'82 JCM 800 4104 Service and Mods

[physics]

Yet another project start(ed/ing).... Picked up a 1982 Marshall 4104 loaded with two EV Force 12's (heatsinked, and with the computer style fonts on the back cover and dustcap) in my new place since all my other gear is several hundred miles away still. Seller said the shop that biased it claims it needs a cap job soon, I just got a minimum set of tools in the mail from my old place so I'll be checking the caps for value, filtration, seal burst, etc. I have a replacement set on hand if needed, and now that I have my tools I should be able to do the job myself. Most recently, it stopped making noise. Knobs have no effect, including the master. Powers on fine, fuses are correct values and are not blown. Hopefully this weekend I'll be able to check voltages and trace the signal.

[K O'Connor]

Hey Physics

I wouldn't jump right to replacing caps yet; that is an easy cash grab techs do particularly when they do not know what is wrong with the amp.

The amp is a simple MV 800-style with a rather spacious chassis. If online photos are to be believed the MV is a standard type, although by that time some models used a PPI-MV with a dual pot - terrible for clean sounds.

As far as "no sound" goes, there are lots of reasons for that. The most expensive ones to rule out (easily) are an open OT or open speakers. You can check both with an ohmmeter and the amp unplugged and caps discharged. More typical reasons are cold solder connections, oxidised jack contacts, dirty pots and similar things.

If you intend to keep the amp, adding proper bias controls and meter jacks would be the first mod to do. It allows you to use alternate tube types and to mix tubes, plus easily set the bias to a safe level for any tube and hum-balance the output stage. BMK2 as our kits go, or get the parts from the usual places.

It would be handy to have a power limiting safety socket for the electrical testing

Have fun

[physics]

Hey Physics

I wouldn't jump right to replacing caps yet; that is an easy cash grab techs do particularly when they do not know what is wrong with the amp.

The amp is a simple MV 800-style with a rather spacious chassis. If online photos are to be believed the MV is a standard type, although by that time some models used a PPI-MV with a dual pot - terrible for clean sounds.

As far as "no sound" goes, there are lots of reasons for that. The most expensive ones to rule out (easily) are an open OT or open speakers. You can check both with an ohmmeter and the amp unplugged and caps discharged. More typical reasons are cold solder connections, oxidised jack contacts, dirty pots and similar things.

If you intend to keep the amp, adding proper bias controls and meter jacks would be the first mod to do. It allows you to use alternate tube types and to mix tubes, plus easily set the bias to a safe level for any tube and hum-balance the output stage. BMK2 as our kits go, or get the parts from the usual places.

It would be handy to have a power limiting safety socket for the electrical testing

Have fun

That's a good point, and yes I've been suspicious of the diagnosis from the previous tech. There were some symptoms before it stopped working that seemed possibly more in line with a bad connection somewhere, but I don't know the exhaustive list of issues caused by bad filter caps. The issues I was seeing were that sometimes the output would get quieter, more distorted in a crackly/noisy sense, and a little intermittent before fixing itself up. At one point the amp made those sorts of sounds with the guitar unplugged while powering off. The plan is to check the things you mentioned, and then trace a signal through the amp with the MV turned down and find out where it stops. It happens on both input jacks, so it seems it's not a jack, but could be wrong. I recently found out that my guitar's wiring has issues as well, so until I fix that there is still some unknowns regarding what was the guitar and what was the amp.

As for the master volume, it's a single gang pot. The 100k resistor mod is in the queue for after I get this working again, and I'll be posting questions about that down the road (in short, the transfer function I've (mis?)calculated doesn't match the description in TUT & TUT4, or the master volume thread on this board).

I'd like to do the bias mods and stuff as well, and I have the kit back at my old place so it shouldn't be hard to get it mailed out. Power Scaling would be cool too, and useful in my apartment/non-stadium life. Despite my dislike for most Marshall sounds I've heard, this amp is quickly becoming my favorite amp I own and maybe played too. Perhaps I just don't like Celestions, and the EV's in this amp are the reason I like it. I certainly liked a friend's rig a lot more after he switched to EV's.

The PLSS is a good idea, mine is still back home. I'll have to dig up some parts at the local Ace Hardware for that.

