Valve Itch phono

I think a well choosen IR is by far more linear then 1N4148 at those currents. A bypass, like a low noice 100k from top anode to diode anode might get the current up to linear levels, and maybe to levels where grid current stops with the strong bias level of one diode.

Edit. Maybe not on second thought. It looks hopelessly nonlinear up to some 50 mA. http://www.nxp.com/documents/data_sheet/1N4148_1N4448.pdf

Would a low noice 1k-2k be worse noisewise then the unlinear effect of the diode to the sound?
 
Last edited:
diyAudio Chief Moderator
Joined 2002
Paid Member
This tube we use has good GM to small Ik ratio presenting low hiss plus very quiet mechanically, and it got slowly recognized for phono input use these days, obscure if not used at all for that back when used in the Itch, so its worth solving for Ig carefully by sample basis. They generally differ by sample when it comes to grid current negative bias threshold, but its relatively low always.
 
I made an attempt to measure grid-anode and grid-cathode capacistance after running a few 6N2P's in the AVO. Since a difference of 1 pF can matter quite alot, especially concerning that Cmiller= C(1+Af), important for the second stage and riaa comp values it was interesting to see that Cga and Cgk measurments differed quite a bit on how I held the crocs. If I added a socket on the tube it increased by nearly double. Then I tested just with the crocs distance and got a difference of like 2 pF just by holding them near or far away from each other. A lession for when doing interconnection cabling, yes indeed. Croc/cabling distance is far more important then tube sample differences concerning grid capacistance.

Measures made with DER EE DE-5000
 
diyAudio Chief Moderator
Joined 2002
Paid Member
The method used for determining Miller is comparing -3dB upper limit in a common cathode circuit between two different source resistances (virtually series to the grid) then calculating back from the measured Av. It takes gen & scope.

With the LCR you could roughly say a type has more parasitic than another when steady for wires, but to be sure of the absolute figure is quite tricky.