Soekris dam1021 vs dam1121

[transcendent]

Hi guys,

I haven't found proper explanation in soekris.dk site nor in this forum what is the difference between the two. Also if anyone had a chance to listen to both DACs - any feedback would be appreciated.

Thanks !

Simonas

 

[analog_sa]

The obvious advantages of 1121 in 2-channel mode:

- integrated inputs, including usb
- simplified controls for input/filters which do not require a micro
- reclocking after FPGA

A possible disadvantage is the inability to directly drive headphones. It also requires an external source of dc.


Judging by the 1941 i would expect it to sound considerably better than a 1021 under similar conditions.

 

[analog_sa]

Please disregard the above Not sure why i was thinking the 1121 is a non-balanced version of the 1941. It is not.

 

[wcwc]

Couple of differences.

1021
Wider range of Powerability Direct AC 7 or 8V or DC from +/- 7 to 15V.
Buffered outputs via an opamp or direct ladder out
Isolated and non-isolated serial connections
larger board
only operates asynchronously
I2S, 2 SPDIF inputs

1121
+/- 5V DC power input
direct ladder out
can be operated asynchronously or synchronously with an external clock (only using an Si570)
can operate multiple boards synchronously
I2S, one SPDIF input
Smaller board

 

[transcendent]

Awesome, thanks for quick replies! I don't really care about AC part and buffered output, and might want to have multiple boards working in sync in future so dam1121 seems like obvious choice.

 

[euro21]

1121:
- hard to access -and use- tiny connectors;
- cost is relatively high (compare to 1021).

My 1121 is in the box, I bought 1941 (as "early bird").

 

[analog_sa]

Am i wrong thinking the 1121 is reclocked after the FPGA? This alone may be sufficient for better sound.

 

[transcendent]

1121:
- hard to access -and use- tiny connectors;
- cost is relatively high (compare to 1021).

My 1121 is in the box, I bought 1941 (as "early bird").

1121 is actually cheaper and comes with more precise resistor values as I see it..

 

[bogdandascalu87]

Am i wrong thinking the 1121 is reclocked after the FPGA? This alone may be sufficient for better sound.

Clocking scheme is as follows

Si5xx===LVDS====> u2——ttl——>U8,U11——>stcp pin12 of shift registers
| ^
| Fpga
| ^
|___ttl______|

U2 lvds to ttl converter
U11,u8 flip flops



Also, dam1021 is using 7/8x4 shift register outputs,so a sign magnitude segmented dac with 1msb, 2 bit thermometer, 25 bit r2r (as per soekris)
Dam1121 is using 8/8x4 shift register outputs. So 7 msb are thermometer like, 25 bit r2r?

AD3-F3706-5863-4-F28-9-D4-F-BF5470-A35-E1-C — imgbb.com

 

[TNT]

Am i wrong thinking the 1121 is reclocked after the FPGA? This alone may be sufficient for better sound.

No, this is correct. Big blunder not to in the 1021...

I own both borads and to me, there was no doubt about the advantage for the 1121. Change/take away the output cap and you are good to go.

//

 

[TNT]

Clocking scheme is as follows

Si5xx===LVDS====> u2——ttl——>U8,U11——>stcp pin12 of shift registers
| ^
| Fpga
| ^
|___ttl______|

U2 lvds to ttl converter
U11,u8 flip flops



Also, dam1021 is using 7/8x4 shift register outputs,so a sign magnitude segmented dac with 1msb, 2 bit thermometer, 25 bit r2r (as per soekris)
Dam1121 is using 8/8x4 shift register outputs. So 7 msb are thermometer like, 25 bit r2r?

AD3-F3706-5863-4-F28-9-D4-F-BF5470-A35-E1-C — imgbb.com

So both are 25b r2r? What is the net difference really?

One would hope U11 and U8 are from Potato!?

//

 

[bogdandascalu87]

So both are 25b r2r? What is the net difference really?

One would hope U11 and U8 are from Potato!?

//

Take a look at figure 3a in this article, where a segmented dac is used as example https://www.analog.com/media/en/training-seminars/tutorials/MT-016.pdf


The thing that confuses me is how Soren is using the 7 resistor array of the thermometer segment of dam1121. As 3bit from 32(8x4 shift register outputs) or as 7bit from 32?

In either way using thermometer dac for msb it’s really usefull. Thermometer dac is inherently monotonic and not so sensible at resistor tolerance, so having say 3 bits from 16 resolved by thermometer, the rest of 13lsb are resolved by r2r segment, and for 0.01% tolerance resistor this is a doable task...

Could any help?

 

[dewdrop]

No, this is correct. Big blunder not to in the 1021...



I own both borads and to me, there was no doubt about the advantage for the 1121. Change/take away the output cap and you are good to go.



//

Taking your advice seriously, the 2 output caps are going away soon!

 

[TNT]

Sören changed the type caps on later products. If it was due to "market pressure" or own revelation - we will never know

As an EE, I think there should be a cap there - but a sonically transparent type.

//