Early Dual Rectifier Problems

[AdmiralB]

I found an early DR (serial in the 200s) that doesn't work for fairly short money. I figured it'd be an easy flip since I found a previous non-working amp (serial in the 400s) last month and made some quick profit on it.

The problem with this amp is the same as the previous one - power tube arced between pins 2 and 3. In the previous amp, this destroyed the 100-ohm and 150-ohm ground reference resistors in the 6VAC supply. For some reason Mesa used carbon comps here..the ONLY place they used them - and they went up in flames. Fortunately there wasn't any collateral damage, and new resistors and removal of the carbon trace did the trick.

This amp is tougher - the 100-ohm resistors were OK but the 150s did a full burn, and I had to grind out about a 3/8" hole in the board. I got new components mounted (it also took out one of the rectified 6V filter caps and blew a transistor apart), and connected the broken traces under the board, and it fires up and holds steady. BUT, there's hum in the preamp (sounds like 120Hz pre-gain and 60Hz pre-master) and very little gain.

Going into the loop return, it works fine. But it seems as though at least some of the LDRs aren't working in the switching circuit. The 5W 220-ohm resistor that precedes the diode in the 50VAC-to-9VDC supply has about 18VDC and about 25VAC on the diode side and gets REALLY hot; the following 100-ohm (68 ohm on the Schematic Heaven schem) drops another 10V or so and runs cool. There's about 8VDC in this rail, which seems a little low.

I swapped the 6N193 LDR/Darlington, no joy.

Voltages on Q1-Q3 look OK.

What should I be doing next?

[AdmiralB]

OK, got some more narrowed down. At least a couple of the LDRs are flaky...I had one spare VTL5C9, but not enough, so I'm waiting on a package from Allied.

The loop, however, has been freaking me out and I finally figured it out. I've been doing all my testing with it in bypass mode. Turns out, bypass doesn't work right.

The first 250 or so DRs have a unique circuit board (mojo-filled, apparently). It's Rev C, and it's the first one, so it seems Revs A and B never made it out of Petaluma. They went to Rev D until serial 503, then changed to Rev F. The first 503 tend to command an inordinate price - I sold the Rev D I referenced in my previous post for an $800 profit.

In any case, the only schematic I've seen is Rev F. And it shows the loop working as I expect from these amps - with the loop control in bypass, a DPDT relay completely bypasses the entire loop circuit.

When the loop is in any other mode, the relay brings the loop in between the channel masters and the PI, and when it is switched out, it is done by LDRs that shunt signal around the send/return jacks - so the send level control and loop master volume are always in the circuit.

Turns out, on Rev C the bypass mode doesn't work. They screwed up the PCB and reversed the polarity on the relay coil, so it never bypasses the loop.

Must be why Rev C didn't last long.

[mykey]

[QUOTE=AdmiralB;34300]OK, got some more narrowed down. At least a couple of the LDRs are flaky...I had one spare VTL5C9, but not enough, so I'm waiting on a package from Allied.

we've never seen any mesa amp last very long.
now you know where the term "there she blows!" comes from.

[Enzo]

SO VActrol LDRs just automatically become unreliable when they sit in Mesa products, but not elsewhere?

What do you think Mesa is doing wrong to cause these failures?

What have you done in your shop that seems to reliably correct the problem?

[mykey]

For some reason Mesa used carbon comps here..the ONLY place they used them - and they went up in flames. Fortunately there wasn't any collateral damage, and new resistors and removal of the carbon trace did the trick.

This amp is tougher - the 100-ohm resistors were OK but the 150s did a full burn, and I had to grind out about a 3/8" hole in the board. I got new components mounted (it also took out one of the rectified 6V filter caps and blew a transistor apart), and connected the broken traces under the board, and it fires up and holds steady. BUT, there's hum in the preamp (sounds like 120Hz pre-gain and 60Hz pre-master) and very little gain.

now you have it figured out, carbon traces formed on the board rendering the entire unit FUBAR. Arching and burning between circuit traces is the primary problem with all mesa amps. And many times these traces form inside the fiberglass where you can't see the carbon.
taking the tubes and components off the board, use an ohm meter to find the carbon burns, particularly between plate and grid of the preamp tubes.
you may find 2 megs of resistance or less between the plates and grids.
you also realize that drilling the carbon tracks off the board is really the only way to fix the problem.
sound like you still have more carbon traces (burns) to track down.
another major problem is where resistors lay on top of the circuit tracks, tends to arch there too. lifting the resistors up off the tracks (to form an air gap) will solve a number of problems and noises.
takes a lot of time and effort to get one of these working again and most techs don't have a clue, you are already miles ahead of most who try to fix these (pathetic pieces of crap).
shorted diodes in the power supply can also cause that kind of hummmmm, etc, etc.....

