Zen I/V Converter

For your simulated circuit, you should consider changing :

1) C1,2,5,6 to 10uF minimum
2) R8,16 to 10k (using 500ohm to drive a 1k load is a bad idea)
3) add caps (100uF) to gnd after R9,10,17,18

I would use a complimentary follower with a single pair of 2SK170/2SJ74 and no source degeneration. Hardly any distortion with 10k load.


Patrick

I will try all of those suggestions

Regarding 1) I thought I could lower the caps because the circuit is now seeing 1 Meg as opposed to 10k or 20k.
 
so biasing source at AVCC/2 (or close enough, the exact V doesnt seem to matter all that much) is no problem due t o being Vground right? for sabre users input Z really should be quite a bit lower, i'm not even sure paralleling the jfets will get us there for best DNR; but the sound is another matter I guess. I have DCB1 PCBsready and waiting for a reason and this seems like a good one, but I need to tick a few more boxes yet. BTW this would be for ackodac not buffalo II. same diff though as far as numbers that matter here.

linear systems JFETS would seem the logical choice here no?
 
Do you have any data on that? The manufacturer does not
mention voltage degradation in the data sheet.

:cool:

TDA1541A datasheet (page 7) says:
"Notes to the characteristics
1.To ensure no performance losses, permitted output voltage compliance is ±25 mV maximum."

I made 3 DACs with that chip, tried various I/V solutions and found out that it really sounds best with DAC's output as close to 0V as possible.

Snap from datasheet:
 

Attachments

  • ds.gif
    ds.gif
    11.1 KB · Views: 1,970
Last edited: