Marshall w/ melted caps next to tube sockets

[VintageGrooves]

Hi. My name is drew, and I fix (mostly solid state) amps at a shop called Vintage Grooves. Anyway, the other day a Marshall TSL602 comes in with the ht fuse blown. The tubes tested fine, but one of the el34s was visibly arcing at high volume, so I replaced both of them. The amp still popped the ht fuse, so I popped the hood and saw that two of the contacts on the back of one of the el34 sockets had arced. I cleaned the arc as best I could, hooked it all up, and voila - all is well. Well, for a few minutes, anyway. However, after a few minutes at high volume the amp began to have that "something is burning on the pcb" smell. I popped the hood again, and noticed that the little ceramic disc caps right next to the el34 sockets (bottom left and right of pic 2) had black oooze under them, like they were getting way hot. (see pic three - the cap is under the grid resistor) So, I'm wondering - do I just replace the caps (as though they were originally damaged by the bad tube) and hope for the best, or do I consider that possibly there is another problem somewhere? As I mentioned, we mostly do solid state stuff here, so I'm really stumped on this one - any help at all would be greatly appreciated!
Sincerely,
drew

[Enzo]

C26 and C27 are little 22pf 500v disc ceramic caps. When they burn up, replace them with 1000v caps. Or leave them out if the amp seems stable without them. They are wired between plate and screen on each tube - pins 3 to 4. They fail short, and then they short the plate and screen together, making it a triode essentially. Hum noises and low gain result.

Usually we get tube guys who are afraid of solid state. it is refreshing to have a SS guy wondering about tubes. In either case, troubleshooting is troubleshooting, and you seem to be doing OK.

[VintageGrooves]

Thanx, Enzo - you rock! I finally found a schematic on the drtube site, and see what you're saying. Two more quick Qs, if you don't mind: 1) by "stable", do you mean that the amp doesn't oscillate? Also, if it's stable w/ the tubes I just put in , will it necessarily be stable w/ others? 2) I've never seen these caps on older tube amp schematics I've looked at - do you know why they're there?
Again, many thanx.
drew

[Enzo]

They are there for... stability.

Yes, not oscillating, not making parasitics on peaks. WHich is also oscillating, really.

WOuld other tubes be less stable? Perhaps. No guarantees in this life.

You can operate the amp withiout them, and in fact I would, just to verify that it works OK otherwise. I do replace them, I have a bag of suitable parts for it. If I didn't have tham, I would be seeing if they were necessary or not by testing.

[Jasonp]

Do you think this cap effects tone or dampens the feel or dynamics of the amp to gain stability?

[Enzo]

Not that I can detect. I certainly would not remove it if it were not bad. Not replacing it is more a matter of expedience than tonal nirvana.

[Jasonp]

Not that I can detect. I certainly would not remove it if it were not bad. Not replacing it is more a matter of expedience than tonal nirvana.

Can you please explain to me what that cap is techinally doing for the circuit in that spot? .. voltage wise? Is there signal runing through it?

[Enzo]

it couples the plate and screen together at really high frequencies. This reduces the available gain at those freqs. This is way above anything coming out of the guitar.

On the screen side the boltage is more or less the B+. On the plate side, the very large plate signal is added to that. In my view, the 500v cap is not enough for the occasional large spike that comes flying back from the transformer

[Chris G]

it couples the plate and screen together at really high frequencies. This reduces the available gain at those freqs. This is way above anything coming out of the guitar.

Well, I'm not so sure ... I remember doing some corner-frequency math involving the caps in my ancient CBS-era Twin, which had them, and I seem to recall they rolled off everything above 3K or so.

I mean, guitar amp speakers don't put out too much above 5K, but still ... clipping them out made everything noticeably less dull.

[52 Bill]

Well, I'm not so sure ... I remember doing some corner-frequency math involving the caps in my ancient CBS-era Twin, which had them, and I seem to recall they rolled off everything above 3K or so.
Those caps in the Twin were used differently and were much higher in value than the ones that are being discussed here.

[Enzo]

Yes. Chris if you look at the Fender, they are MUCH larger caps and are wired from grid to ground. These are small 22pf caps wired screen to plate.

[Jasonp]

Very intersting! Thanks for the explain guys.
I would bring my amp to you guys in a heart beat and pay you guys top dollar for what ever my amp needed or fixed.

