Quick question about the Marshall Treble Peaking circuit...

Cirrus

Curiosity got the better of me and I opened up the Mesa Stiletto combo to have a look inside. Before V2 there's the parallel resistor/ cap arrangement similar to in a JMP/ JCM etc.

However, there's actually two caps in parallel with a 430kohm resistor; a 750pf and 150pf. I know 900pf caps aren't that common, but surely a single 1000pf would have been as good? Or is there a cunning reason to use two like that? 

ChrisMusic

I'm not sure if the Stiletto Stage 2 circuit is the same as either your or my Ace combos, but here is a link to the PDF.

http://www.thetubestore.com/lib/thetubestore/schematics/Mesa Boogie/Mesa-Boogie-Stiletto-Stage2-Schematic.pdf

and a link to Marshall's 2204 schematic / circuit diagram.

http://www.thetubestore.com/lib/thetubestore/schematics/Marshall/Marshall-JMP-Mastervol-50W-2204U-Schematic.pdf

I'll wait to find if there is indeed a cunning reason though, apart from nailing the frequency response...
(but I suspect they are fairly wide tolerance components anyway?)

Cirrus

OK, interesting, so they're actually more of a bright cap system there in parallel with the gain knob and a 470k resistor (the one in mine is labelled 430)

And there are a couple of treble peaking cap/resistor combos just on that first page... this schematic is gold, thanks so much!

ICBM

They certainly do use a lot of very non-standard values in it! I can only assume that it's been tweaked by ear with trimmer pots in various critical positions and then the values measured.

Cirrus

Never let it be said that I'm not a moron - while chatting to the drummer and putting the chassis back in the cab, all excited to hear what my changing of the tone stack slope resistor had done, I put the valve rectifier in one of the power tube slots and vice versa.

Funnily enough, when I turned it on it initially fizzled a bit (I assume this is what rectifiers sound like while they're blowing) but then sounded ok and behaved almost as normal, except presumably on 25w of power so I could tell something wasn't *quite* right, walked over and smelled the smell of the screen resistor. Turned it off straight away, probably only on for 30 seconds all in, but it had melted its solder and fallen right out of the pcb, a charred cylinder.

Aside from the Valve rec (is it safe to run without one if I don't use that mode?) and said fried resistor, is there anything else that's likely to be fucked? Will the EL34 have blown? Not sure what pin corresponds to what...

I'm such an idiot. 

ICBM

How is the rectifier valve socket wired? On the Dual Rectifier, the unused pins are connected as well, which increases the potential damage to the EL34... I haven't seen inside a Stiletto.

I think the OT should be OK because the rectifier valve doesn't have pin 3, and the bias supply should be OK because it doesn't have pin 5, so it can't have shorted those.

Cirrus

Rectifier socket underside - looks like there are 4 pins which aren't connected to anything...



And just for funsies, the bombsite;


Quite the scorch marks. According to the schematic there are two "5352" diodes connected in series between the B+ and the blown resistor. Think they're underneath the board...

ICBM

Good - no connection to pin 7. That means the EL34 has probably survived.

EL34:

Pin 1 - suppressor grid, pins 2 & 7 - filament, pin 3 - plate, pin 4 - screen grid, pin 5 - control grid, pin 6 - no connection, pin 8 - cathode.

Rectifier:

Pins 2 and 8 - filament/cathode, pins 4 and 6 - plates. Pin 1 is usually present on the valve but not connected, 3, 5 and 7 are not present.

So… EL34 in rectifier socket - didn't power up, so all it had to do was withstand a high voltage from the screen to the (cold) cathode, which it should have been able to.

But - rectifier in EL34 socket… because the filament is centre-tapped to ground, and the cathode is grounded in a fixed-bias amp, there's 3V between pin 2 and pin 8, which is enough to partially power up the rectifier. It would then conduct directly from pin 4 to pin 8, which blew the screen resistor. If you're lucky, the rectifier valve itself might have survived because the resistor would limit the current, but I wouldn't count on it.

The amp should still work with no rectifier valve if it's set to solid-state, but obviously you'll have to replace the screen resistor first - and it's also quite likely the Zener diodes are blown.

Cirrus

Ok, thanks @ICBM . Those Zeners - Since I can't see them and only have a basic multimeter, I measured the resistance across what I think are the pins. Each of the two I *think* are for the non-blown EL34 sockets (some of the traces are on the underside, I presume) measured 800 ohms, the two I *think* are for the blown one measured 300 and 450 ohms. Aware I could be barking up the wrong tree there.

Is there any reason they need to be on the underside of the board? Would it matter if I soldered replacements in on the easy side? 

ICBM

Cirrus said:

Ok, thanks @ICBM . Those Zeners - Since I can't see them and only have a basic multimeter, I measured the resistance across what I think are the pins. Each of the two I *think* are for the non-blown EL34 sockets (some of the traces are on the underside, I presume) measured 800 ohms, the two I *think* are for the blown one measured 300 and 450 ohms. Aware I could be barking up the wrong tree there.

Is there any reason they need to be on the underside of the board? Would it matter if I soldered replacements in on the easy side? 

That doesn't sound good. If they've blown they're probably shorted (which would explain the lower resistance, although that still seems surprisingly high if they are) so they will need to be removed first - although you may be able to do that by heating the joints and poking the legs through with a pin so they drop out. You can certainly mount the replacements above the board if you want.

Does your multimeter have a beep continuity tester so you can work out where the traces go on the underside, even if you can't see them? You'll need to do that to be sure you orient the replacements correctly, if you can't see how they're fitted now.

Cirrus

Yeah, it does - I'll make sure I'm right before I start randomly unsoldering things. I'd kinda hoped the resistor being 3.75w and the diodes being 5w would save them. Of course, it could be that my late night quick check was wrong, and I was measuring across both diodes in series to get 800w then only measuring across 1 each to get ~350-400... I'll investigate. Thanks as always for the help @ICBM, you're a good egg.

Cirrus

I can only try it at tv volume right now but I think I've fixed it...

ICBM

Make sure the voltages on the screen resistors are the same if you haven't replaced the diodes - if the diodes are shorted, the amp will work but the screen voltage will be too high.

Whether the resistor died before the diodes depends on the current draw rather than the power rating, but I do think there's a good chance they've survived.

Cirrus

I replaced them - they were definitely reading lower resistance then the other pair so *something* wasn't right. I'll measure the voltages later just to be sure all is well.

Cirrus

Well, all seems to be good. I lowered the tone stack slope resistor from 39k to 25k to get some more upper mids relative to the very top end zing. Sounds pretty good, am pleased.

And it definitely sounds better when the valves are all in the correct sockets...