Negative LTP CCS supply

[Klimon]

Is a negative supply for the CCS in a LTP's tail simply a method to circumvent the need for a very high B+ or are there other advantages?

Thanks

Simon

[ray_moth]

If the LTP is the first stage, as in an all-differential amp, then the alternative to a negative supply woud be to bias up the grid that doesn;t receive the signal and ground it through a capacitor, Connect that grid to the input grid via a 1 Meg resistor and use capacitor-coupling for the input.

If the CCS is made up of BJTs, such as the simple and effective cascode CCS favored by Morgan Jones, you would need a negative supply of about -40v, or you could bias up the grids by 40v. Clearly, the negative supply solution does avoid the need to elevate B+ by about 40v but that's bot really a lot to worry about.

I don't see any other advantage to using a negative supply with an SS CCS but I can see a disadvantage in having to provide a negative supply itself. Of course, if you want a pentode CCS, it's a different story, because that will require over 100v to operate!

[pilli]

...if I may take advantage of this thread for a beginner's question, is it correct in the attached schematic to put something like -10V or -15V to get a few mA out of the JFET?

(The place where I copied it only mentions "a small negative supply", but what is "small" when handling vacuum tube supplies? the B+ here is 300V...)

By the way, the triodes are ECC88.
And the whole thing should provide opposite phases out of C1 and C2.

Thanks.

_

[Bandersnatch]

Originally posted by ray_moth


I don't see any other advantage to using a negative supply with an SS CCS but I can see a disadvantage in having to provide a negative supply itself. Of course, if you want a pentode CCS, it's a different story, because that will require over 100v to operate!

The 100V is an approximation, average likely tube choice etc. It is easy to see a 200 mA CCS that only needs 60V to operate with; with 100V, 500 mA is quite easy...
cheers,
Douglas

[Klimon]

but I can see a disadvantage in having to provide a negative supply itself.

Exactly the reason why I'm asking.

If the CCS is made up of BJTs, such as the simple and effective cascode CCS favored by Morgan Jones, you would need a negative supply of about -40v, or you could bias up the grids by 40v.

Thanks, that should help me along the way.

Simon

[Bandersnatch]

hey-Hey!!!,
A battery biased, or self-biased MOSFET ccs made from Supertex depletion-mode MOSFET's like DN2540N5 and DN3545N3 are good with about 15V of headroom; the capacitance change is mostly over by ~10V.
cheers,
Douglas

[Klimon]

A battery biased, or self-biased MOSFET ccs made from Supertex depletion-mode MOSFET's like DN2540N5 and DN3545N3 are good with about 15V of headroom; the capacitance change is mostly over by ~10V.

Have you got a schematic?

Simon

[Eli Duttman]

Simon,

A big advantage to using a B- supply is the elimination of caps. in the signal path. Only the couplers between LTP and "finals" are present. Fewer phase shifts imply much greater stability.

Look at the "El Cheapo" schematic. The CCS shown is a MOSFET cascode devised by Doug (Bandersnatch).

[ErikdeBest]

Hi Klimon

Here you go:

http://www.diyaudio.com/forums/showt...11#post1603111

If you want I can sell you a couple of DN2540's IC's, for 2,50 euro/each.

Erik

[Klimon]

Erik, I sent you a PM. SY's CCS seems to fit the glove for my 'simple' project (target is 2c51 LTP driving ML6 push-pull or differential).

Eli, what extra cap are you referring to when not using a negative rail? Thanks for your patience

Cheers,

Simon