[Jasonp]

Yes. Chris if you look at the Fender, they are MUCH larger caps and are wired from grid to ground. These are small 22pf caps wired screen to plate.

what does pf stand for? I take it this value isn't very large..?

[Enzo]

pf mean picofarad.

A 0.001uf (microfarad) cap is also a 1000pf cap. WHich is also a 1nf cap (nanofarad)

SO the 22pf cap we speak of here is if you prefer, a 0.000022uf cap. Yes, very small.

In the old days when I was learning, we never used "pico" for anything, the term then would have been 22uuf or 22mmf, either of which meant "micro-micro farad." If you see mmf or uuf on an older schematic, that is what it means, it means picofarad.

[Jasonp]

pf mean picofarad.

A 0.001uf (microfarad) cap is also a 1000pf cap. WHich is also a 1nf cap (nanofarad)

SO the 22pf cap we speak of here is if you prefer, a 0.000022uf cap. Yes, very small.

In the old days when I was learning, we never used "pico" for anything, the term then would have been 22uuf or 22mmf, either of which meant "micro-micro farad." If you see mmf or uuf on an older schematic, that is what it means, it means picofarad.


wow, amazing small value for sure! Now does this mean that a very small (current?) is allowed by the cap?

[Enzo]

Does what mean that? It is a cap, so when it is working right, it blocks DC, and only allows high freq signal to pass through unimpeded. WHen it shorts, then all kinds of current can pass through it.

[Jasonp]

Does what mean that? It is a cap, so when it is working right, it blocks DC, and only allows high freq signal to pass through unimpeded. WHen it shorts, then all kinds of current can pass through it.

value is what I mean. Basically the smaller the value the less signal will go through the cap?
Thanks for answering my dumb questions, as you can read I'm a total newbb

[Enzo]

Basically the smaller the value the less signal will go through the cap?

Um, not like I think you mean. The impedance of a cap goes down the higher the frequency is that tries to flow through it. Changing cap values is one way to adjust the frequency response of an amp. It is also how you set up the crossover points in a speaker crossover assembly.

Smaller value caos will only let higher frequency signal through. Just the same way the crossover in a speaker only lets treble high frequency stuff through to the tweeter. The simplest crossover is just a plain old cap in series with the tweeter.

There is something we call a brightness cap in some amps. The BRIGHT switch on most any Fender amp is an example. Turn it on and the sound gets a little brighter - not louder, just brighter. Brighter means a little more high frequency than without it. SOme amps don't use a switch for it, some like old Marshalls wired it permanently on one channel and left it off of the other. Instant bright channel and normal channel from two channels that were otherwise identical.

That brightness cap was something like 100pf. The 22pf cap is much smaller in value, so the freqs it lets through are even higher than the extra brightness in our Fender example.

SO it is not like a smaller cap would be turning down the volume, a smaller cap would be like turning up the treble, or in this case maybe turning down eveything else.

But remember, this cap is not in the signal path, this is bypassing some of the amp gain. So it reduces the gain at frequencies much higher than the guitar makes. Think of this cap as turning down the ultra-treble. Sorta...

[Jasonp]

Um, not like I think you mean. The impedance of a cap goes down the higher the frequency is that tries to flow through it. Changing cap values is one way to adjust the frequency response of an amp. It is also how you set up the crossover points in a speaker crossover assembly.

Smaller value caos will only let higher frequency signal through. Just the same way the crossover in a speaker only lets treble high frequency stuff through to the tweeter. The simplest crossover is just a plain old cap in series with the tweeter.

There is something we call a brightness cap in some amps. The BRIGHT switch on most any Fender amp is an example. Turn it on and the sound gets a little brighter - not louder, just brighter. Brighter means a little more high frequency than without it. SOme amps don't use a switch for it, some like old Marshalls wired it permanently on one channel and left it off of the other. Instant bright channel and normal channel from two channels that were otherwise identical.

That brightness cap was something like 100pf. The 22pf cap is much smaller in value, so the freqs it lets through are even higher than the extra brightness in our Fender example.

SO it is not like a smaller cap would be turning down the volume, a smaller cap would be like turning up the treble, or in this case maybe turning down eveything else.

