Very early Laney Sound 60W Head-Amplified Zener to drop B+ not droping enough expected voltage

[SoulFetish]

Customer brought us an early vintage Laney "Sound" 60W head for service. Came back a couple of times with the customer saying that it blows the HT fuse on him (Seems to have a tendency to overload it's EL34s, causing them to fail). I'm not the tech who worked on this originally for us, but I hear about it.
The high voltage supply is up in the high 600's, and without having been in the amp to know what the screens run at, I posited that the power supply might be stressing the newer production EL34s and maybe dropping the supply might be an appropriate measure. I told my boss that I had a salvage NTE2973 from a VVR circuit I built into one of my first amps back in the day and an a bunch of 59V zeners from a surplus lot, so we could do an install of Keen's amplified zener circuit as an option. I actually used that mosfet for that very purpose to drop the B+ on a Transformer spec'd for a quad of EL84s which I was using in an amp using only 2 x EL84s, and it worked exactly as expected.
Anyways, we got around to installing the zener/mosfet circuit today. Here is the exact schematic as modified:



So, when we powered the modified amp back up, there is only about -20V across the fet. Any idea what is going on??

[Helmholtz]

Did you have a load connected?
Circuit needs some load current and might not work as expected in standby.

What is actual screen voltage? It might be more important to bring the screen voltage down by more than 60V.

Did you exclude leaky coupling caps?

[SoulFetish]

Did you have a load connected?
Circuit needs some load current and might not work as expected in standby.

We did connect a load. Interestingly, however, the tubes were barely pulling any current (so the bias supply needs to be modified as well) and I was suspecting that this might have something to do with in as well.

What is actual screen voltage? It might be more important to bring the screen voltage by more than 60V.

We're back at it tomorrow, so i'll take some readings.

Did you exclude leaky coupling caps?

I'll find out, but this was probably done during the first time the amp came in for service.

[J M Fahey]

Also *scope* the MosFet.

It is a peak limiter so I bet you lowered +B by 60V or so, BUT a regular multimeter will give you a much lower average reading,

Agree with "helmholtz" that it would be better to lower screen voltage, plates can take a lot.

I´d use a MosFet to get some 400V DC for them.

[SoulFetish]

update: went to the shop late last night (one of the benefits/detriments of having moved down the street from my work). Made the adjustment to the bias supply to allow the '34s to draw a more appropriate bias current, and now the measured B+ is right around 606V.

[Helmholtz]

EL34s can stand considerably higher plate voltage than 6L6s.
I have a Siemens amp that operates them at 800V/25mA.
But it's essential to keep screen voltage below 450V max.

[J M Fahey]

^^^^^^ that.

When I was a kid, repaired a ton of Philips 100W PA amplifiers, much used in Churches (the other popular brand was Geloso) and train/bus stations, Tango clubs main (only) PA, etc.

Very popular because Philips had a huge factory here so EL34 were "local", while 6L6 were "imported" and so hard to find, and 6V6 unheard of, but mainly because Philips published not only the schematic but more important Power and Output transformer winding.

Supply used a bridge rectified centertapped transformer, so it offered both 800V +B and very good 400V too, excellent for screens, let alone the preamp.

Ordering or self winding the OT was a need because it had 11K primary; published secondary was 500 ohm only.

Fine for PA and distributed speakers, typically transformer fed reentrant horns, a single amp could easily cover 10-15 street blocks with Music/Voice/News as a sort of "Community Radio" equivalent, not exactly usable for regular Guitar/Bass/stage PA, so part of my business was rewinding OTs to more usable 4-8-16 ohm output or building crazy amps with ... count´em ..... SIX microphone inputs!!!!!

Old memories

[alexradium]

yes,leave alone the plate voltage,then regulate the screen and Pi/preamp to 400 Volts using the mosfet and a couple 200v zeners to ground,that helps a lot with noise too,and the power dissipation will be acceptable.

[Helmholtz]

yes,leave alone the plate voltage,then regulate the screen and Pi/preamp to 400 Volts using the mosfet and a couple 200v zeners to ground,that helps a lot with noise too,and the power dissipation will be acceptable.

Not sure how that would be wired, but I don't think regulated screen voltage is desirable in a guitar amp.
That would eliminate the (desirable) sag induced screen compression.

A series "constant voltage dropper" (zener) will preserve the compression.