In my build I used one transformer and Miro's psu1 for the dac board and a separate transformer and power supply to feed 5v to the DIR9001 board I'm using for input. As I understand it, ideally the draw from a dual rail supply should be even. But given the DIR9001's current requirements are so small (6ma), can I just power it from the psu1's 5v rail?
I mentioned earlier a passive IV resistor of ~27R and with 100x voltage amp using AD8429.
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7138645
Well, brave man tried and said it was much better than he expected.
Patrick
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7138645
Well, brave man tried and said it was much better than he expected.
Patrick
@propitious @NIXIE62 - Thank you. This should help clean things up a bit. As I'll eventually need to add another power supply to implement EUVL's Folded Cascode Cen IV
Looks interesting, can you share the schematic?I mentioned earlier a passive IV resistor of ~27R and with 100x voltage amp using AD8429.
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7138645
Well, brave man tried and said it was much better than he expected.
View attachment 1105781
Patrick
This is not so good approach... Current output of the DAC chip is loaded directly with Reactive load and then going to Riv.
After that is SRPP stage with local feedback (without Ck at Rk)
Value of Riv is large. And should be tested fist smaller values and appropriate instrumentation low distortion OP amp gain stage.
But maybe it could be achieved with smaller Riv because that Riv is in the same time a R load for the passive filter. Generator is the internal impedance of the DAC. That should be checked too.
...
Better way is
- first to determine right Riv value for each specific dac by measurements
- Make some slight HF filter with some C parallel at Riv to some higher F, to be slight capacitive load for the week DAC output but still nave moderate roll off...
- employ some buffer after the Riv stage. It is quite convenient to use some JFET simple SE buffer because of the +- 12V supply from the analog stage of dac. In this point is good to use passive filter of higher order following with the same buffer. Because Rg is from very high value in JFET gate as input.
- After that we have small signal but low impedance source. One option is to go with transformer for isolation + SE tube gain stage with tube buffer.
link. for passive filters
https://rf-tools.com/lc-filter/
cheers
.
That was not from me.
I just bumped into that in my computer somewhere.
What I would do with passive iv is posted here :
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7138645
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7164070
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7160371
But I don't use passive IV myself.
Patrick
I just bumped into that in my computer somewhere.
What I would do with passive iv is posted here :
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7138645
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7164070
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7160371
But I don't use passive IV myself.
Patrick
Hi PatrickThat was not from me.
I just bumped into that in my computer somewhere.
What I would do with passive iv is posted here :
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7138645
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7164070
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7160371
But I don't use passive IV myself.
Patrick
This is actually better concept.
But could be improved. For instance multiply x 4 Q1 with about 2mA Idss That will be 8mA cca. Because upper J111 is with more transconductance...
Divide R2 b 4 too.
Next You don't need J113 in the buffer stage? But J111 cascade or not... Or also multiply x 4 J113
And one the same current source instead of R4
For more gain, and without local feedback, R3 should be bypassed. But for these lower values Cs will be around 2200uF min.
Grid stoppers of 100r almost does not doing anything with some medium HF. They stoping in MHz region. Should be from higher values as 470R range.
.
J308/309-310 could be used too for the same concept.
.
I did it years ago with 2SK170 (for some TDA chips) and it was very good
Last edited:
Will that lower distortion ?
LTSpice says not.
Distortion is same, unless the load impedance is much lower than 10k.
Noise is lowered by 6dB.
You can also use 9x // to reduce by 10dB, or 16x by 12dB, etc.
Too much matching required.
And Samuel Groner once studied how mismatch will increase distortion.
Can't find it off hand.
For just-for-fun project, 2Sk170 is a waste.
Cheers,
Patrick
thanks rfbrw,
both have still a resistor for the task if I am not wrong ? It is just after an active device. It was nice to see again Rbroer thread and Jocko Homo shematic... The AD1862 was is preffered dac chip.
Hi
No it is not the same task at all.
Resistor @ DAC Iout is one element passive IV converter with phase shift voltage of -180deg
This one from Lin A. (or other similar concepts) is part of discrete active multi element circuit with dedicated power supply...
Even the resistor value could not be same. This and other active circuits showing for low values of Re more distortion and hum. Which is the opposite case for lower values of only one passive Riv at the end of dac in most cases...
I would say it is key to the operation of the circuit as a whole. Stick the circuit in spice and see what happens without it.
There are other i/v out there with a much higher component count. I'd put them up if I can find them but I'm not sure they belong in this thread.