Zen I/V Converter

Mr Pass,

Isn't Witten he also a senior officer in the Symmetry Police? :)

Back to the circuit...

In the last paragraph of the article you suggested that C3 and C4 could be removed entirely to increase the low frequency performance. Is this a typo?

I am considering paralleling multiple pairs for a Saber I/V stage and it would seem that removing these all together, even in the circuit as drawn, would have a negative effect on the circuit at all frequencies.

The "to avoid DC arguments" a few paragraphs above is prety clear, but I thought it worth asking to avoid confusion for anyone trying to parallel multiple current outputs (or multiple devices) for a single channel.

Thanks for the great circuit.
Dave
 
input impedance is too high and transconductance a bit too low for best current mode output with the sabre. i'll still try it out once people more skilled than I have made a few tweaks to the circuit. indeed the simulations look very good; but the real world measured performance of owens rendition of the D1 mosfet IV are better. i'm already running an older version D1 IV PCB with some modifications and am very happy as is, but yeah its possible papa past trumps papa new WRT sabre dacs
 
Ok here is an off the wall idea for low current R2R DAC's. Set the gain for 0.5V (lowest distortion), paralell Jfets to give lower input impedance. Take one of the outputs and DC couple to a tube anode follower with a parafeed transformer. I know tubes and DAC's don't mix well but this might be a good integration without the need for passive I/V or step-up transformers.
I see a lot of NOS tube designs here and this would be much better than the passive I/V's I see on all of them, I might try it but with a PMD100-AD1862 oversampling DAC. Without a tube this Zen I/V seems to have too many compromised for classic +-1mA R2R DAC's.
 
Hi all,
i'd like to try an I/V stage with this concept on my AD1865 dac, that actually uses a 200 ohm resistor for I/V conversion, but didn't like the push pull idea (it's difficult for me to find matched 2sk/2sj). Did anyone think that a SE circuit like the one below can work well, even with reduced gain due to the missing negative part of the circuit?

(sorry for the link, didn't got a host that support direct linking)

Picasa Web Album - bonalux.lc8

If there is a chance that it can work i will try it, or any suggestion will be appreciated.

Cheers.