PSRR For 6SN7 Buffer?

[G]

Hi All,

Would a RCLCLC filtered supply bring the supply ripple down low enough to be acceptable for powering this buffer? The caps are 220 uF except for the 10uF cap right after a 5AR4. I'll post another diagram later. Also in question is the sound quality to be had from a tube buffer. I think that it will work better than a passive pre but I'm not sure. Opinions?

[fdegrove]

Hi,

PSRR and ripple are not quite the same thing.

You may have good ripple filtering but still have to deal will poor PSU rejection.

In the case of your proposed PSU it can be good but IME the CF is best served by a stiff PS since it has by nature not so good PSRR.
One way to aviod is by using a bipolar PSU where a good deal of PSU related noise is canceled.

Regards,

[G]

So you think that it needs more capacitance in the PSU? I've never thought about using a bipolar PSU for a tube circuit. Interesting concept.


G

[G]

Here is the diagram for the PSU:

G

[fdegrove]

Hi,

So you think that it needs more capacitance in the PSU?

Well,you could try and give it a listen for a week,then remove some till you find what suits you best.
Stick with the same brand/type of cap so you're not bringing other parameters in.

I've never thought about using a bipolar PSU for a tube circuit. Interesting concept.

This is nothing new (to me at least),when you have to drive the impossible output tube this usually does it,and has the added advantage of cancelling PSU noise.

As far as the IXYS goes,I've no idea if you can float it in a negative supply but basically I don't see why not.
If not,there may be a counterpart PNP/NPN whatever it is.

Ciao,

[G]

Peter Millet says that the Ixys part acts basically like a high gm triode. I'm also not sure if it will work as I have it shown. If not then I can use a CCS made with discrete components. How would you apply a bipolas PSU to a tube. I assume that you would not put a bias voltage on the grid.

G

[fdegrove]

Hi,

Biasing of the grid can be done in the usual various ways:

One clever example of how to derive it from a simple supply is shown in the Futterman circuit:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=7010

Take a look at the second picture.

When you look at this circuit you notice the 12B4As are biased at
-24VDC.
B+ = 157VDC on the anodes.

The upper tubes get this voltage from a voltage divider put in the negative rail.
The lower tubes which have their cathodes at a higher negative potential due to the extra PSU 30VAC and get their negative bias supply from this .
B- = -157-24VDC = -181VDC.

In this particular case both triodes are stacked but it works exactly the same with a single triode.

You could just as well use a seperate bias supply,I'm just pointing out the basic principles.


Ciao,