• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Reference DAC Module - Discrete R-2R Sign Magnitude 24 bit 384 KHz

My problems is how to improve sound quality .About alsa driver, all is in kernel docs,it's simple & clear; and I get 384KHz i2s as bcm2835 datasheet.There is no problem in I2S output signal for now.I can't find anything else left for ALSA-dev to do,

After skimming through your code, the following comes to my mind:

1. pcm formats: SNDRV_PCM_FMTBIT_S32_LE - IIRC dam1021 won't process 32-bit samples, although it uses 32-bit word alignment; 24-bit is the maximum allowed.
2. code duplication when rates are defined btw of SNDRV_PCM_RATE_KNOT [1]
3. this patch will probably break some configurations that use linear rate plugin. and apps that use simple PCM API cannot utilize higher sample rates anyway (see alsa-lib, pcm_simple.c@157).

It does not mean that your code doesn't work as intended, but there might be subtle corner cases that it fails. To be honest I'm not kernel developer. Probably neither are you. That's why I suggested patch submission to alsa-devel for a fair peer review.

the next will be how to figure out the problem of sound listening,I thinks so.

IMO pushing higher sample rates to dam1021 won't improve anything in audible terms (ceteris paribus). Try vref mods if you are keen on (de)soldering.

[1] [alsa-devel] [PATCH] ASoC: cs42l73: Don't mix SNDRV_PCM_RATE_KNOT with specific rates
 
After skimming through your code, the following comes to my mind:
........
[/url]
Thanks for your reply!

1. pcm formats: SNDRV_PCM_FMTBIT_S32_LE - IIRC dam1021 won't process 32-bit samples, although it uses 32-bit word alignment; 24-bit is the maximum allowed.
---This field represent the format of hardware supported.dam1021 support 32bit i2s signal, and in my memory, it's sensitivity is about 28bit in someway.

2. code duplication when rates are defined btw of SNDRV_PCM_RATE_KNOT [1]
3. this patch will probably break some configurations that use linear rate plugin. and apps that use simple PCM API cannot utilize higher sample rates anyway (see alsa-lib, pcm_simple.c@157).
---Yes,you're right. it maybe cause some app which use simple PCM API work abnormal, and confuse others to reading code.But when work with MPD, there's no differences between my ugly code & correct code,They result to same hardware state both.And I do not want to do patch submission to alsa-devel, because as you said before:I am not a kernel developer.

My problems is :
when same music play out from amanero & raspi, there is something differences between them.
At first,I found that mpd or alsa driver do resampling when play dsd file or 352.8KHz file on raspi board, it will resample to 192KHz, it's unacceptable for me,so I add 352.8/384K playback support to driver.Now there is no resample left, but it's still something different.
My next problem is to find out something else result to this?

Raspi board did output i2s signal as datasheet, is there something left to do in alsa driver code?
I can't see anything left for hardware working process.So I'm confused why you keep asking me to
patch submission to alsa-devel for a fair peer review. -_-!
Sorry, my files is not a official alsa patch, it's just written for figure out the cause of above problem.

Maybe I'll do some test to compare output audio wav in some days.
By now, I have fallen into ""Fallout 4"" on ps4 ^_^
 
How come the ladder is fed +/- 4v but the raw output is still single end. How does that work?

//

+/-4v is required to make an output with no DC offset. It really has nothing to do with single ended versus balanced.

To make a true balanced DAC with this, you need two boards, so you have 4 channels. Left +, Left -, right + and right -.

And for that, you still need +/-4v. The left + and left - both go positive and negative, but the are 180 degrees out of phase from each other.

Randy
 
Thanks for your reply!

1. pcm formats: SNDRV_PCM_FMTBIT_S32_LE - IIRC dam1021 won't process 32-bit samples, although it uses 32-bit word alignment; 24-bit is the maximum allowed.
---This field represent the format of hardware supported.dam1021 support 32bit i2s signal, and in my memory, it's sensitivity is about 28bit in someway.

Those additional bits form headroom for DSP.

I can't see anything left for hardware working process.So I'm confused why you keep asking me to
patch submission to alsa-devel for a fair peer review. -_-!

Well, that's an overstatement. It was just kind advice. At the end of the day it's you, not me, that complain about sound quality, and it seems you have a hard time figuring out what's going on.
 
