[tranhieu]

So far there's been no reply in the main thread in diyaudio so I bring the question over here. Hope someone can enlighten this for me.
http://www.diyaudio.com/forums/group-buys/156149-tube-i-zator-professional-pcb.html
In short I'm sourcing parts and I'm not sure which value for the Rk I should pick. Lampizator said he chose 39R, while someone I know who built the DAC-end with similar output stage told me to go with 100R. But what I really want know is based on what formula or equation can you get the correct value?
 
My DAC is the Buffalo es9008, in the manual it says that
 
Quote:

The differential output of the buffalo is 0.924*1.65V (3.1746 Peak) + 1.65 DC bias and the series output resistance is nominally 195 Ohms at each output.

I'm thinking about picking the 6n2p or the 12ax7.
Thanks for reading.

 

[nikongod]

You pick the resistor based on the operating point of the tube you want to use, and the voltage offset from the DAC chip. 
 
I could not find a schematic with a few clicks through that thread, do you have one?
 
I'd lean away from the 12ax7 unless there is a SS Buffer after it. 
 
The output impedance of a 12ax7 SRPP (I think the lampizator is SRPP, schematic?) is relatively low which is generally good, but the relatively low currents the tube is capable of make driving anything lower than a 50Kohm load iffy which IMO outweighs any advantage from output impedance. When you consider cable capacitance its not hard to push a 12ax7 over the edge.
 
I should note: I'm not such a fan of tube DACs, or the 12ax7, but maybe the goal is to run them ragged to sound more tubey? 

 

[tranhieu]

Below is the schematic for the pcb. 
Why is there DC offset here? I thought it would have been filtered off before reaching the output stage?
The signal will be fed into the B22 with a 50k attenuator so I'm not worried too much about the impedance mismatch. Perhaps not in the near future.

 

 

[tranhieu]

Opps, just realized I misread the diyaudio posts. It should be 200R for the Rk, but for the Ri/v Lampizator said 39R while Brian from TPA said 'as close to 0 as possible'. Is there anything I'm missing here?