Marshall Valve Problem

[Dunkmop]

Hey all,

I am an aspiring Guitar player/ guitar tech. I have just really started to get into the more technical side of the amps. I am a electrical/electronic engineer, so i have a nice background to work on this stuff.

I have recently decided to retube and re-cap my amp - Marshall 2555SL

It sounded so much more awesome when i changed the original el34s and 12ax7s (Marshall JJs i think?); i have replaced with EH12ax7s and thought i'd try out these GT EL43Ms. Whether or not i know if these are valves for me; im unsure (plz offer me any advice on this subject if you would like).

The valves have been a bit of a problem (GT EL34Ms). I looked into these semi-hyped tube, and from the sounds of it, they seemed pretty good sounding/reliable/bang for buck.

Ive recently found out they arent as reliable as i thought. i had to sent away for another quad when the filament looked like it had dropped down and shorted the cathode - big boom. i have ordered a new set that came next day chucked em in and set them up.

I dont know how to treat new valves, so advice i want really - is it true they should survive a 40 hour 'burn in'?

Any help would be much appreciated. Cheers

Forgot to say that the new valves have just gone too

[MWJB]

I thought the GT tubes were already burned in? Typically they are burned in so that they can be matched & graded, lots of vendors do this.

What plate current are you getting with the new tubes? If they bias up to 15-17W each that should be fine, monitor them for a while if you like, see if they drift...but not 40 hours, it's really not necessary. Nevertheless the tubes should survive 40hours at idle, no problem.

[Dunkmop]

Yeh they do burn them in. Are you supposed to burn them in with the stand by swtiched on? or is it just the filament - i was under the impression it has to be on full power. Technically the valves should last hours and hours left on, this is why im dubious.

I phoned them to see if they will replace these ones, but the are adament something is wrong with my amp. The filament shorted before, and this time the plates went red hot and then it failed.

I set these up coldish, because this is where many ppl said they shined. average 33-35mA @ 455, so 15W or so. Compared to the original tubes; they were set drawing 42 or so which i thought was quite high. when the GTs went in i had to turn the bias up a fair amount though, which im stumped about.

Power transformer is running warm/ hot to touch nearly

[Chuck H]

You may want to consider the possibility that it could be the amp. If you aren't blowing fuses and there is nothing visibly wrong I would start by removing all the tubes and checking voltages at the power tube socket plate, screen and input grid lugs. Also check the heater winding (it is AC). Post your findings. Does the power tranny still run hot with no tubes in?

Chuck

[MWJB]

"Are you supposed to burn them in with the stand by swtiched on?" Yes.

Plates shouldn't be going red hot at 33-35mA @ 455vdc, 15W isn't cold by any means.

Are all the tubes failing or is one specifically the culprit? Check the sockets for signs of arcing/shorts (if a plate pin has shorted to a filament you could have a track on the socket), replace any sockets that you can't clean up sufficiently. If you have a pair of useable tubes from the original 8 then double speaker load (to match impedances) and try them in the inner & then outer sockets, checking plate, screen, grid & cathode voltages.

Different brands of tubes bias up wildly differently at the same negative voltage on the grid.

"Power transformer is running warm/ hot to touch nearly" Heaven knows how hot it can run before failing...it could be warm because of an excessive load, it could be being heated by the tubes...what are your loaded AC heater voltages like (pin 2 to pin 7 of a EL34, use clips & connect up with amp unplugged from wall, check secure & no possible shorts)?

Sure your new filters are good/correct polarity (though i'm sure that one would have let go by now if wired wrong)? Pull out ALL the tubes and measure current draw accross the standby switch...shouldn't be any.

Do not fit 4 new tubes, all at once, until you have made these checks, if you still find nothing - start with one tube, then power down fit a second (2 either side of the OT - 2 inner or 2 outer), then power down & reset speaker load for 4 tubes & fit a 3rd tube, power down & fit the 4th...this way you may discover a faulty tube or socket.

Out of interst how are you measuring plate current?

[Dunkmop]

Only two valves failed and this was on one side of the amp. It was making a slight buzzing noise, which progressively got louder over a few seconds - i noticed it and looked at the amp, and two of the valves plates were going red hot; so i switched it off immediately. I came back later tried to switch the amp on again - the filaments heated up np, but as soon as i switched the standby switch on, it took out the main fuse in the plug (3A) instead.

