Why did this Marshall 8008 start smoking

[rockhorst]

I was at a band rehearsal today, checking out some preamp pedals I made with a power amp my friend owns. Stuff was sounding very nice but at some point we noticed that the level was actually quite low (loud, but not loud enough for a full band context). The power amp is a Marshall Valvestate 8008. So we turn it off and have some discussions about impedance when at some point somebody point out that something smells burnt and there's smoke coming out of the amp. It probably was already smoldering when the power was on, but the smoke started coming out a minute or so after the power had been off. After it had a cool down outside we opened up the top plate and there were burn marks on the plate and resistor R8 seems to have been the culprit (black burn marks on the PCB around the resistor). On the schematic R8 is located in the bottom right hand corner, below the diode rectifier and connected to the chassis. It's a 1/2W 10 Ohm resistor.



My guess is that something else failed which caused the resistor to handle more power than it was designed to do. Any clues?

[Slowpoke101]

It started to smoke when the amp was off? Unless there are some wiring problems with your building's wiring (mainly the earth wiring) there is no way that the amp itself is at fault. The power required to smoke the resistor would have to be external to the amp. The inputs to the amp and the outputs from the amp.
The most likely cause would be an effect that was positive ground (germanium fuzz, etc) using its own power supply that has its negative connected to the mains power earth. The amp would also have to share its mains earth with that power supply. Current could then flow from the effect via its cable (shield) connection to the amp. The amp's 0V rail (referred to as GND) connects via the 10R resistor to the amp's chassis. The chassis should be connected the the mains earth. This completes the circuit and smokes the resistor.
I'm most likely wrong but have a look. Maybe someone else has a better clue.

Cheers.

[PRR]

R8 smoked strongly suggests a live "ground" at that venue.

But that would not make it play weak. (And R8 may have been toasted long ago.)

[GiovannyS10]

Hey dude!
First of all remember to use a in-series light bulb when turn it on again, and if you want to troubleshoot too.
I guess that the R8 isn't your main problem. You said the sound was going low slowly, right? In the end you continue hearing a low sound or any sound? Or turned you the amplifier off before check it? Before you turn off the amplifier, the sound starts to be a bit noisy? Like an little overdrive? Both channels are with low sound or just one?
Just check R53, R54, R66 and R65 and TR18, TR4, TR1 and TR14, that last ones are just a precaution. They common burn and cause these symptoms... Low and noisy sound. And can produce a bit of smoke when burn.
I would check TR4, TR8, TR15 and TR6 too. When it burns, normally the around resistors gone too.

If all this are okay, please turn it on *using a series light bulb* and measure the voltage between C11+ and C14 +; and between C11+ and C14 -. And let me know the result.

Remember: i am considering that you plugged your amp in the current voltage, and it was correctly grounded... And considering that its burnt while was on, and you just saw the smoke after turn it off. Not the smoke starts when it was off Hahahah

Good luck.

[Rob Strand]

As an alternative to PRR's comment.

If you were using a positive ground effect and it was powered from a power supply who's negative terminal was connected to ground, then that power supply would effectively connect across the 10ohm resistor.
The effect might even keep working. 

You will need to check a lot of things to determine the cause.

At the end of the day I'm assuming the amp still works?
It might have a bit more buzz if that 10 ohm resistor is fried open.

[rockhorst]

Thanks for the input so far. Let me elaborate and clarify a bit. Some stuff below is based on memory and assumptions.

It's a solid state stereo power amp in a steel or aluminum chassis. The amp wasn't rack mounted but was just lying bare on a 4x12 Marshall speaker cab. The 4x12 was most likely 16 Ohms and the amp is designed for 4 Ohms or more. The impedance mismatch could at least in part explain the lack of volume. The amp sounded fine otherwise, no buzzing, slow changing or dying of the sound was registered.

It states that input/output B is the mono option. The guitarist has however also used only input/output A for part of the evening. This could have been most of the evening or just before the smoke was noticed, I'm unsure. This may have had some impact? I'm not too sure. I have the updated 120W version of this amp and there it seems to make no difference which terminal you use and the manual is actually very vague about it.

I believe the venue had proper wall sockets with safety ground, but I'm not entirely sure.

About the effects: all stomp boxes were sleeve negative, tip positive (so +9V) and they were daisy chained and standing on top or next to the power amp. When someone pointed out the smoke I noticed a bare daisy chain plug lying on the amps chassis. The wall wart used was a One Spot, so a two prong device without safety ground. With one ground (sleeve of PSU) touching another ground (amp chassis) my initial thought was that this may have setup an unexpected circuit, for instance for the capacitors to discharge when we switched of the amp. I'm not sure if this is a valid idea or total bogus please chime in.

I have not turned the amp on since, I will at least first replace the 10 Ohm resistor with a heftier version.

[Rob Strand]

I noticed a bare daisy chain plug lying on the amps chassis.

Sounds like a possible culprit.

If the tip (+) of the PSU touched the chassis of the amplifier then that plug will connect +9V  to "EARTH" and the hence the 10ohm resistor. Assuming the amp has isolated sockets, the effects will connect the PSU 0V to the internal "GND".  So you end up with 9V across the 10 ohm resistor.

Since  power P = V^2 / R = 9^2 / 10 = 8.1W   which will fry the resistor.   The PSU would be pushing out 0.9A, if it can supply this much.

[rockhorst]

Yeah, the more I think about it, the more likely this is. I know R.G. once warned about something like this, and since then I tape off bare daisy chains. I indeed think this is the most likely culprit now I've got everything lined up (which is another useful thing about starting a topic...you get to arrange your thoughts).

Thinking about it a bit more, I think it could actually also explain why it only started smoking after we turned the amp off, not while the amp was on (which it was for about an hour or so) and maybe even why there was a little loss of output volume?

[Rob Strand]

I think it could actually also explain why it only started smoking after we turned the amp off, not while the amp was on (which it was for about an hour or so) and maybe even why there was a little loss of output volume?

The way these things usually happen is something sits there unnoticed in precarious position then some random event pushes it over the edge.  For example when you turned the amp off you  knocked something, or while you were talking you just put your hand on something and bumped it, or kicked the cable and it tugged something.    If the PSU voltage dropped under load and if the effects had transistor gain stages then it's possible the output level from the transistor dropped.   It's all speculation but not unfounded.     If you were chasing down why a plane crashed you would re-do the whole experiment.  You might set up the same effects chain and say put a separate 10 ohm load on the power supply and see what happens.    I do this type of thing chasing down bugs or weird behaviours in equipment.   It all boils down to how much you want or need to know!

[Mark Hammer]

It took all my resolve to NOT respond to the subject of this thread with "Because its friends were a bad influence".



Shame on me.  I should know better.  Carry on.

[rockhorst]

I was just laying a trap for you Mark, well spotted