Advice needed- best way to lower plate voltage

[jutsin]

Hello everyone,
I've been lurking around the forums here, and have got a question for you smart guys :-)
I've got an EL84 amp (Mesa) that I installed a choke in (bypassing the resistors that were in it's place) and now my power tubes are gettin whopped with a ton of plate voltage (a little over 450). What's the best way to bring the plate voltage down without putting back in resistors that rob dynamics and feel? I heard someone suggest using a zener diode. Has anyone done that, or have any other suggestions? Any help is greatly appreciated!

[Enzo]

WHICH Mesa amp is it? That will make it a lot easier for us to look at schematics.

[jutsin]

Mesa Boogie DC-3. It's the B version (slightly different than the A which is the schematic found online). Is there a way to post the schematic here?
It looks like the only difference in the high voltage supplies is that the B version has 6x 680 ohm resistors in series instead of the 5x 470 then 1 680 in the A version.

[melvin]

I believe your power tubes should see about the same voltage with or without the choke because it is typically placed after the center-tap connection to the output transformer like the string of resistors I am expecting you replaced. Since the power tubes didn't see any dropping resistors on the B+ in the original schematic, they shouldn't see a higher voltage after the modification.

Are you testing the plate voltage with the tubes in? When out, the plate voltage will test higher, when in they should bring the B+ down (marked A on the schematic). This is due to resistance in the power supply up to that point and the current draw of the EL84's in idle.

Also, with the choke in place of the resistors, your screens on the EL84's will see a much higher voltage, not to mention the preamp tubes. This could lead to instability, so be careful. Furthermore, I would use the Sovtek EL84M's. They seem to be the only ones rated for 400V on the plates. But I haven't tested others in a Mesa yet. They may work fine with the correct bias current.

[jutsin]

I believe your power tubes should see about the same voltage with or without the choke because it is typically placed after the center-tap connection to the output transformer like the string of resistors I am expecting you replaced. Since the power tubes didn't see any dropping resistors on the B+ in the original schematic, they shouldn't see a higher voltage after the modification.

Are you testing the plate voltage with the tubes in? When out, the plate voltage will test higher, when in they should bring the B+ down (marked A on the schematic). This is due to resistance in the power supply up to that point and the current draw of the EL84's in idle.

Also, with the choke in place of the resistors, your screens on the EL84's will see a much higher voltage, not to mention the preamp tubes. This could lead to instability, so be careful. Furthermore, I would use the Sovtek EL84M's. They seem to be the only ones rated for 400V on the plates. But I haven't tested others in a Mesa yet. They may work fine with the correct bias current.

Hi Melvin,
thank you for the reply. i just measured the screen voltage, and it's about 472V. I am measuring the plate voltage with the tubes in. I thought that the choke shouldn't effect the plate voltage as you mentioned, but it is clearly effecting the screen. The only other things I have changed is upgraded the OT and put in new power/bias supply caps (same values as OEM).

[ThermionicScott]

So the screens are at 472V, and the plates a little over 450V? I'd recheck your wiring...

[jutsin]

So the screens are at 472V, and the plates a little over 450V? I'd recheck your wiring...

That's what I got when both measured to ground. If I measure from cathode to plate I get 474V.

[melvin]

Perhaps you should attach the schematic, the cathode should be at ground according to mine.

[tubeswell]

Hi justin

Can you measure all the pin-to-ground DC voltages under idle conditions on all the tubes and report back - specifying which pin # gives what voltage on which tube? (That way we will be sure which voltages are plate voltages and which ones are screen voltages)

[Prairie Dawg]

Hi justin

Can you measure all the pin-to-ground DC voltages under idle conditions on all the tubes and report back - specifying which pin # gives what voltage on which tube? (That way we will be sure which voltages are plate voltages and which ones are screen voltages)

If your measured plate voltage with the tubes installed is above 350v or so it's giving those EL84s a beating.

According to the schematic plate voltage should be 400v and the screens should measure around 300v.

So something's going on there that's odd.

You can lower your B+ with a zener diode. I used a 50w NTE to lower the B+ 50v in a Rose Morris s/s rectified AC30 and it settled right down and stopped all that redplate foolishness. I would try and find a good place on the chassis that has some steel around it because the zener will throw off a fair amount of heat. A good heat sink compound is a must, and if real estate is at a premium I might think about an external heat sink. I do not think I like the Weber V-Dump because mounting it on a transformer screw will dump all that heat right into the power transformer. Weber does have a wattage calculator which is useful, but I'd say a 50w rated zener will do fine, you can select them for pretty nearly any voltage drop you want up to and including 100v.

http://schematicheaven.com/boogieamp...aliber_dc3.pdf

[jutsin]

Hi justin

Can you measure all the pin-to-ground DC voltages under idle conditions on all the tubes and report back - specifying which pin # gives what voltage on which tube? (That way we will be sure which voltages are plate voltages and which ones are screen voltages)

Hi,
yes I'll do that as soon as I get off of work today and post the findings!

[Steve Conner]

I think you bypassed the wrong resistors when installing your choke.

Even if they "rob dynamics and feel"- which I don't believe for a minute- they also were probably supposed to stop your tubes going into meltdown by dropping the screen voltage. Now people are talking about adding zeners to bring it back down, but I think we're trying to fix something that wasn't broke until you started messing with it. Just put those dropping resistors back in, and splice the choke in further upstream, where it doesn't cause such a voltage rise.

Edit: I looked at the DC3 schematics, and there isn't anywhere else you can put it! Except maybe bounce it off the nut of whoever told you that resistors kill tone.

Steel is a lousy heatsink. It has something like one-eighth the thermal conductivity of aluminium. I remember once trying to heatsink some TO-3 transistors to a steel sheet, and it was a total failure. A 1" circle around them got sizzling hot and the rest of the sheet stayed cold.

[jutsin]

I think you bypassed the wrong resistors when installing your choke.

Steel is a lousy heatsink. It has something like one-eighth the thermal conductivity of aluminium. I remember once trying to heatsink some TO-3 transistors to a steel sheet, and it was a total failure. A 1" circle around them got sizzling hot and the rest of the sheet stayed cold.

I think I did as well haha. The guys at M/M said those were the ones to remove so I went with that, but it doesn't make sense when you look at the schematic- and see the screen voltage results in my amp!
As far as the heatsink portion goes I'll have to do some planning (perhaps an aluminum bracket similar to Weber?) or a heat sink. Thank you for the replies!

[Steve Conner]

Oh, well if you still want to use the choke, I'd put it ahead of the first filter caps, and add another filter cap in there. But that's starting to get more complicated and take up more space.

450V still seems awfully high. Even if you bypassed the droppers, there shouldn't be any voltage above 400 in there... What's your line voltage? Or maybe Mesa bumped up the plate voltage for the Rev.B and that's why they increased the dropper resistors, to keep the screen voltage under control.

[melvin]

If you really want more dynamics and feel, you might try bypassing the whole solid state effects loop/ graphic eq section.