[physics]

Small update: after pulling the chassis and turning it on again, it's making sound. Not sure if something fixed itself in the process of being moved around or not. I was able to reproduce the distorted output though, only on the high sensitivity input. Interesting... Hopefully I'll have more time later this weekend to poke around. Also, seems like mine is from early '82, it's plate voltage sits about 400V even after warming up. My understanding is that later on they went to around 450V.

[physics]

Further update: I can reproduce the symptoms mentioned, but only soon after power-on. Moving the master volume around might affect it, but I couldn't tell. So as a guess, it's heat-related and goes away after something warms up, and is somewhere in the first gain stage. Not sure if the master volume is involved yet or not. Thinking a can of freeze spray would be the next step, to see if I can force the issue some time after power-on.

[physics]

Hmm.... Sufficient monkeying with the input cable seems to cause the same issues.... It seems like messing with the guitar-end gets the most results. I got some results while monkeying with the end plugged into the amp (pulling in and out repeatedly, moving around, etc.), but I'm not sure if I was shaking the guitar's connection, which was on the table at the same time whilst I was forcefully inserting and removing the cable. I have to move on to getting other stuff done today, but it seems like the search zone for next time is from the guitar up to the high-sensitivity input jack.

[physics]

Alright, I've done some measuring. The input jacks are arranged such that when plugged into the "high" input the signal flows from the first stage, through the "low" jack tip switch contacts and to the pre-volume that sits before the second stage. Nothing new for those of you familiar with the circuit, but I mention it because that switch measures 40 Ω across the contacts. A far cry from the < 1 Ω specified. Today I replaced the two input jacks with new UK-made Cliff jacks, and now I measure a fraction of an ohm across that switch in the circuit. The old jack I pulled measures all over the place out of the circuit, anywhere from 15 Ω to 1.5 MΩ. So at the very least, that connection was an issue.

After replacing the jacks, I have not been able to reproduce the issue I was having. No matter how softly I play, the signal doesn't cut out or get overly distorted. Also, I could be wrong, but the amp seems "cleaner" and less "fizzy" after the jacks were replaced. There were low single notes that had a fizzy haze on them previously, and that doesn't seem to be the case anymore. We'll see how things hold up after some more playing and if the recordings I get sound any different than they did before. Thinking the next thing to do, if this fix holds, is to improve the ground scheme. BZZZZZZZZZ!

[K O'Connor]

Hi Guys

Oxidised jacks were mentioned in post-2. The problematic jack is usually the one that is never cycled inasmuch as a cord is always in it or never in it. The contacts oxidise and at the point where the status of the cord is changed, the oxidation acts like a rectifier, imposing a threshold over which the signal has to overcome or "break through". This happens a lot with effects loops, as TUTs report.

In an amp like the 800 with the High and Low jacks, the contacts on the Low jack are never cycled and build up oxidation. The High jack closed contact for the tip merely shorts the input with no cord plugged on, so that tends to be less problematic.

An issue with every plastic jack, especially with the smaller square ones made by Switchcraft and others, is that the spring tension of the contacts is fairly low. A few cycles or the simple passage of time shows off the weakness of the design. 3mm minijacks are exponentially worse and are ubiquitous with computer and mobile equipment proliferation.

The open-frame jacks of yore have larger components and generally better spring tension. Their weak point is that all the components are held together by a crimp of the sleeve end. Rotation of the wafers and metal parts can cause a short between things that should not be shorted, and it also makes it more difficult to tighten the jack in the chassis hole.

Have fun

[K O'Connor]

Hi Guys

Intermittent problems are the hardest to figure out. Physics did a great job finding a way to induce the symptom, allowing the root cause to be sussed out easily.

[physics]

Indeed, bad jacks were mentioned in post 2! I forgot about that once I hopped down the rabbit hole. "Rectification" is a pretty darn good way of describing what I was hearing. Once I noticed the threshold behavior I was reminded of something I read somewhere about oxidation creating a diode, either in the context of DIY AM radios, or in TUT, or probably both, and that helped tip me off that it might be the jack(s). Maybe I should save the bad jack and make a radio with it.

Thanks for the input and help!