Topic: Soren's R-2R Dac (Read 15989 times)

Chris1967 · « **Reply #110 on:** May 25, 2015, 09:43:20 AM »

Thanks uncola!!

There are 3 outputs from the Soekris:

1. Unbuffered single ended
2. Buffered single ended
3. Buffered balanced

I had originally rigged it for all 3 outputs for test purposes, that's why you see in the picture two single ended rca outputs and one blanced xlr output.

Now that i have tried them out, i prefer only the unbuffered single ended output, and you see in the picture that i have only that one connected.

The balanced xlr and one of the rca pair are not connected any longer.

MisterRogers · « **Reply #111 on:** May 25, 2015, 11:52:15 AM »

Quote from: Chris1967 on May 25, 2015, 07:07:51 AM

Nearly done now,

Salas Reflector-D on the DIYINHK receiver (@3,3V) and two Salas Bib on the Soekris (@12V)

Nice work Chris!

Chris1967 · « **Reply #112 on:** May 25, 2015, 06:17:40 PM »

Thanks man!!

fishski13 · « **Reply #113 on:** May 26, 2015, 02:13:27 AM »

yes, awesome sauce.

aive · « **Reply #114 on:** May 26, 2015, 09:10:30 PM »

Quote from: hans030390 on February 17, 2015, 10:19:58 PM

NOS filter slightly fatiguing over time but detailed? That's not what I would expect.

You should check it out Hans. The NOS square wave waveforms posted on DIYA are impressive - the DAC acts like a signal generator.

Analogue LPF at ~250 kHz. If you input 352/384 kHz it bypasses the first FIR filter. The second 8x filter can't be avoided, but you can set it to NOS mode only, which only 8x zero pads in time domain but doesn't interpolate - this upsampling shouldn't produce any aliases in audible domain and aliases should be removed via 250 kHz analogue filter.

If I get one, I'd do my upsampling in PC (going to try pre-upsample my files to 352/384 kHz via "exact interpolation" process (process I think Yggy filter is based on)) and just run Soren's DAC in NOS.

Edit: Yeh I don't recommend running NOS at 44 khz... aliases up the wazoo...

bettersimple · « **Reply #115 on:** June 08, 2015, 11:36:52 AM »

Hello Priidik,
as I can see we have a similar setup:

Quote from: Priidik on May 04, 2015, 02:46:14 PM

I use a simple regulated LPS for power and JLsounds USB interface.

My DAM1021 is connected to the JLSOUNDS I2SoverUSB. I use a 5V linear supply for the galvanic isolated side of I2SoverUSB and a 3.3V ultralow noise TPS7A4700 regulator for DAM isolator chips.
I configured DAM input as automatic (INPSLCT0 and INPSLCT01 open)
I use Foobar player.

I've got a very annoying issue: every track change suffers from a 3 sec pause/fade in.
By default the issue reported concerns sample rate change but I have it when every track of the same album starts.
So I loose the first 3 secs of every track: I can see DAM signal lock led starting to blink (no signal or no lock) for 1 sec every new track of the same album, then the music starts with the 3 secs fade in.

Do you have a similar issue? Have I to configure DAM input as manual to I2S ((INPSLCT0 and INPSLCT01 to GND)?

Tnx

Nevod · « **Reply #116 on:** June 09, 2015, 07:28:09 PM »

That dac could benefit from factory-made nonlinearity measurement and corresponding compensation, then it won't have any disadvantages compared to laser-trimmed R2R chips. Though that could require a more powerful FPGA.

Priidik · « **Reply #117 on:** July 20, 2015, 09:25:40 AM »

Hi gents!

Anyone tried to mod the +/-4V ref yet?
(It had quite a bit of signal related pollution peaking at 12kHz, person to thank for this discovery is Alecm from DIYA)
If not, i highly recommend so. In my setup previous mods (clean supplies for osc-s and dam itself) have made 3-5% improvement. Filters perhaps 10-15%, if i must quantify.
The 4V ref mod is more like 30% improvement. Significant step up in soundstage size and precision, better dynamics contrasts. Bass is slightly better defined and has more impact. Much more space between sounds now. Going back, the stock sound is kinda raw and grainy now vs the upgraded 4V ref. Yes, i went back a couple of times i messed up.
The term used here gray vs black background is also valid here i think. Background is emptier, darker.

I replaced the smd ceramics on 4x supply bypass to 47uF ones + added a good quality Wurth Elektronik 10mOhm, 630uF polymer on each shift register's Vcc -- GND. (overkill, i know)
I have not touched the supplies opamp gain yet (i'm too lasy to track down those smd resistors). Reportedly the 10ohm resistor in the feedback loop is the main culprit there.

MisterRogers · « **Reply #118 on:** July 20, 2015, 11:56:37 AM »

Awesome work mate! I flipped mine - needed the cash for 'inbound' gear (Yggy, HE1K). I'll be building another one soon - we'll, soon being after I have funds saved for my EC Studio :-) Not surprised to hear about the results of the 4V mods.

Feste · « **Reply #119 on:** August 09, 2015, 07:24:26 PM »

Hey guys -

Was wondering if anyone here had tried a different output stage they liked better than the unbuffered SE?

Author Topic: Soren's R-2R Dac (Read 15989 times)

Chris1967

Re: Soren's R-2R Dac

MisterRogers

Re: Soren's R-2R Dac

Chris1967

Re: Soren's R-2R Dac

fishski13

Re: Soren's R-2R Dac

aive

Re: Soren's R-2R Dac

bettersimple

Re: Soren's R-2R Dac

Nevod

Re: Soren's R-2R Dac

Priidik

Re: Soren's R-2R Dac

MisterRogers

Re: Soren's R-2R Dac

Feste

Re: Soren's R-2R Dac

(adsbygoogle = window.adsbygoogle || []).push({});

Author Topic: Soren's R-2R Dac (Read 15989 times)