), confirming my suspicions. I strongly suspect that's my clue, and the Marshall way is the right way.

I should NOT follow the schematic, yes?

[EL34]

That diagram looks correct to me.
The tops of all 4-3300uf caps are more positive than chassis ground.
You can double check by measuring the left side of the bridge with respect to chassis ground.

[PRR]

Schematic is wrong.

Omit C71 C73.

No "ground" on the DC end of the heater system. (It will be approximately balanced around ground via the 2x100r in the AC section; ideally it should have a separate winding for the DC circuit).

[PRR]

C71 is clearly reverse-voltage as shown.

Ripple current is very close to rating on typical caps. It may be closer to 1,000-hour torture-test conditions than it should be. It really wants more physical bulk to shed heat.

How precious is the amp? How roomy? A small separate DC supply would be a fine idea. If you can wire V1 V2 in series, 12V 0.3A is an easy spec to find, possibly in high-quality surplus.

[HotBluePlates]

Schematic is wrong.

Omit C71 C73.

No "ground" on the DC end of the heater system. (It will be approximately balanced around ground via the 2x100r in the AC section; ideally it should have a separate winding for the DC circuit).

Doesn't it look like they got "too clever"?

It seems like C71 and C73 are intended to be anti-polarity, like a non-polar e-lytic, which then spans the d.c. wiring and has the chassis as a mid-point reference. They're balancing to ground with caps instead of resistors, which seems like a poor idea given e-lytic tolerances and the fact it just ain't needed.

[gmoon]

Thanks, guys, plenty to chew on.

Just to muddy the waters, I'll throw in the Marshall JCM2000 schematic...since C71 was (intentionally?) reversed on the PCB.

[PRR]

> throw in the Marshall

If that's what they copied, they got it wrong.

Yes, over-clever.

[gmoon]

Yes, over-clever.

OK. 'nuff said.

The original orientation (like the Marshall) might be OK, if I'm getting the gist.

Easier than a separate DC supply re: PRR's suggestion, at least for now. Which actually is do-able, the chassis has a bit of room.

The replacement caps I bought are somewhat higher speced than the originals (16V, 105C instead of 10V, 85C). But otherwise are the same, so they fit right on the board.

[stingray_65]

I'm wondering if I'm reading the schematic in the first post right.

Are the heaters fused?

Isn't it a bad thing if there is HT to the tube and no heaters on? (If the fuse blew)

Ray

[gmoon]

Are the heaters fused?

Isn't it a bad thing if there is HT to the tube and no heaters on? (If the fuse blew)

Yep, there are two fuses on the heater supply.

One of these was blown when I got the amp. The fuse holder is directly beside the broken trim pot (now fixed) and I think when the previous owner broke it, the screwdriver probably shorted the fuse.

Bad thing to fuse the heaters? Don't know, but someone will. The Marshall has 'em, too.

[sluckey]

Isn't it a bad thing if there is HT to the tube and no heaters on? (If the fuse blew)

I've never done this but I can't see that any harm would come with B+ but no heaters. The tube plates and screens would just look like be a chunk of metal if cathodes are cold. No current can flow until the cathode is heated sufficiently to boil off electrons.

[gmoon]

OK, it's alive.

My initial impressions were so-so. The clean channel was really weak, kinda noisy, the reverb weak and the gain channel was just OK. Then I reverted to the preamp tube compliment that the previous owner used--swapped out one of the 12aT7s for a 12ax7. Huge difference. If anything, it's quieter than before. The clean channel volume is now more on the level of the gain; they both sounds really nice now. Maybe the 12at7 is bad (I have some 12ax7s, 12au7s and 12ay7s, but no 12at7 tubes on hand). I'll have to try a good one.

I hate to gush about amplifiers--it's all so subjective. But it exceeds my expectations, although it's still not my "type" of amp. It's freaking loud, of course. Too loud.

These early Madisons are really just prototypes. A couple of the controls don't work (components have been removed from one board) and there are kludgy mods done at the Madison shop. The original owner of this amp was a Madison endorsee (got the amp for free) and said it was modified for him personally. Maybe they custom modified the circuit for the preamp tube change.

Thanks for the help, guys.

[eleventeen]

Not enuf knobs!

[gmoon]

 

My favorite thing about all those knobs--the original owner cut/broke the shaft off a tone control so it could only adjusted with a screwdriver. Nobody messes with my Mids!

(A cool dude, though. Wouldn't take more $$ for the head, even when I offered.)

[gh1]

Hey, I tried to buy that amp but never got a response from the seller. Lucky you   

[gmoon]

Yep. Craigslist is all about timing. The ad was top-of-the-page when I saw it. (Does the CL dance  )

A note about finding information on recently defunct companies-- I found a lot on the wayback machine at archive.org, including a PDF of the owners manual.

The cashed websites also have the "factory" info on biasing (in case anyone else is looking for it):

-- EL34: 35-40mv per tube (70-80mv per outside pin)
-- 6L6GC or 5881: 40-45mv per tube (80-90mv per outside pin)
-- Ruby 6L6GCMSTR: 50-60mv per tube (100-120mv per outside pin)

Other than the filament supply caps, the other mod I did was to fix the broken bias trimpot. Rather than try to find an identical "side adjust" trimmer, I used a top-adjust I had handy. To make that work I drilled a hole in the chassis for access (and a matching hole on the other side, in case I decide to replace that one). You can see the unchanged trimmer on the right, and barely see the replacement on the left.

A top-adjust trimmer is a lot harder to break off the board, I reckon.

[hehasfallen]

I'm hoping one of you can answer a question about this amp: I need a foot controller for it, yet they do not make them anymore, any suggestions? please help

[HotBluePlates]

If someone posts a complete schematic, we can tell you if the footswitch is anything more than just switches outside the amp.

[John]

http://music-electronics-forum.com/t25068/  <--- forum page where I found this, so you can see comments, and

http://www.fjamods.com/Temp/ashton_vp100.pdf  <--- the complete (I think) schematic. How accurate all this is, I have no idea, but hope it helps!

[HotBluePlates]

Okay, it looks like jacks 9 and 10 (JK 9, JK 10), on the left side below the input jacks, are where you'd plug in your footswitch. It appears there are panels switches to do the same functions, and they simply make a connection to ground.

So, it looks like you could simply have 4 switches, which close to make a connection between the jack hot and shield. This is like any old-school footswitch out there.