Marshall 1959 SL100 (4 x EL34) Grid leak resistor size

[tubeswell]

Can't quite make out the writing on the attached schematic. Are the EL34 grid leaks mean't to say 120k each or 220k each? (Only, it would seem inappropriate if they were 220k in this amp)

[Tom Phillips]

The marking looks like 120k to me.

[Enzo]

I am going to vote 220k. Aside from 120k looking low, the almost identical 1959T is clearly printed as 220k. Also the newer versions of 1959. like the 1959SLP or the "handwired" reissue, also all show 220k.

[tubeswell]

I just serviced a 90s reissue SL100 head and it has 220k grid leaks - go figure

Even with new EL34s and the bias voltage constant at -41V, the tube current seems to gradually creep up from 30s into the 50s over a 25 minute period. It seems to run okay on about '3' happily for hours, but when its cranked, then after a while it gets crackly. This would seem to be a problem with grid current and bias drift maybe?

[Chuck H]

IME most Marshalls with EL34's use 220k grid leaks. It's a popular mod with current EL34 offerings to change them to 150k. I've even read about noted EL34 guru Ken Fischer (of Trainwreck fame) advocating for this. Though he didn't do it in his own amps. Probably because he sold them with selected, usually NOS tubes. The lower grid leaks limit gain a small amount but also reduce the coupling cap discharge time and, so, grid loading.?. How this ultimately affects the tubes working conditions WRT reliability is not entirely clear to me, but I think it has something to do with certain types of distortions when clipping causing voltage spikes that find a lower impedance path through the screens than the plates potentially causing failures in the newer, more tender EL34's. Also recommended pretty much everywhere is the use of 1k individual screen grid resistors. This certainly increases screen grid circuit impedance and limits screen current for higher frequency anomalies. I used 150k grid leaks and 1k screen R's in my EL34 build just because it's recommended. I never tone tested any other component values.

[frus]

IME most Marshalls with EL34's use 220k grid leaks.

That's for 2xEL34
A shared grid leak for two parallel tubes (as in 4xEL34 amps), is supposed to be half the resistance (grid current is double, so bias shift is the same).
So 100k -120k would be the right value

[Chuck H]

That's for 2xEL34
A shared grid leak for two parallel tubes (as in 4xEL34 amps), is supposed to be half the resistance (grid current is double, so bias shift is the same).
So 100k -120k would be the right value

Well, I didn't say 220k was "the right value" WRT design. I said that most Marshalls with EL34's use 220k grid leaks. And that statement, being true, doesn't require correction. In fact, notice your statement above that I have highlighted bold. Since I am specifically addressing Marshall circuits, and you are addressing my post with a quote, you are, in fact, wrong. 100k to 120k is absolutely not the right value for any Marshall amplifier that uses EL34's whether 100W or 50W. And yes, double the load/half the resistance, in general. But 100W Marshalls don't sound like 50W Marshalls and some people don't want them to either. That 100W Marshalls use the "wrong" value is partly responsible for their tone. My suggestion of the 150k grid leak value is considered a tonal compromise by some and they would rather live closer to the edge with the 220k value. But thanks noting the circuit differences WRT the shared grid leak on the 100W vs. the inherently individual grid leak for the 50W since it may become significant to the discussion.

[mozz]

That schematic shows 220k, you can see the 2 nubs of the number, the back is just cut off. A 1 would be a straight line.

[frus]

I meant, if they specified 220k for 50W amp, then 100-120k would be a sensible value for 100W.
Of course, WRT bias shift due to grid current, the tone is another matter completely

Btw, I've heard assumptions (somewhere on this or some other forum), that "classic" amp designers used to put larger-than-max-specified grid leaks, but used to bias them a bit colder than "theoretically" possible, to avoid thermal runaway. And that maybe even that the famous "70% bias rule" stemmed from that. Any thoughts?

[diydidi]

For EL34 use 220K. Also change grid stoppers to 5k6.
Change bias circuit accordingly.
What is your feedback res value?

[Enzo]

What grid current are we talking about? Unless we drive the grids positive, there is little current flow. No steady DC current anyway.

[tubeswell]

Grid leak - a few nano amps at most. What is at the heart of the matter (for me) is whether it might be more practical on a 4 x EL34 model to drop the grid leak resistance - given the generally disappointing quality of new EL34s - even though the impedance bridging would be degraded somewhat. (Either that or fork out for half-way decent EL34s).

I think 120k is probably too low (because of the impedance bridge), but seems like it might be a practical workaround to use 150k or 180k (without having to redesign/rebuild the entire amp to use a different type of PI/driver arrangement). I was curious about what was written on the (1970 dated) 'unicorn' schematic, because I was pondering whether Marshall engineers had tried this back in the day, but eventually decided that 220k was better (because of the impedance bridging). I daresay EL34s in 1970 were rugged enough, or cheap enough, to mean that changing the shared 220k grid leak resistor value wasn't worth thinking about.

[g1]

Not sure if this is relevant, but in all the amps that offered different power tubes (JCM800 with EL34 for UK and 6550 for US, JCM900 etc. with EL34 or 5881) they always stuck with 220K for the EL34's. The other tubes usually got 150K's. I think it would have been easier for them to just go with 150K across the board for all production in these cases. So there must have been some preference for the 220K with EL34's.

[tubeswell]

Not sure if this is relevant, but in all the amps that offered different power tubes (JCM800 with EL34 for UK and 6550 for US, JCM900 etc. with EL34 or 5881) they always stuck with 220K for the EL34's. The other tubes usually got 150K's. I think it would have been easier for them to just go with 150K across the board for all production in these cases. So there must have been some preference for the 220K with EL34's.

Yes I've noticed this. Probably because a quad of EL34s do sound good this way. And in the latter 1990s, I understand that Marshall negotiated with (the real) Svetlana, to make more reliable EL34s, so the RI's of that time (which this amp is) could be marketed hand-on-heart as not likely to expire before the warranty was up.

[Chuck H]

I have an early 1959 reissue. I think it's "SLP" but not sure. Supposedly these early ones had the better sounding transformers. I was nonplused by the stock amps overdrive tone to be honest. I modded it with my own version of a channel stack and a master volume, which is how I used it for most of my gigging years starting some twenty five years ago. Around that time I installed the old school brown base TESLA tubes. That helped. It sounded great after the mods and the tube change (I wish I'd bought another five sets of those!). It was the first amp I ever modified. I haven't even played the amp in ten years and I haven't used it regularly for eighteen! My plan is to rebuild it stock with an eyelet board, maybe mustard caps, etc. and sell it for whatever the market will offer... It has 220k grid leaks.

Just a slightly related story I suppose.?.