Potential Manufacturing fault in Marshall DSL40s

[SoulFetish]

In fact, I would say manufacturing fault is likely...

Within the past year, we've had 2 different DSL40's come in with the same failure mode. In our examples, the failure mode shows up as follows:

- No output in Low power mode, and very low output in High Power mode.

When pulling the chassis, visual inspection shows R49/4k7 power supply dropper is burned and measures basically open. After replacing and installing known good tubes, amp shows tested good briefly until I noticed rapid discoloration and smoke arise from the replacement resistor. Amp was turned off immediately, and we pulled our notes from the previous repair.
We compared the current amp on our bench to the notes from the previous amp. Further inspection showed that the board had arced through the trace and bleeder resistors in the stacked filter caps circuit, just as it had in the previous amp.
During the course of service, I made mention of the discoloration on the glue used to secure the components to the PCB. When we discovered that the amp had the same exact mode of failure, my boss asked the question "I wonder if the glue has anything to do with it". So, I looked at the PCB again, and realized that the discoloration in the glue only shows where there is PCB trace directly underneath. This shows a clear pattern indicating that the glue is reacting to the high voltage, current (heat), or both.
The arcing through from trace to resistor occurs where the trace runs under both bleeders, and glue is deposited.
I will upload some pictures to show the condition, and schematic when I can.

If anyone else has observed this as well, please feel free to share.

[SoulFetish]

Here are some photos to illustrate my description. The first one shows the damage to the trace and resistor from the resulting failure. The rest show a clear reaction in the adhesive directly over traces carrying high voltages and power supply ground currents:

[loudthud]

Generally, the green solder-mask is pretty tough stuff. It can take the heat of wave soldering and whatever is used to remove the flux after soldering. Probably water soluble with the surface mount on board. Not sure what the residue is by the OC Pass sticker. Will alcohol and mop up with a tissue remove it ?

Is this amp ROHS compliant with lead free solder ?

[Mick Bailey]

Marshall amps are all ROHS compliant. That's an interesting fault and here in the UK I would ring the guys at Marshall and let them take a look, but in the USA it may be worthwhile emailing them for an explanation. When I found a fault with some Mode4 boards they took it seriously and thanked me for alerting them.

[SoulFetish]

Not sure what the residue is by the OC Pass sticker. Will alcohol and mop up with a tissue remove it ?

that is benign. Probably just a micro layer surface film from using canned air to blow off excess 99% isopropyl I used to clean pcb where I worked.

Is this amp ROHS compliant with lead free solder ?

Without looking, - probably.
This amp is still under Marshall’s manufacturer warranty.

It’s weird, LT. nor only was there a degradation of the solder mask layer, but the resistor insulation was pierced, exceeding its breakdown voltage.

Just spitballing here, but I wonder if the adhesive compound here uses a solvent cure. I know acetone can be a highly effective curing agent, but an acetonal cure silicone might cause sever damage to composites and insulating materials.

[SoulFetish]

Marshall amps are all ROHS compliant. That's an interesting fault and here in the UK I would ring the guys at Marshall and let them take a look, but in the USA it may be worthwhile emailing them for an explanation. When I found a fault with some Mode4 boards they took it seriously and thanked me for alerting them.

Thanks Mick. You have to figure with two of the exact same faults within a year in that model is cause for concern. In the odd chance that were the only/first ones to come across it we probably won’t be the last. And could end up being a huge headache for Marshall.
We are submitting these as a warranty service, so I’ll be sure to call Marshall USA’s service department and give them a heads up so they can take a look. Maybe a they get out ahead of this.

[tdlunsfo]

I just had this exact failure on a similar spot on a DSL100. In this case the customer heard the arc and shut it down. When I powered it up I could visually see the arc from the trace to the bleeder. Fixed it by scrapping away the glue and installing a new bleeder off the board. I posted in another forum that I hadn't seen the glue age like this. It was brown/green in spots almost like an omeba or something. Really odd.

[loudthud]

Some people have told me not to use RTV type glue on electronics because it contains acetone. I've never seen the results of what it can do. This glue looks much more viscous than RTV.

RTV is sort of a rubbery glue you might use to prevent leakage around an automotive wind screen (windshield).

Edit: I think RTV stands for Room Temperature Vulcanizing.

[Mick Bailey]

I've not seen any reference to acetone in the composition - rather, I think there's a confusion with acetoxy which releases acetic acid in the presence of atmospheric moisture. The acid-cure causes corrosion and in some cases - mainly in humid conditions - can be quite damaging. I built an electric fence energizer on PCB and used acetoxy RTV inside the enclosure. The whole of the inside rotted away and after three years it still smelt of acetic acid. I now use Loctite neutral cure silicone. It's a little soft, but has a high dielectric strength and high temperature rating. A lot of amp manufacturers use hot melt glue due to the rapid set-up time.

