[dave slagle]

Romy put up a similar schematic from a russian site and shines and I were scratching our heads at it.


I entered it into spice and had to take some liberties with the first tube (used Stephies D3A model) but the basic picture is here



Ignoring the bass resonance for the moment it is clear that th second time constant is missing. Romy then mentioned that the design relied on using the inductance of a MM cart as one of the filters at which point the 22K variable resistor at the input made sense. I put the 250mhy of L into the sim and twiddled the 22K value to get this.



Looking much better. Now for the resonance. 14Hy's is a small value for the 6e5P and upping that lowers the resonance frequency.


I know in reality there will not be a a 30dB peak at 20hz and even though more inductance will push it lower (as will increasing the bypass cap) I think a better approach would be to supply the screen voltage from it's own supply.


This gets us close enough to give me confidence that by tweaking the values we could get better response. I'll attach the file below if anybody wants to play.

dave

[dave slagle]

For some reason I am not comfortable with using the cartridge L as part of the filter and chances are I'd want to use a MC anyways. Simply adding a cap across the 1K5 resistor gets us the needed constant that the MC cart inductance did.

I twiddled the values and everything now is looking pretty good.

[reVintage]

Hey Dave,

I have been playing a similar RIAA for the last 30 years albeit with cascode insted of pentode and 1*gain second WF-stage. Works great probably due to the lack of EQ series-resistor. Think the EQ I use is simpler to adjust though. The second transformer-out stage seems nice but must have an ample overload margin as the first stage will have 30-40dB gain.

Funny you should bring this up now as I am currently working on a hybrid version with an innovative topology.