• 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.

dam1941 - Next Gen Discrete R-2R Sign Magnitude 24 bit 384 Khz DAC module

Member
Joined 2009
Paid Member
Building on four years of experience with the dam1021, it's now time for the next generation OEM/DIY Discrete R-2R Sign Magnitude DAC module, the Soekris dam1941. Based on the well tested Soekris R-2R DAC core, it now include all the needed digital interface: USB, SPDIF RCA, BNC and XLR and Toslink, all fully galvanic isolated, plus an extra I2S port, suited for connection to example a streamer board like a Rasberry Pi.

It support up to 24 bit / 384 Khz PCM and DSD 256 input, with FPGA based reclocking and custom digital filters, < 0.3 ps jitter clock generator, with 27 bit resolution. Firmware can be upgraded though a standard serial interface.

The two things there are many diverting opinions about, the Power Supply and the Output Buffer, are now external to the main DAC board so it can be customized and optimized.

The board also have connections for an optional advanced user interface, with LED indicators for interface and speed, with support for an encoder to control the volume, with option for a two digit seven segment LED display, much like our end user products, no programming needed. There is even a receiver for a remote control.

There will be two basic models, the dam1941 will have four full Sign Magnitude R-2R networks for fully balanced outputs, the cost reduced dam1921 is the same board but will have only two full Sign Magnitude R-2R networks mounted for single ended outputs. The basis R-2R network has an output voltage of 1.3V RMS and output impedance of 625 ohm and can therefore drive a lot of things directly....

We have designed a couple of low cost bare PCBs to support the dam1941/dam1921, we now have a flexible Power Supply PCB, a flexible Output Buffer PCB and a full User Interface PCB, those are available as a set only.

Please visit Soekris Engineering ApS, Products dam1941 for details, specifications and pricing.

**** Now Shipping from Stock ****

This thread is to be used for dam1941/dam1921 technical discussions, lets try to keep product specific questions in the right threads
 
Last edited:
@Soren

Have some initial questions, eagerly awaiting manual.

What is the pitch of the power supply connector? (J7 ?)
J7 is the same pitch as J14?

Part number for connector ends?

The initial Soekris.dk page show power requirements are +5V 0.6A and -5V 0.1A

Is only +5/-5V power rails required? No +12/-12V ?

If one does not have the LED/Control board can one set the startup volume via serial? Permanently?

Assume it will also be -40 like the 1541?
 
Member
Joined 2009
Paid Member
@Soren

Have some initial questions, eagerly awaiting manual.

What is the pitch of the power supply connector? (J7 ?)
J7 is the same pitch as J14?

MTA156 type, 0.156" for power, with +5V Digital and +-5V analog.

J14 is the same but really for if you want to solder output connectors to the PCB with wires, the outputs are also available on other connectors.

Part number for connector ends?

Will be in manual.

The initial Soekris.dk page show power requirements are +5V 0.6A and -5V 0.1A

Is only +5/-5V power rails required? No +12/-12V ?

The dam1921/dam1941 itself only need +-5V. The Power Supply PCB will also have support for higher voltages for buffer boards.

If one does not have the LED/Control board can one set the startup volume via serial? Permanently?

Assume it will also be -40 like the 1541?

Everything will be configurable by the serial port, just like the other dams. The dam1921/dam1941 firmware will have support to be used as basic dac without leds/controls to full led/controls just like the dac1541.
 
Member
Joined 2009
Paid Member
Hi,
Some questions here:

- will it be possible to use the 1941 as 2 single ended stereo DACs (or 4 channels DAC) ?

No.

- how is the sound quality without output buffers ?

Great, like all the Soekris DACs - what else did you expect me to say :) Be aware of output impedance and level.

- how/where are the output connectors ?

On the optional output buffer board.... If not using an output buffer board there are holes on the dam1941 to attach wires....

Thank you in advance and have a nice day.
 
1941 features

Hi Soren,

I am a very happy customer of a 1541 - and now looking for a DAC board for a DIY project. As you put it: Nobody seems to agree on power supplies and buffers - I am no exception ...

Two questions on the 1941:

1) Is it possible to get ms coding from the 1941 (have the sum signal L+R on one channel and the difference signal L-R on the other one) instead of LR coding?
2) Will the cross-feed option of the 1541 be available on the 1941?

Thanks a lot!
 
Member
Joined 2009
Paid Member
Hi Soren,

I am a very happy customer of a 1541 - and now looking for a DAC board for a DIY project. As you put it: Nobody seems to agree on power supplies and buffers - I am no exception ...

Two questions on the 1941:

1) Is it possible to get ms coding from the 1941 (have the sum signal L+R on one channel and the difference signal L-R on the other one) instead of LR coding?

What's ms coding ? and whats for ?

2) Will the cross-feed option of the 1541 be available on the 1941?

Yes.

Thanks a lot!
 
Hi Soren!
I am an happy owner of the dac 1321, it is sounding great. Wil the 1941 or 1921 better the sound of the 1321? Otherwise I wil change to the diy board. Is the 1021 still a good board the work with soundwise? Or is the 1921/1941 better?
As diyer the work on the 1021 looks more as a challenge, but betters it the 1941/1921? Can you tell me?
Thanks allready!
Teake
 
Hi Soren,

ms coding (or ms encoding) stands for "mid-side-encoding"; it is frequently used in mixing.

Normally, a DAC sends the L signal down one channel and the R signal down a second channel. If you do ms encoding instead, one channel converts the mid or mono signal M=L+R and the other channel converts the side signal S=L-R. On the analog side, the two signals are mixed back together: 2L=M+S and 2R=M-S.

