Forget Opamp I/V stages.

Sep 14, 2006 at 6:03 AM Post #16 of 25
Quote:

Originally Posted by jcx
for the op amp based I/V converter it is interesting that so many have forgot/never learned what was commonplace app note knowledge ca 1970:

In the early days of monolithic op amps and Iout dacs it was well known that the highly desirable low pass filter forming feedback C in the op amp I/V is a feedforward short circuit for high frequency switching events, this majorly jerked around the asymmetric NPN output stages of early op amps causing distortion, the known solution was to add a low Z Class A buffer in the loop biased to give a low Zout with the worst case current step



i think that the op wrote the above, only in comon english.
maybe i just didnt understand what i read properly.
 
Sep 14, 2006 at 1:57 PM Post #17 of 25
Thanks for that link Jcx. Looking at the TOC there may be a lot of useful info in there for future reference.

Hey T52 since you showed interest in this in the first place I've made several changes for the better to the DAC. I'll post the updated schematic tomorrow.
 
Sep 14, 2006 at 8:11 PM Post #18 of 25
Quote:

Originally Posted by Garbz
Hey T52 since you showed interest in this in the first place I've made several changes for the better to the DAC. I'll post the updated schematic tomorrow.


yep, very cool... i'm awaiting the package from ezkcdude with his proto board and all needed chips... for his reference i'll finish the dac board 100% after his schematics, but i plan to test other i/v approaches as well. one of that may very well be the modded d1 stage.
 
Sep 15, 2006 at 6:14 AM Post #19 of 25
Ok all done. New schematics are http://www.garbz.com/amp/dac/passd1-mod.pdf

Changes are to the current source for the source follower. The resistor returns to the -ve rail rather than ground allowing a higher value and thus a more robust CCS. The supply rails were also labeled as +/-32V, output resistor was removed, and a resistor to ground was attached to prevent massive DC at poweron while the cap charges.

With 32V the power through the new resistor is 1.016W thats why there's 2 in parallel. A single one will work but you'll end up with a resistor that runs at 95 degrees, that said it is working in my case.
basshead.gif


Also I was having slight problems with noise the other day. The entire I/V stage was horribly unstable resulting in AM pickup on my interconnects. The problem turned out to be me drawing more power from the powersupply than the transformers were capable of delivering. Those pissy little 5VA 30V xformers wern't enough
eek.gif
so now I replaced them with 2x 30VAs
very_evil_smiley.gif
. Btw many thanks RescueToaster for helping diagnose the problem as well as suggest some of these changes to the simplified design. If I keep upgrading things like this I may just add a proper CCS to the source follower.
 
Mar 16, 2009 at 1:45 PM Post #20 of 25
Pedja Rogic's [/url]stage, and Hawksford’s Stage presented in the paper “Current steering transimpedance amplifiers for high-resolution digital to analogue converters”, (AES, 2000), which I will not post but feel free to contact me for it.
.



Can you please mail me the above paper

Thanks
harsha
 
Mar 16, 2009 at 2:04 PM Post #21 of 25
It's not a big deal to outperform the OPA627 with any no-nonsense discrete design as this op-amp is simply poor sounding. You can take a V-out DAC like the latest from Cyrrus or Sabre, build the buffer/filter using the AD797BRZ and there will be serious problems for any I-out DAC to compete this set with any kind of I/V stage. Sure, a very thorough tube or transformer design could be better but what you get with the AD797BRZ is far from the sad OPA627 SQ level. You could go even further using proper bipolar op-amps allowing for custom compensation but it's another story, not yet completely verified by me.
 
Mar 16, 2009 at 4:42 PM Post #22 of 25
Quote:

Originally Posted by harsha /img/forum/go_quote.gif
Pedja Rogic's [/url]stage, and Hawksford’s Stage presented in the paper “Current steering transimpedance amplifiers for high-resolution digital to analogue converters”, (AES, 2000), which I will not post but feel free to contact me for it.
.

Can you please mail me the above paper

Thanks
harsha



I don't know why Garbz didn't want to post it, considering it's on the university of Essex website: http://www.essex.ac.uk/DCES/research...0amplifier.pdf
 
Mar 17, 2009 at 3:05 AM Post #23 of 25
Quote:

Originally Posted by majkel /img/forum/go_quote.gif
It's not a big deal to outperform the OPA627 with any no-nonsense discrete design as this op-amp is simply poor sounding. You can take a V-out DAC like the latest from Cyrrus or Sabre, build the buffer/filter using the AD797BRZ and there will be serious problems for any I-out DAC to compete this set with any kind of I/V stage. Sure, a very thorough tube or transformer design could be better but what you get with the AD797BRZ is far from the sad OPA627 SQ level. You could go even further using proper bipolar op-amps allowing for custom compensation but it's another story, not yet completely verified by me.


My modded Rotel 855 (TDA1541a DAC, 4xOS) with an rbroer discrete output stage held its own quite nicely versus a Twisted Pear Sabre DAC with Ivy ouput stage that I had here at home for a couple of weeks (naamanf's DAC). Standard TDA1541a (no double crown, etc.). I haven't even done anything to the DAC PSUs, etc. yet.

And if it is a v-out DAC, it really isn't an I/V stage now, is it?
biggrin.gif
 
Mar 17, 2009 at 10:45 AM Post #24 of 25
Quote:

Originally Posted by Pars /img/forum/go_quote.gif
And if it is a v-out DAC, it really isn't an I/V stage now, is it?
biggrin.gif



Sure, there is none but I had to mention a V-out DAC because bipolar op-amps, especially some of them like NE5532 or AD797 have high input currents which doesn't pair well with I-out DACs. You have to use a JFET input one and they don't sound best. Actually, I don't quite like the JFET "sound" even if discrete.
 
Jun 28, 2012 at 7:50 AM Post #25 of 25
digging up a really old thread, sorry guys, I actually think its worthy of a bump anyway. Myself i'm running a modified D1 by opc called NTD1 with my ess, as well as Patrick's SEN current steering IV. i'm quite happy with both, particularly the NTD1, but I do have a weakness for well done multiloop opamp designs, even if only an intellectual exercise.
 
Garbz, did you ever get the Hawksford transimpedance multiloop IV going in a stable manner? I love the idea of this design, but anything I read about it seems to suggest extreme troubleshooting to get stability, or people like jcx giving up as they cannot even get a sim stable. i've got enough suitable opamps on hand to build it, but given it seems its prone to oscillations i'd rather not have them go all explody.
 
any help gratefully received
 
cheers
 

Users who are viewing this thread

Back
Top