But remember, this cap is not in the signal path, this is bypassing some of the amp gain. So it reduces the gain at frequencies much higher than the guitar makes. Think of this cap as turning down the ultra-treble. Sorta...

Ok, this is starting to make sense.
By gain do you mean volume?

[VintageGrooves]

For what it's worth, there are some folks who won't even wire in a 1-ohm resistor between cathode and ground to simplify checking bias, bc they feel that it complicates the signal path and adulterates the sound or something:). I dunno, must take an awfully good hear to hear differences like that...

BTW - the marshall sounded great w/o the caps, so we went with it that way. Many thanx again,
drew

[VintageGrooves]

omg - i'm so sorry - I was looking at page two, and responded to a question that was like 10 comments ago - sorry for the non-sequitor!!!

For what it's worth, in response to the current discussion...
If it doesn't make sense that a smaller cap value could also take away high frequenies, just think of the little cap that's wired to the tone pot in your guitar. It works the same way: when you open the treble pot (which turns down the treble), what you're doing is opening up a "valve" that lets some of the signal bleed off into ground. However, along the way is the cap, and it only lets high frequencies go through it. So, when you turn down the treble, you're letting some of the treble bleed off to ground through the cap, and it is this same cap that also stops any bass from bleeding off the same way. So, depending on where it's at in the signal chain, a cap could either let more treble into the signal path or bleed more treble to ground.
hope that helps,
drew

[Enzo]

There you go, great example.

[Jasonp]

omg - i'm so sorry - I was looking at page two, and responded to a question that was like 10 comments ago - sorry for the non-sequitor!!!

For what it's worth, in response to the current discussion...
If it doesn't make sense that a smaller cap value could also take away high frequenies, just think of the little cap that's wired to the tone pot in your guitar. It works the same way: when you open the treble pot (which turns down the treble), what you're doing is opening up a "valve" that lets some of the signal bleed off into ground. However, along the way is the cap, and it only lets high frequencies go through it. So, when you turn down the treble, you're letting some of the treble bleed off to ground through the cap, and it is this same cap that also stops any bass from bleeding off the same way. So, depending on where it's at in the signal chain, a cap could either let more treble into the signal path or bleed more treble to ground.
hope that helps,
drew

No the more we talk about this stuff the more I can try to understand what you guys are talking about. Pretty cool.

[Jasonp]

interesting thing happened today. I was playing at about 4 1/2 volume wise on the amp.. Super loud on the distortion channel. I looked at the back of the amp and I see the SED power tube on the last socket (power tranny side) start to red plate. So I stop playing and turn the amp off quick, leave it on stand by. So I power it up again all ok, check bias both sides at about 88 mV. Correct me if I wrong here but this tube is on the same circuit as that cap?
Started playing again and was fine. I bet I have a tube going out.. so I changed them out for JJ's and lower the bias to 78 mV a side. Sounds good.

[Enzo]

That is the socket with the cap. Take that tube and swap places with it and one of the others. Now pplay again. If the same tube red plates, then it is obviously the tube. If the other tube now red plates in the same socket, then the amp has an issue there.

I suspect you have a failing tube. The fact it will bias out doesn't mean there is not some little piece inside breaking loose, or just the tube not being able to handle operating under load, regardless of idle bias.

[VintageGrooves]

ok, back again:) I removed the caps, and the amp sounded great. I advised the customer of the option of replacing the caps or just leaving them out, and he opted to leave them out (the amp seemed totally stable). But, 2 days later, here comes the amp again, this time with a blown filament fuse (the fuse that is in series w/ fil1 and fil2 on the two power tubes and the phase inverter.). The fuse is a 6.3A slow blow, and the one that was in it was the kind with the little bubble in the middle of the filiment. Anyway, I hooked my current meter up to the two sides of the fuse, cranked up my PS real slow, and watched what happened. After confirming that nothing was going to blow up, I powered up quickly and watched the fuse current. When the tubes were totally cold, the current would run up to 8-10A for just a second, and then would settle down around 4.5A.
Is this current normal? If so, should I just assume it was an old fuse? (I realize that this is rare:)). Also - could this have anything to do with my cutting out C26 and C27?
many thanx,
drew
p.s. I can't tell you if the amp will blow another T6.3, bc my radio shaft doesn't stock anything even close, so I have to order - arghh!