Zen I/V Converter

[EUVL]

You can use the reference pin plus a buffer.


Patrick

[Nelson Pass]

Originally Posted by cviller

1. Would it be an improvement to have a capacitor on the "wiper" pin to gnd? 2. Could the jfet gates be connected to the reference pin on the dac chip instead of the pot? (I don't know the ESS that well, but many TI chips have a reference pin)

1) You would likely want some resistance in series with such a cap to
prevent parasitic oscillation, so you can just use the pot for that.

2) If you want the Sources to be at the voltage of the reference pin, and
the fets are adequately matched, don't see any reason you couldn't.
Keep in mind that the reference pin may or may not be as quiet as you
want, so RC filtering might be appropriate.

[EUVL]

Hope Nelson would forgive me for advertising for Jan Didden.

I wrote an article for Jan's Linear Audio (Volume 2) in June titled "Zen->Cen->Sen IV".
It made some interesting comparison between 3 minialisitic IV designs, all based on a pair of matched JFETs and a few components.
Both in simulation and hardware.

I shall open a new thread at the digital forum to show some details and pictures after the Volume 2 is open for order.
Should be some time next week.


Patrick

PS
Correction -- it is already open to order now at http://www.linearaudio.net/.
I shall post later today.

And yes, Jan has my admiration and support for devoting his time and finance after retirement to publish "Linear Audio".
For our benefit rather than his.

.

[EUVL]

Posted here :

Zen -> Cen -> Sen, evolution of a minimalistic IV Converter


Patrick

[fireworks]

Interesting discrete circuit for an I/V converter.
However, I cannot help but wonder: what is the linearity of this circuit ?
A high end 24 bits DAC has a very high linearity. If this I/V converter has worse linearity than the DAC, the resulting linearity will be correspondingly worse.
In other words, the I/V converter has a very hard task ahead, otherwise it is the weakest link in the channel.
Is this circuit up to this task ?

[rsdio]

Originally Posted by fireworks

what is the linearity of this circuit ?
A high end 24 bits DAC has a very high linearity. If this I/V converter has worse linearity than the DAC, the resulting linearity will be correspondingly worse.
In other words, the I/V converter has a very hard task ahead, otherwise it is the weakest link in the channel.
Is this circuit up to this task ?

Excellent question. Would SPICE simulation be able to provide an answer? I've used simplified current output models for high-speed DACs to aid in simulation of the I/V converter (using op-amps), so I assume that the FET/JFET version could be simulated as well. It's tough to find a SPICE model of the entire DAC, but sometimes you can do well enough with just the current output circuit, including any parasitic capacitance on the outputs.

[tiefbassuebertr]

arround the three I/V (I/U) converter topologies about
http://www.passdiy.com/pdf/Z-IV.pdf
there are a wide range of posts, and I am not able to read them all.
I have heard only various op amps like OPA637, AD829 and OPA827 against a normal resistor with follow high input impedance line amp so as the D1-topology. From my view the D1 topology sounds at best, follow from the "only one resistor" topology about
http://www.dddac.de/pcm63/DAC_I-V_Resistor.htm
I/U converter that used op amps like the above mentioned still sound "over analytic" and a bit harsh like a tart wine.
The topology with jFETs so as various other discrete op amp solutions I have never heard.
Where can I find a description about the perceived audible differences between this various topologies?
Is there an I/U topology in the meantime, that beats clearly the D1 I/U converter ?

[Zero Cool]

I don't quite understand how this went from a simple schematic with one 2SK170 and one 2SJ74 to something far more complicated??

Has anyone built and tried the circuit as Nelson posted it??

[jazz]

Yes, I built it and it hasn't been out of the system since.

grtz
Joris

[PA0SU]

Originally Posted by tiefbassuebertr

Where can I find a description about the perceived audible differences between this various topologies?
Is there an I/U topology in the meantime, that beats clearly the D1 I/U converter ?

on my web site: by-rutgers.nl