The Official Dual Rail Thread

[troposnet]

radiostar, its done. thanks.

negative bias with multi-tap transfo

[thegunch]

Hi all....first post here. I've got a question that applies to dropping screen supply voltage in order to achieve more output power. For reference I have a build that has a Marshall/Fender style power supply, running 2xKT88 in fixed bias with ~510V on the plates and a few less on the screens. Clearly Hiwatt, Ampeg and a few others provided lower screen supply via dual PT secondaries and a two rail configuration, as this thread is largely about, however im curious about modding an existing amp with its current transformer. There is the Mosfet voltage dropping circuit as described by RG on page 3 of this thread, which certainly looks useful *and* easily adjustable, but i'm wondering why a simple voltage divider was not offered up a an option? Is there some reason im missing that this s not feasible? Just two resistors in the screen voltage supply to drop it to the desired level.

[SoulFetish]

Hi all....first post here. I've got a question that applies to dropping screen supply voltage in order to achieve more output power. For reference I have a build that has a Marshall/Fender style power supply, running 2xKT88 in fixed bias with ~510V on the plates and a few less on the screens. Clearly Hiwatt, Ampeg and a few others provided lower screen supply via dual PT secondaries and a two rail configuration, as this thread is largely about, however im curious about modding an existing amp with its current transformer. There is the Mosfet voltage dropping circuit as described by RG on page 3 of this thread, which certainly looks useful *and* easily adjustable, but i'm wondering why a simple voltage divider was not offered up a an option? Is there some reason im missing that this s not feasible? Just two resistors in the screen voltage supply to drop it to the desired level.

Well, you could. But there are better more effecient circuits that achieve a better screen supply. Depending on how much you want to drop the voltage, you may place unnecessary loading. Plus you dont want to run the screens directly from a divider as you end up with too much supply ripple. Im guessing the circuit RG posted was in the relm of a source follower with a zener to set the gate voltage? (forgive me if I don't go through six pages of thread from 2013). Those circuits are super useful and probably better suited for a screen supply IMO. What voltage do you want to run the screens at?

[thegunch]

Well, you could. But there are better more effecient circuits that achieve a better screen supply. Depending on how much you want to drop the voltage, you may place unnecessary loading. Plus you dont want to run the screens directly from a divider as you end up with too much supply ripple. Im guessing the circuit RG posted was in the relm of a source follower with a zener to set the gate voltage? (forgive me if I don't go through six pages of thread from 2013). Those circuits are super useful and probably better suited for a screen supply IMO. What voltage do you want to run the screens at?

Re: what voltage somewhere closer to 300V, I think. Old Genalex/etc. datasheets specify 560 plate, 300 screen for 100 watts under typical operation. And if I did this i'd be dividing the voltage taken off the already existing node supplying the screens, so I dont know that any additional ripple would be introduced.

[SoulFetish]

Re: And if I did this i'd be dividing the voltage taken off the already existing node supplying the screens, so I dont know that any additional ripple would be introduced.

Then I see no problem with that. I wanted to double check some historical references, so I read the Audio power amplifier chapter in RDH4 and it seems fairly common design practice to feed screens from dividers in reference to ground, cathodes, or other nodes.

[hylaphone]

Hi folks, just read through this thread and have a question regarding PT specs.

I'd like to try this with separate transformers for plate and screen supplies.
Should they each be rated for their max current? For example 800VCT @400mA, full wave rectified for plate supply, then maybe 250V @ 100mA bridge rectified for screens and preamp?
Or do I need higher screen current available?
Thanks--

[Jazz P Bass]

"I'd like to try this with separate transformers for plate and screen supplies."
Why would you want to do that?
A simple high wattage resistor will do for each screen.
If you use a lower voltage than the plate then you will see a drop in gain.

[Helmholtz]

2-level power supplies are used when plate supply is so high (say considerably above 500V) that you would need to drop maybe 100V or more for a safe screen supply voltage.

[hylaphone]

Well, for example the GEC KT88 datasheet has the example of 100W 2x KT88 amp with 560V plate, 300V screen supplies. My understanding is the screen supply needs to be pretty stiff hence all the different ideas in this thread of dedicated and or regulated supplies.
I'm looking to use this topology but scale it back slightly for 70W.

Instead of stacking the plate supply on top of the screen supply, why not separate them completely?
This way the plate is rated for max plate current, and the screen transformer is rated for max screen current + preamp. Am I missing something?

[Helmholtz]

Well, for example the GEC KT88 datasheet has the example of 100W 2x KT88 amp with 560V plate, 300V screen supplies. My understanding is the screen supply needs to be pretty stiff hence all the different ideas in this thread of dedicated and or regulated supplies.

Exactly.
Dropping that much voltage with resistors would result in extreme screen compression.

The KT88 indicates a max. signal screen current of 30mA (@ 100W) for both tubes. I think 100mA rating for the screen supply transformer should be fine.

[hylaphone]

Exactly.
Dropping that much voltage with resistors would result in extreme screen compression.

The KT88 indicates a max. signal screen current of 30mA (@ 100W) for both tubes. I think 100mA rating for the screen supply transformer should be fine.

Thanks for confirming. I got derailed by post #78, until I realized they were talking about a stacked supply.

[SoulFetish]

Hi folks, just read through this thread and have a question regarding PT specs.

I'd like to try this with separate transformers for plate and screen supplies.
Should they each be rated for their max current? For example 800VCT @400mA, full wave rectified for plate supply, then maybe 250V @ 100mA bridge rectified for screens and preamp?
Or do I need higher screen current available?
Thanks--

you could use 2 separate transformers, but it you were looking to run the screens of by a separate supply, i would consider this the least efficient way of doing it.
I would design or look for a power transformer with either dual taps in each leg of the ht secondary, or run it off a separate dedicated secondary winding altogether.

[Helmholtz]

In principle a separate screen PT would allow for a stiffer screen voltage - if that's the purpose.
Some tube Hifi amps even use active screen voltage regulation for max. dynamic response (= least screen compression).

[nevetslab]

In principle a separate screen PT would allow for a stiffer screen voltage - if that's the purpose.
Some tube Hifi amps even use active screen voltage regulation for max. dynamic response (= least screen compression).

That's what Audio Research has in their aged D70 Stereo Tube Power Amp that I just received from MEF member sgelectric. Amp has two pairs of 6550's with a third 6550 as part of the regulated screen supply. At present, the screen supply is blowing fuses, so I'll have to dig into circuits to find the cause and solution. I'll post schematics when I get a chance.

[Helmholtz]

That's what Audio Research has in their aged D70 Stereo Tube Power Amp that I just received from MEF member sgelectric. Amp has two pairs of 6550's with a third 6550 as part of the regulated screen supply. At present, the screen supply is blowing fuses, so I'll have to dig into circuits to find the cause and solution. I'll post schematics when I get a chance.

I wonder if/how they took care of Vhk limit of the series regulator tube.