[Tom Phillips]

Standard RTV inside a sealed electronics enclosure is a bad combination as you have stated. The "electronics grade" RTV, as some people call it is very expensive. It also cures slower than standard RTV but it gets the job done for certain applications.

[loudthud]

I think you are right about the acetic acid. It's been too many years for me to remember clearly.

[SoulFetish]

I've not seen any reference to acetone in the composition - rather, I think there's a confusion with acetoxy which releases acetic acid in the presence of atmospheric moisture. The acid-cure causes corrosion and in some cases - mainly in humid conditions - can be quite damaging. I built an electric fence energizer on PCB and used acetoxy RTV inside the enclosure. The whole of the inside rotted away and after three years it still smelt of acetic acid. I now use Loctite neutral cure silicone. It's a little soft, but has a high dielectric strength and high temperature rating. A lot of amp manufacturers use hot melt glue due to the rapid set-up time.

I went back to a few Manufacturers websites to brush up on RTV curing types. I may have confused acetoxy and acetone cures.
Interestingly, many neutral cure RTV silicones are Acetone and Alkoxy cure types but do not create corrosive byproducts like Acetoxy cures, and are safe for use in Electronics (such as securing electronic components to PCBs).

[DaiTheFlu1967]

Im replying after seeing your post because I had this EXACT same fault about a year ago (DSL100HR). I turned on the amp and heard the arc and immediately shut it down cause i know that sound all too well. Luckily i had taken note of the the glue darkening prior to this so it was the first thing i checked. I cleaned off all of the glue (really took the time to get it all and restore the board its previous aesthetic condition as well with some micro mesh) and measured the resistors (i have a schematic but i forget off the top of my head what value they were) and after that i cautiously used my Variac to slowly power the amp which is a pain in the ass with the DSL100HR's because the switching board doesnt work properly if it doesnt get enough voltage so cycling the amp is necessary afterwards i've found. After that it worked good as new and has ever since.

But im worried still about this happening elsewhere since every spot that has that glue has turned either dark brown or black already. Im no tech, im a electronics hobbyist and i do my best to take care of my equipment whether i built it myself or bought it from a large company. I will say im relieved to see a professional come to the same conclusion as i have. What would your recommendation be for me as far as what to do with the rest of the glue that has turned dark as well? Maybe just remove it from the places where its in close proximity to high voltage or just remove it all to avoid any future headaches?

Again im really glad i saw your post, youll see this is my first post here. I frequent Marshall Amp Forum and DIYstompboxes etc but after seeing your post i felt like it would be useful to be a member here as well. A reply with your thoughts would be greatly appreciated.

*********
-The 1st picture was taken right after the amp arched and i turned it off an pulled the chassis. I took the picture cause it was my best explanation for the issue and i wanted to document it before i cleaned it off.

-The 2nd picture was taken partially through cleaning off the burnt remains of the glue. It was late so i didn't finish until the next day and i apparently forgot to take a picture of it when i was done. But it is now completely cleaned off and the board is scratch free.I just finished carefully removing pieces of glue with tweezers then for the scratches i used some micro mesh. Youd never know anything happened there if you were to look at it now. I would be more than happy to take a picture of it now if you or anyone else is interested in documenting the repair. Im seriously kicking myself for forgetting to take a picture of it once i finished lol.




Halfway through cleaning glue off***

[DaiTheFlu1967]

I just had this exact failure on a similar spot on a DSL100. In this case the customer heard the arc and shut it down. When I powered it up I could visually see the arc from the trace to the bleeder. Fixed it by scrapping away the glue and installing a new bleeder off the board. I posted in another forum that I hadn't seen the glue age like this. It was brown/green in spots almost like an omeba or something. Really odd.

If you dont mind reading my post above this one replying to the OP and giving me your professional thoughts about removing the rest of the dark glue from around the board I would appriciate it. My amp is one of the first off the line (bought second month after release serial date of 11-17). And i had this failure happen over a year ago and its been working flawlessly ever since i fixed it but im questioning if i should remove the glue from the other high voltage area just to be safe. I put a choke in a few months ago and when i was removing the choke resistor (i know thats not the correct terminology but im drawing a blank) i noticed another resistor pair close to the other set of filter caps that is starting to look the same as the pair that had the arch occur. If youd like pictures or for a more specific explanation let me know and ill look at the schematic and we can figure out what kind of voltage that other pair is close to. Thanks for your help!

[DrGonz78]

I would remove any glue that was starting to be suspect. One thing I noticed from your pictures is that R14 appears to be cracking. Not a big deal but thought I would mention it.