[mykey]

SO VActrol LDRs just automatically become unreliable when they sit in Mesa products, but not elsewhere?

What do you think Mesa is doing wrong to cause these failures?

the design of the circuit board is inherently defective, insulation between circuit traces is woefully inadequate. mesa has known this for many years and has done nothing to correct the problems.

What have you done in your shop that seems to reliably correct the problem?

most techs can't fix them. drilling holes in the circuit board to stop the arching is one reliable approach.

[Enzo]

Whatever disagreements we may have, Mykey, do not insert words into text of mine within a quote. The third sentence qwithin the quote was not made by me.

You quoted Admiral where he says his Vactrols were flaky, then made your own remark about the amps blowing up. The Vactrols are not blowing up. This is not answering Admiral's question.

most techs can't fix them.

Speak for yourself.

[Steve Conner]

Welcome back mykey, did you figure out how to use the quote tags yet?

[AdmiralB]

Whatever disagreements we may have, Mykey, do not insert words into text of mine within a quote. The third sentence qwithin the quote was not made by me.

It is also not the problem with this amp. This amp cooked the ground reference resistors in the heater supply, because a power tube arced from plate to filament (and if I was guessing, it'd be because they labelled the 8-ohm jacks "8-16 ohm", and flyback resulted from a 16-ohm cab).

The area I had to grind out isn't anywhere near signal traces, so I don't think that's the problem. I won't be able to track the hum down until I get the LDR issues resolved, however.

[drewl]

Oooohhhh, you go girl!
I had one like this recently.....short between plate and heater of an output tube, fried low voltage power supply and made a general fried mess of things.
They are repairable if you have good pc repair skills,yo.
Word....

[mykey]

Oooohhhh, you go girl!
I had one like this recently.....short between plate and heater of an output tube, fried low voltage power supply and made a general fried mess of things.
They are repairable if you have good pc repair skills,yo.
Word....

yes, thats what happens. most techs who work on these things think the solder connections are bad and don't realize carbon traces form between the circuit traces, shorting the circuits out. in many cases the carbon traces are so small these are not visible. at one time mesa put ceramic beads on the resistor leads to lift the resistors above the circuit tracks. then they started coating the boards with insulating shellac. but the shellac did not prevent the arching. the factory techs did not do much better, and mis diagnosed the problems.
i have never seen an amp so expensive with so many design problems.
and yes they do blow up, and catch on fire. there have been many such reports for many years. i have seen several with charcoal on the boards.
i have worked on the first ones made and the recent ones, still the exact same failures.

[AdmiralB]

I got a recommendation to call Mike Bendinelli about the loop issue. So I did.

He maintained that the loop relay worked right on all revisions, and that there must be something wrong with the amp.

I stressed that the polarity was backwards on the coil, and that it could not be the result of mis-repaired damage upstream, since it would smoke the diode in parallel with the coil if that were the case.

His reply? "Polarity doesn't matter on coils".

Well, that rather depends on the relay.

[Enzo]

Polarity of a coil? I have to agree with Mike. The coil in a relay makes a magnetic field which attracts a metal plate, and that moves the contacts. There is no polarity. Apply current either direction, and the field forms and the armature plate pulls in. Some relays run on AC for that matter.

In a circuit, a DC relay coil will have a diode in parallel to snub spikes when current is removed. This diode is not in the relay, it would be a part on the circuit board. If you connected power to the existing circuit backwards, then that diode woul be forward biased and short out the supply, but that is unlikey to happen in a piece of gear in for repair.

The only thing that will smoke the diode is applying reverse power supply voltage to it. That will happen even with the relay removed from the board.

The coil is no different from a speaker voice coil. Reversing the speaker wires will have no effect on the electronics of an amp. It might not sound good when sitting next to other speakers, but that is acoustics.

[mykey]

I got a recommendation to call Mike Bendinelli about the loop issue. So I did.


His reply? "Polarity doesn't matter on coils".

Well, that rather depends on the relay.

the admiral is correct, DC relays are sometimes polarity dependent

[Enzo]

Provide an example and explain how that is the case.