There is no sign of arcing in the tube sockets and all the tubes fit nice and snug.

I tested the amp without tubes first - my findings;

Pin 1 (Grid 3) - chassis
Pin2 (Filament) - 3.3V AC
Pin 4 (Anode) - 478V DC
Pin 4 (Grid 2) - 478V DC
Pin 5 (Grid 1) - -42.3V DC
Pin 6 ----
Pin 7 (Filament) - 3.3V AC
Pin 8 (Cathode) - chassis

Now i test the tubes with a working set of tubes;

Pin 1 (Grid 3) - chassis
Pin 2 (Filament) - 3.1V AC
Pin 3 (Anode) - 457V DC
Pin 4 (Grid 2) - 452V DC
Pin 5 (Grid 1) - -42V DC
Pin 6 ----
Pin 7 (Filament) - 3.1V AC
Pin 8 (Cathode) - chassis

To measure the plate, I am using a TAD Biasmaster - adaptor sockets read the Pin 3

The transformer runs cool when no tubes are in place. I just wanted to know if anybody knows a typical temperature for the power transformer, it seems to be running hot generally, so i expect this is its normal operating temperature.

[MWJB]

"i noticed it and looked at the amp, and two of the valves plates were going red hot; so i switched it off immediately. I came back later tried to switch the amp on again" You don't say what action you took to prevent the same thing from happening again? If you didn't change anything, then the amp's going to do the exact same thing every time you turn it on.

Are these voltages consistant accross all 4 sockets? Any variance in voltages/current from one side of the PT to the other? What is the plate current at these voltages?

The transformer will run cool with no tubes, as they're not drawing any current (no load). Operating voltages are recorded under load. PT operating temperatures will vary greatly from one model of amp to the next. Secondary voltages are more of a giveaway if the PT is under stress (low B+ or low 6.3VAC supply).

6.2VAC on the heaters seems a bit low, this would usually be >6.3VAC. May be indicative of excessive current load one one of the PT secondaries?

It's a good idea to wind up the bias pot to give a higher than normal negative voltage at pin 5 (say -50 to -55vdc?), with no tubes installed, to ensure that the tubes are on the colder side when installed, then you can fine tune the plate current with less chance of overload.

-42v seems a bit low to me?

[Dunkmop]

I knew the tubes were gone when they started to red plate. I was working nights and had just got in lol, thats why i came back later. I just wanted to see what part of the valve had gone i.e. last time the filament went on one, but this time it was something else. Big blue flash and loud noise etc followed.

The voltages are consistant (within 1-2 volts of one another)

I've put the original Marshall jj tubes that were in there before and returned them to the previously set bias. Any suggestions of how to correctly approach setting up a new set of valves would be appreciated i.e. 'It's a good idea to wind up the bias pot to give a higher than normal negative voltage at pin 5' things like that. im trying to gather all information as possible.

So you add both filament voltages up then? its around 6.2 - 6.3 volts, the readings fluctuated slightly when testing them.

Is it a good idea to test under a maximum negative Grid 1 Voltage, and would you like me to do that?

[Chuck H]

More important to measure all the power tube sockets individually. Do the failures follow the same socket/s? Are you using any kind of attenuator?

Chuck

[MWJB]

"Is it a good idea to test under a maximum negative Grid 1 Voltage, and would you like me to do that?" Not necessarily, you just want the grids more negative than normal top save redplating unknown tubes on initial installation.

Trouble is you're only giving us tidbits of information...We need to know plate voltage, screen voltage and plate current - all at the same time...negative voltage at pin 5 on it's own means diddley squat.

The fact that your heaters read borderline is suspicious, assuming that you have a decent quality meter.

[Dunkmop]

It is an old fluke, so maybe a little un-calibrated. I will get my fluke from work and test again. Previously i tested every socket and most values came out the same, or within a few volts on the larger voltages.

What voltages would you typically expect from the heaters then? would this cause this kind of problem?

Im not using any sort of attenuator. Ive been running my original jj's in there for a couple of days now and seem fine for the moment.

[MWJB]

I would expect heaters to be more like 6.4VAC upwards. I don't suspect the heaters are causing this, more that a big load on the B+ rail might be pulling them down.

When the original tubes ar installed & running fine, what current are they drawing?