Marshall 8040 Valvestate repair

Repairing my daughters amp ready for sale.
It worked several years ago when put away in the loft.

Power up, either nothing or it powers up.
Example, on the bench, plugged in to measure DC volts at smoothing caps - nothing.
Unplugged while considering what to check next.
Plugged back in to measure transformer AC at board.
Did nothing else, nothing moved, no switch operated. Only plugged in to wall.
Power on plop from speaker, hum, we have power.

This is likely not be the only fault.
Before dismantling I had sound from boost channel but nothing from clean channel.
Boost channel was much quieter than it should be.

I have seen a suggestion to reflow solder joints as allegedly these amps used early lead free solder that is susceptible to failure.

Question, joints look ok to me.
Are there visible signs of failed solder joints or should I go ahead and reflow them all regardless?
The short answer is yes, you can typically spot a failed joint with a magnifying glass.

However, your description doesn't appear to have established that that is the problem you are facing, so definately not worth the attempt to reflow everyghing at this point. If it's intermittent joints, wiggling should at least make portions of the circuit come alive.

I think you just need to methodically go through troubleshooting, i.e. incrememental tests figuring out what is working, starting on the power side first. For example, if not powering up, is power getting to PT primary? Secondary? Filter caps? PS output? Power to main board? Daughter boards? Then move to signal.
  • Like
Reactions: 1 user
You are not describing one problem but various unrelated ones.

Start by checking the mains wiring: plug/cable/IEC connector/fuse/switch/primary wires.

You must have confirmed continuity, no measurements are possible without electrical power.

Then measure transformer secondry voltages, DC rails, if not remembering wrong there should be a small HV transformer for the tube or a main transformer HV winding, guess flaments are DC, fed (and attenuated to 12V) from one rail or you might have a 6.3V winding.

Post schematic here if you want real answers, not "it must be some thingie inside".

Apparently power chip is working, good because it is unobtainium for decades now.
Thanks for answers so far.
I didn't post schematic as I was not really looking for specific fault finding advice, just a general question on PCB soldered joints.

It is 1 large PCB, single sided, thru hole.
Working on is tricky as power tranny is bolted to case, short wires. IEC inlet is riveted to case, everything hard wired, not socketed.
I have PCB inverted & solid plastic sheet between PCB & metal case to prevent shorts.
Thus I don't fancy wiggling & poking too much.

You say unrelated problems but the common factor so far is intermittentness.
I reflowed the power supply side of the board and now have consistent power.
On this basis, I will reflow other areas that are subject to heat or weight - valve socket, large caps, power resistors.
I will (for now) avoid reflowing around ic pins (lower risk of failure, chance of overheating chips in the process).
If all seems ok I will replace pots (they always were noisy).

On the 'minimal improvement' side, I intend to fit input protection diodes to protect the input ic and (if no objections) a third PCB standoff.
As can be seen in the photo, a large area of board is completely unsupported.
Maybe it is supposed to flap about as the reverb is fitted in that corner?


  • DSCF6498.JPG
    498.7 KB · Views: 176
  • 8040-iss8.pdf
    100.9 KB · Views: 110
OK, umpteen joints reflowed.
Didn't take long and I consider it time well spent.
I now have sound from both clean and boost channels.

I now feel confident that replacing the pots is now worth while.
VRI, clean channel volume goes open circuit at slightest provocation resulting in earthquake rumblings.
Remainder are mostly noisy or stiff, might as well do the lot as leave the 2 or 3 that may be ok.

Last question, B+ smoothing cap, large one in picture.
Is this what a bulged cap looks like?
I think the photo shows the top is crowned.
All the others, the top is flat and below the poly sleeve.
Would you agree I should replace it?


  • DSCF6499.JPG
    319.4 KB · Views: 106
Yes, the replacement cap will be a perfect fit. .
Parts ordered, Cricklewood Electronics. Most helpful when I asked them to measure the cap for me.
Half the price of Guitar parts suppliers for the pots and I'm sure they all come from the same factory.

As a final test, I removed the dodgy clean channel volume pot.
Soldered a couple of fixed resistors in its place.
Tested - no horrible noises therefore proved that particular fault was entirely down to the dodgy pot.
Didn't want to wait for parts to arrive then find there were more problems to solve.
So I have:
Replaced all the pots.
Replaced the possibly bulging elco.
Fitted 1.5K resistor in series with pin 3 of the first op amp and 1N4148 diodes from pin 3 to + & - rails for input protection.

All now works.

Just one niggle:
Clean channel volume, I can play at full volume (in a small room) and it is loud but not unbearably so.
Boost channel is VERY loud and for comparable volume the gain & volume controls are set around 1/4 rotation.
This doesn't seem right.
1N4148 (DO 35)
1N4148 (DO 35)
In general, clean channel on 6-7 should reach some clipping because of power stage swinging rail to rail and definitely crunching on 10.

Master volume (if present) on 10 of course.

Not set in stone by any means, but common design practice, "extra" gain allows for weak-ish pickups to still drive amp to full power. .
Any ideas why the clean channel should be relatively quiet?
I can play with guitar full volume & clean channel full volume and it's not unreasonable listening in a small room.
In fact you have to set volume half way to hear speaker rather than the strings on the guitar.

Boost channel gain & volume pots would be set around 1/4 for comparable volume.

The amp has been stored for 15 years and I cannot remember what it was like before but given the footswitch switches between clean & boost for effect then volume should surely be comparable.
Apart from the volume it seems to be behaving.

From the schematic, the boost channel picks it's signal after IC3A so that must be ok. VR1 is correct 100K value (measures 97K in circuit). That only leaves IC3B & IC4A & surrounding components.
All resistors measure ok.
Jumping from pins to power amp in, we have gain after IC3B, loss through passive tone controls and some gain after IC4A.

For full disclosure, 220K pots have become 200K & 22K have become 20K (what was available). I doubt 10% variance matters.

Any ideas how to progress?
Can you measure the voltage across the load? This will give you a true indication of the power.

I have the 8008 and it really can do 80W almost exactly into 4 Ohms per channel.

I would expect you to get 40W into 4Ohms, so you should be able to measure approx, 12.5V RMS (~18V peak) on the output with everything cranked.

I would also be good to quantify the input voltage to make sure you are driving it properly.
  • Like
Reactions: 1 user
Time for an update.
The amp has been parked under my workbench for months while I indulged one of my other hobbies (see pic).
I don't have space to do 2 jobs at once.

I soldered a cheap signal generator kit form Ebay. A fun project.

Using this I injected a signal and got loud volume injecting to the right of R31 and vastly reduced volume injecting to the left of R31.
I assumed IC3B to be the culprit.
Replacing it has restored full volume to the clean channel.

To summarise:
Reflowed soldered joints. This got the dead amp working.
Replaced all pots - all scratchy, VR1 intermittent (could that have caused IC3B to fail?).
Replaced IC3B. Fixed quiet clean channel.

It has been a very satisfying exercise.


  • DSCF6568[1].JPG
    521.4 KB · Views: 29
  • Like
Reactions: 1 users