Marshall JCM800 valve problem

Not sure about the KT88 but I have some Marshall KT66s and they're re-branded Shuguang ones. Does your KT88 look like this http://www.ebay.co.uk/itm/SHUGUANG-KT88 ... 321wt_1161 or like this http://www.ebay.co.uk/itm/SHUGUANG-MATC ... 104wt_1399 ? If so then I guess it's a Shuguang.

VB

Its a Kerry King 2203 if that means anything, just looking for a schematic as there are some resistors gone as well a fuse.

Yup, those look like the variant with the metal base collar peened round the baseplate. If you really want that type you could give HiFi Collective http://www.hificollective.co.uk/catalog ... p-852.html a call and see if they have any. Otherwise the TAD ones http://www.tubeampdoctor.com/en/shop_Ot ... TR_TAD_575 look very similar (except it may only have a single side getter ?). If people try to offer you a 6550 though do be careful. They're not rated to handle the same screen voltage that the KT88 is. Of course in practice they might. But equally they might not.

VB

Now that's interesting....just as an observation. Last year we tested one of these very amplifiers and at some point, one of the KT88's did just what your's did. The heater was dull but everything else appeared OK. Out of interest, I took the valve home and ran it up on the Taylor....virtually no emission, dull heater but plenty of heat. I assumed it had gone gassy but the getter looked OK. I was more interested in the failure mode.

Marshall did reluctantly send us another KT88 which enabled us to finish testing but clearly they were not pleased at the extra expense.......!

Rich

Further examination has revealed a resistor that's open circuit plus a fuse, not having much luck finding a schematic yet. Also found a resistor that's never been soldered at one end and is hanging in mid air...(ah quality)
Thanks for the possible sources for replacement

Its a Kerry King 2203 if that means anything, just looking for a schematic as there are some resistors gone as well a fuse.

There seem to be zillions of JCM800 schematics around but they all use EL34s for output!

Yours isn't so easy to find as a couple of locations no longer exist but these should give you what you want:

https://docs.google.com/file/d/0B9WgrAT ... li=1&hl=en

https://docs.google.com/file/d/0B9WgrAT ... edit?hl=en

Thanks Terry, that's the one , your a star

I see the R120 is a bleeder resistor across one of the smoothers, interestingly C109 has a bulge on the top, wonder if R120 played a part in that.

Why are the grid stoppers for the KT88s specified as 1%? Surely, that is real overkill! The feed a very high impedance so 1k or so either way would make no difference at all.

Well it is a guitar amp, so I suppose the KT88s might be driven hard enough to draw grid current. But the inherent distortion would surely completely swamp the effects of any drive imbalance resulting from mismatched grid stoppers ?

VB

I wonder if the choice is more to do with the resistors being metal film and therefore non inductive?

Al

Metal film isn't necessarily non inductive- the value may be set by spiral trimming of the film.

Inductive on a grid stopper on an audio amp only helps anyway.

Only at VHF and up is the spiral trimming significant on film resistors anyway. 200% values are fine for grid stoppers. They are 1% MF just to have one type of inventory.

I can understand the inventory reason and the common tolerance of professional metal film Rs being 1% but why mark it on the schematic as such? Not all resistors are marked 1% which suggests that the designer thought it necessary as it was to do with the phase splitter drives to the KT88s.

Yes they did fit quality resistors in the phase splitters quite widely.
The push pull sets I have are all 20% resistors except the phase splitter. The resistors are 2% high stability as they have the pink band on them.

Phase splitter gain setting / load resistors need tight tolerance unless there is a preset, which might be a better idea.

But Grid stoppers or Grid "leak" resistors are irrelevant to gain and not part of phase splitter/drivers.

Phase splitter gain setting / load resistors need tight tolerance unless there is a preset, which might be a better idea.

But Grid stoppers or Grid "leak" resistors are irrelevant to gain and not part of phase splitter/drivers.

Grid stoppers, yes. But grid leaks are often, as far as AC's concerned, in parallel with the anode loads. Their value is usually higher than the anode load so the effect of it being different from nominal is not so large. It may well not be negligible though.

VB

I look at it as a compromise.
If an anode load is 10K and on the other side of the coupling capacitor there is a 100K grid leak resistor a 10% change in the grid leak will only make a 1% difference to the anode load.
It does make sense to me.

I do too, so for example in the Quad II the EF86 anode loads are 180k and the KT66 grid leaks are 680k. So a 15% change in the grid leak (not uncommon in those high-value Eries) causes a 3% change in the effective AC anode load. It's not large, but it's not negligible. Fortunately it's a push-pull amp and all the resistors tend to drift in the same direction (although not usually by the same amount). So the balance tends not to be as badly upset as the individual circuits are.

VB