Balanced or parallel DAC chips

[WTS]

Hi; Quote....One common practice that seems guaranteed to mess up
the DAC's linearity is to use a resistor on the output of
a current-out DAC for the I/V conversion rather than an
op-amp summing junction. Saw some curves that a
friend did recently and the resistor version was pretty
pathetic.

I'm sorry but I would have to disagree with this. But it would also depend on which dac you were using I suppose. If you are using the AD1861 or 1862's by Analog Devices, then using a resistor for the I/V conversion is the only way to go(imho.. these dacs are the best you can get period, the 1862 being the better of the two). I used two dacs in a balanced config.(inverted data feeding the second one) feeding a bridged power amp direct and made the input R ... I beleive the value I used was a 562R, Caddock MK132....the load for the dac(i/v). To balance out the amp I used the same value on the inverting side. For volume adjustment I use a bit shift setup before the dacs. I say without a doubt that getting away from any opamps in the signal path is the best thing you could ever do to improve the sound. The amps are my own design( or should I say a compulation of designs and knowledge gathered over the past 20 some years).

And one other note, if you want to improve the sound of your equipment even more, get rid of those gawd awful electrolytic caps,all of them and replace them with polypropolene. In my power amps, which are basically bridged 200 watt modules, in the base amps I use a total of 800uF per rail..72 v rails and in the amps for the upper ends I use 600/rail... Solen 200uF/400v. Talk about clean and powerfull bottem end. I didn't know what good bottem end was until I made the swap.

[BrianL]

WTS

Well, you may disagree, but them's the facts. You are correct
that some current-output DACs may behave differently than
others, but I've seen enough to say that in general, the linearity
(and hence distortion) will be significantly worse with a
resistor "I-to-V" converter.

You can debate whether it "sounds" better or not, but
the objective fact is that the DAC feeding the "zero" impedence of
transimpedance stage of suitable quality will have less distortion
than a DAC driving a resistor (and possibly some active
stage after that).

My experience has been that better objective
measurements have led to sound that was more pleasing
in the long term. There is valid debate as to which "objective"
measurements are best correlated to the audible experience,
but I've not seen and cases where consciously choosing
circuitry with less good performance has created long-term
satisfaction.

[WTS]

Hi BrianL;
Well those may be the facts, but there is a group of us in the city which I live and we all have a electronics background right from techs to profs and dean's of science's, as well as some members of the local symphony. Now you can have all the smarts in the world and do all the tests you can and want to say this should sound better than that, but it all comes down to how it sounds to your ears. As far as this change or implementation of the dac goes, we found without a doubt that it was by far more accurate to the ear than the conventional methods. Now none of us claim to have golden ears or anything, its just our findings.
Thanks.
And let me say that these listening tests were not done using second rate supporting equipment or done in one seating or in the same location.