Two advantages:

One, you can easily change the stereo width on the analog side by amplifying the S signal differently from the M signal before mixing.

Two, more subtly, the structure of the error signal from the DAC changes. In LR encoding, both the L and the R signal will acquire a noise signal from the DAC, and these two signals are independent of one another. Therefore, the ear is presented with two additional, independent signal sources, one on the left channel and one on the right channel. With MS encoding, usually the M signal is much stronger than the S signal since most of the music is located close to the center of the stereo panorama. If we assume (extreme case) that all signal is M, then in LR encoding, left and right channel still get different error signals from the DAC. In MS encoding, left and right channel get the same error signal from the DAC. The ear is presented with one additional signal source which is located at the center spot.

Supposedly, this sounds a lot better. There is quite a bit of talk on this on a German forum, where people program their streamers to send M and S signals into the DAC instead of L and R signals, and then mix things again on the analog side. For the analog part of this, there even is a commercial product available called Audiovero Cleaner.

I think implementing such an option into an FPGA should be simple (just add one addition and one substraction), and compared to implementing it in the streamer it has the added benefit to apply to all digital sources connected to the DAC.

So, again - do you think MS encoding is possible with the 1941? I am optimistic since the 1541 already switches L and R on the DAC channels as you switch from the XLR output to the Phones output channel ...
 
Member
Joined 2009
Paid Member
Hi Soren,

ms coding (or ms encoding) stands for "mid-side-encoding"; it is frequently used in mixing.

Normally, a DAC sends the L signal down one channel and the R signal down a second channel. If you do ms encoding instead, one channel converts the mid or mono signal M=L+R and the other channel converts the side signal S=L-R. On the analog side, the two signals are mixed back together: 2L=M+S and 2R=M-S.

Two advantages:

One, you can easily change the stereo width on the analog side by amplifying the S signal differently from the M signal before mixing.

Two, more subtly, the structure of the error signal from the DAC changes. In LR encoding, both the L and the R signal will acquire a noise signal from the DAC, and these two signals are independent of one another. Therefore, the ear is presented with two additional, independent signal sources, one on the left channel and one on the right channel. With MS encoding, usually the M signal is much stronger than the S signal since most of the music is located close to the center of the stereo panorama. If we assume (extreme case) that all signal is M, then in LR encoding, left and right channel still get different error signals from the DAC. In MS encoding, left and right channel get the same error signal from the DAC. The ear is presented with one additional signal source which is located at the center spot.

Supposedly, this sounds a lot better. There is quite a bit of talk on this on a German forum, where people program their streamers to send M and S signals into the DAC instead of L and R signals, and then mix things again on the analog side. For the analog part of this, there even is a commercial product available called Audiovero Cleaner.

I think implementing such an option into an FPGA should be simple (just add one addition and one substraction), and compared to implementing it in the streamer it has the added benefit to apply to all digital sources connected to the DAC.

So, again - do you think MS encoding is possible with the 1941? I am optimistic since the 1541 already switches L and R on the DAC channels as you switch from the XLR output to the Phones output channel ...


Seems ok doing it in the mixing process, but I'm sorry, just can't see any advantage in a DAC.... If the L+R and L-R calculations are done digitally, assuming at high resolution, why would it matter where it's done ? Why would anybody send L+R and L-R to the DAC, you might as well do the calculation back to regular stereo just before sending the digital audio to the DAC ?

Not that it could not be done, addition and subtraction are simple and easy in a FPGA, but I don't want to add unnecessary functions, just more stuff to maintain and test.
 
Member
Joined 2009
Paid Member
Hi Soren!
I am an happy owner of the dac 1321, it is sounding great. Wil the 1941 or 1921 better the sound of the 1321? Otherwise I wil change to the diy board. Is the 1021 still a good board the work with soundwise? Or is the 1921/1941 better?
As diyer the work on the 1021 looks more as a challenge, but betters it the 1941/1921? Can you tell me?
Thanks allready!
Teake

The dam1921/dam1941 are based on the dac1541, so it's better, how much is up to your ears to decide. The advantage of the dam1921/dam1941 is you can adjust it to your taste....

The dam1921/dam1941 is improved over the dam1021, mostly functional, so the dam1021 is still very good.
 
for a DAU

I bought an OEM board from you a few months ago. But I had to cancel the whole project because it turned out to be too complex for me. I lost a lot of money!

Basically I am still interested in a R2R DAC. Especially a new version like your 19x1. But I don't want to spend that much money for a non result again....

As I use a single ended preamplifier the 1921 will be enough, right?
My primary source is a USB connected HTPC with FLAC and DSD files.
Your website says: "The OEM just need to add a Power Supply and Output Buffer, and if needed, an optional small PCB with the user interface."
Do I get these Power Supply, Output Buffer and the user interface PCB from your side? If so, when and at what prices?
Since I am quite inexperienced, it has been shown that I am not capable of more than "Hypex NC400" DIY. Put components together and maybe a little (!) soldering. Is a 1921 then something for me?

greetings
 
What is the recommended impedance of the next element of the chain to connect the dam directly (in balanced), without buffer ?



I want to connect the dam directly to my ncore 400 amp. I guess it's ok because the datasheet of the ncore recommend 1k ohm, and anyway it's input impedance is very high. But i am curious to know what would be the minimum impedance in which a buffer is needed, in the case if one-day i want to connect the dam to something else.



Didn't find much on the internet, something like a 1:100 ratio ?