parafeed - Gyrator vs CCS anode load

[adamus]

Hello all.

I am currently messing around with parafeed headamps.

Does anyone have any comments regarding the difference in sonics between a gyrator and a CCS load?

CCS load is cascode pnp with led Vrefs (morgan jones).

I could knock up a gyrator, i think i have a couple of dn2450 and a variety of film caps. Anyone have a prefered schematic?

Thanks

adam.

 

[bigwill]

Gyrators are great because you can manually set the quiescent anode voltage, and you set the current with the usual cathode resistor (or LEDs if you're cool). No worrying about adjusting a CCS, the valve will draw what it wants

 

[Hearinspace]

I don't whether it will fit your needs in this case but Wavebourn's SVCS is a beautiful thing.
It's mentioned a few places on the site. Search SVCS

 

[adamus]

Thanks for the link. I should add this is fed by a salas shunt supply, so I have a nice clean supply. Could I get by with a single dn2450 or is more sand required. Afraid I am a follower here so advice welcome.

 

[Wavebourn]

You can probably be fine with a single dn2450; all depends on which tube in which regime do you use, and which load does it drive.

In short, gyrator load on AC is identical to CCS, but on DC is not. Similarly to a choke, but not exactly.

 

[adamus]

C3g cap coupled to a 10k edcor.I have been playing with operating points and 14 ma , 190v plate seems to sound best.

 

[Wavebourn]

I can't find data on dn2450...

 

[adamus]

Ooops.... try dn2540n5 http://www.supertex.com/pdf/datasheets/DN2540.pdf

sorry for the typo

 

[Wavebourn]

Ok, thanks!

It looks like you need 50V swing. That means, something like about 40V voltage drop on gyrator is fine. Output capacitance will be still low on max swing, and power dissipation about 0.5W that is fine without heatsink. 230-250V B+ will be needed.

I would use it with something like 360 Ohm in source, for 5V voltage drop. That means 5-1=4V between gate and source. If to use 470K trimmer to set anode voltage and 2.2 megaohm between gate and drain, you have plenty of headroom to trim working point for your tube.

Here is the example, but using BSP225 enhancement mode MOSFET that required different values of resistors (C3 should be 4-10 times higher, actually):

http://wavebourn.com/forum/download.php?id=493&f=7

 

[Bigun]

Nice design Wavebourn.

I'm curious (still learning, as ever), why did you opt for a tube (V2) as buffer to drive output tube if you allow sand in the game you could also use a source follower ? or even use another gyrator for the cathode load for V2 ?

 

[Wavebourn]

Nice design Wavebourn.

I'm curious (still learning, as ever), why did you opt for a tube (V2) as buffer to drive output tube if you allow sand in the game you could also use a source follower ? or even use another gyrator for the cathode load for V2 ?

Source follower would add own capacitances. I tried to minimize them.

Here is one more version, without gyrator:

http://wavebourn.com/forum/download.php?id=494&f=7

You can use 6P1P there instead of 12L6GT, with 2K7 cathode load resistor.

 

[adamus]

wavebourn, you may have to forgive my lack of knowledge here... but will something like this work

An externally hosted image should be here but it was not working when we last tested it.

 

[bigwill]

That will work, all I'd do is make R1 bigger and in real life use a zener from the gate to source to protect the gate from overvoltage

 

[bigwill]

Oh, and you can take output directly from the source to get a lower output impedance, which is nice!

 

[ErikdeBest]

That will work, all I'd do is make R1 bigger and in real life use a zener from the gate to source to protect the gate from overvoltage

A question about the zener: it is commonly applied to protect enhancement mode mosfets, but will it also work with depletion mode mosfets, as the DN2540 discussed here?

erik

 

[bigwill]

A question about the zener: it is commonly applied to protect enhancement mode mosfets, but will it also work with depletion mode mosfets, as the DN2540 discussed here?

erik

I'm not sure! I forgot DN2540 was depletion mode. It won't hurt, but I'm not sure what protective role it will have. You might need one the other way around with depletion mode FETs, to protect the gate from excessive negative voltage with respect to the source. Someone more knowledgeable than me will be able to answer this.

 

[zigzagflux]

I put in back to back zeners as insurance in either direction, and have had good results. Some of my designs seem to place tough conditions on the FETs, during startup conditions when the tubes aren't sufficiently heated. Lost a few of them until I installed the zeners.

If I recall they were 9.1V; during normal operation most of my CCS's run with a volt or two on the gate, so there is really no effect when installed.

 

[bigwill]

Back to back seems to be the way to go then, for either type!

 

[bigwill]

Wait - This will limit you to 0.7v on the gates due to the forward voltage....

Edit: Ignore me, back to back is not anti-parallel

 

[Bigun]

How does this gyrator design compare with using a choke load ? I have recently found out that Edcor output transformers are as cheap as $17 each, which in your application would remove the need for many components and provide safety isolation for your headphones against any danger of B+ appearing at the output. Simpler to wire up, no heatsink and could allow a lower B+ depending on how much is being dropped across the gyrator. The only downside I can see is that your B+ needs to be a lot cleaner since PSRR is lower with the choke.