Last edited:
DSD support

All ready, except long awaited DSD FW

Same here, DSD FW...... common Sóren it is November and DAC1101 is nowhere to buy?
Cheers
Danny
 

Attachments

  • kozel-rmaf-2015-image44.jpg
    kozel-rmaf-2015-image44.jpg
    403.5 KB · Views: 894
first: please excuse my poor english...

Is there anybody who can tell me about the sound quality of this DAC compared to a Buffalo, Mambo or any commercial DAC?

I am very interested in the Soekris, and there are lots of technical descriptions and documents about it. But there is no description about that what i need to make a decision: What comes out of this device?

Regards,
Georg

I already ask the same question month ago , but like you , no anwers :confused:

it seems that this device is for electronics fan not for music :(
 
I already ask the same question month ago , but like you , no anwers :confused:

it seems that this device is for electronics fan not for music :(

First off, my heart goes out to the folks in Paris...

Now, the rest of the stuff may seem rude, but it's really just laying it out factually.

Most folks don't bother to respond to people who did not use the search function in the forum. There have been comparisons already made in this thread. A few (who've stopped contributing recently) don't think it compares with the 5k euro TotalDAC. There have been many that think it compares very favorably against Schitt's totl DAC. And betters a few others.

The problem is that this is a DAC module... the results are highly dependent on your implementation (there isn't a single set way to set this up... this is DIY). People have their preferences in music and it is reflected in how they've decided to set up the DAM1021.

The Vref mod alone has a multitude of variations.

The supplies (some batteries... some just a linear +-12v... some like me who've removed the onboard +-5v supplies and powered the whole DAC by my own +-5v supply with another dedicated supply for the 3.3v.)

Then the output stage can be different... some use the onboard balanced drivers... some single-ended straight to amp... some to a buffer... and some to a preamp module.

So what do you want to compare it to - not everyone has the same setup... it's not the DAM1021 vs X... it's DAM1021XX vs DAM1021XY vs X and that's not going to happen... are you going to sponsor one of the DIYers to complete the various setups?

So you have everyone with different setups and different preferences to music and you want a list of which is better than what? Based on the info above, tell me who is able to answer you on that.

I'd say your last line is the most inaccurate of all. I think most DIYers enjoy the tinkering for improvements to suit their taste... they don't necessarily care which retail product is better than the other. They relish in the pleasure gained from improvements from what they are able to accomplish. So the folks who've bought and are enjoying this DAC fit in the category that they are music lovers and not gear lovers.... they love tinkering with the items to get to where they want. It's also not cheaper... I think the money spent on the parts for the mods are close to eclipsing one of the DAM modules on my part... and I'm still trying to get around to tinkering with Arduino to control both my boards... but because it's sounding great to me I haven't moved from that makeshift setup for months now.
 
Last edited:
Nothing rude in your answer , just facts , and they will help me and some others to make our choice , just because we are in the DIY world , the more infos we get the better is ;-)

we all got the point that every setup have different resuts , thats what you can share , if you like ;-)



.
 
Last edited:
Hi I have connected Soekris v2 by raw output to my tube amp and i have problems with wird noise,When i put volume on 12 i dont have that noise, when i put volume lower or higher the nois i coming on, i have heard that problem is with impedance , and i must add bufor to soekris, good and transparent bufor is on 2sk170 j-fets, but for know i dont have power suply for them so i have question that my problem is on impedance ?
 
DAM1021 V2 Pics

Hi,

I got my V2 couple of weeks ago and connected it to a DIYINHK XMOS (non isolated), Salas BIB +/- 12V, 3.3V Salas reflector-D and RPI2 with Runeaudio, DAMfilters, output to LDR>DCB1>Pass F6. Very pleased with the music produced by the set. Still waiting for he Nordmundss V2 companion boards to finish it.

Albert

The requested pics:

DAM1021V2fr (Large).jpg

DAM1021V2bk (Large).jpg
 
Hi I have connected Soekris v2 by raw output to my tube amp and i have problems with wird noise,When i put volume on 12 i dont have that noise, when i put volume lower or higher the nois i coming on, i have heard that problem is with impedance , and i must add bufor to soekris, good and transparent bufor is on 2sk170 j-fets, but for know i dont have power suply for them so i have question that my problem is on impedance ?

I had the same problem.

Fixed it bij putting a 20K pot between the 3,3v line of jour volume pot.

Set pot at 2,9v at max.of your volume pot (12).
 

Attachments

  • res.jpg
    res.jpg
    122.6 KB · Views: 760
  • DAM dac.jpg
    DAM dac.jpg
    965.7 KB · Views: 753
Last edited: