Soekris' DAC implementations

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
Q

quantran

Member
Joined 2004
Hi

My dam1121 has a problem with locking to signal from coaxial input. I built the input circuit with a pulse transformer as recommended by Soren including 1k resitor connected to 1.2V and 0.1uF to ground.

The DAC is playing but it loses locking every few minutes. I am not sure if there is any problem with impedence matching due to my bread board implementation. Should I arrange the pulse transformer close to the the dac board or close to RCA socket?

Anyone has an idea about the potential issue? I have attached a photo.
 

Attachments

  • IMG_1170.JPG
    IMG_1170.JPG
    75.2 KB · Views: 921
F

fedde

Member
Joined 2002
fedde.nu
I currently use a USB to RS232 convertor to communicate with my Soekris DAC. I would like to change this into TTL level so that it become isolated.

I consider the following option:
PL2303TA USB TTL to RS232 Converter Serial Cable module for win 8 XP VISTA 7 8.1 | eBay

But this device only has a 5V power option (and 3.3V TX/RX levels). Do I need to put a 3.3V voltage on the isolators to make this work? So do I need to look for a cable with 3.3V power supply? Any suggestions?

Fedde
 
F

fedde

Member
Joined 2002
fedde.nu
I have another question. Does anyone have good advice on an USB to I2S convertor for DAM1021?

My shortlist contains currently:
1. Amanero combo384
2. Singxer F-1
3. DIYINKH XMOS 384 (or 784 or multichannel 384)

I would appreciate low jitter and (especially) stable drivers. Possibility of usage of USB power is a plus (this applies to the first two options). I heard some good stories about the F-1. but is it really worth the money? It is almost twice as expensive as the Amanero!

Any feedback is greatly appreciated... ;)

Fedde
 
J

JOHNNYCR100

Member
Joined 2010
Newbee question, Kali or not?

Hi sirs, absolute newbee question ...but better to ask prior to spending money;)

I wanted to mate a raspberry pi3 to a sokoris dac, but it known that the pi clock is not great for audio, so the Kali reclocker is very good ( so I read, I don't own one) ...however as the sokoris dac has its own master clock ( and won't run in slave mode?).. Is the Kali any use with the sokoris!

Given that is the pi suitable at all for a sokoris dac or serious audio at all ?.
Thanks for bareing with me, I have not found a definitive answer yet, though no doubt it exists , I may have missed it,

Thank you all answered valued.

Regards
Johnny
 
TNT

TNT

Member
Joined 2003
Paid Member
If the re-clocking in DAM works as the designer describes it - the clock doesn't matter. It cant be improved and also, it cant be worsened. Well, if its wildly varying so that the DAM clocks need to be adjusted every 30 second this is not good - this however requires such a poor clock feeding the DAM so that it can be judged as broken. It just need to be stable (the ppm figure) - jitter (the fs/ps figure) has no influence as it will be cleaned out as the music samples will sit and wait in a memory for a while before used.

Anyone putting a bord with a processor e.g. (RP) or a SMSP close to a DAM must be MAD as it will shower the DAM with HF garbage - both via the connections and airborne. If one really want the most correct reproduction - any switching unit should be kept of arms length of the DAM board. Now wires might be OK as the DAM is isolated but this isolation seem to have some limits - radiated HF however is quite hard to get rid of.

But on the other hand - if you like the little edgy sound which could be taken for resolution - go ahead and use a "computer environment".

In my experience, if you want to squeeze the most fidelity out of a DAM board, these should be your priorities:

- DC power quality. (Non switching regulators close to DAM)
- Lowest possible received EMI. (Use toslink. Anything switching - 50cm away at least)
- Vibration isolation. (Be inventive - experiment)


//
 
J

JOHNNYCR100

Member
Joined 2010
thanks TNT

Hi TNT,

well thats food for thought,- sounds like a kali may not be necessary at all in a Sokoris implementation?---and a PI possibly if not isolated, thats food for thought construction wise...wonder how all these dac hats measure up in terms or RF Hash?....still what is saved on the kali i could buy a tin box and some cork:p

thank you for that intresting reply.. may have saved some cash there?

thank you for that answer.
best regards
johnny
 
H

hop.sing

Member
Joined 2009
Interesting

I have never thought that a raspberry close to the dam could be a problem. Since there is a Microprocessor on the dam itself. I understand no cheap switching power close by, but that is news to me. I am going to test that since I have a toslink source and a raspberry source with a kali hardwired within 15 cm distance to the dam. I will disconnect power to the raspberry and listen with toslink only. Maybe I will hear a difference, hopefully not😀
Ps: the kali made a difference to my ears, but as everybody knows, take that with a grain of salt... conformation bias and problems like that ...
 
F

fedde

Member
Joined 2002
fedde.nu
TNT said:
If the re-clocking in DAM works as the designer describes it - the clock doesn't matter. It cant be improved and also, it cant be worsened. Well, if its wildly varying so that the DAM clocks need to be adjusted every 30 second this is not good - this however requires such a poor clock feeding the DAM so that it can be judged as broken. It just need to be stable (the ppm figure) - jitter (the fs/ps figure) has no influence as it will be cleaned out as the music samples will sit and wait in a memory for a while before used.

Anyone putting a bord with a processor e.g. (RP) or a SMSP close to a DAM must be MAD as it will shower the DAM with HF garbage - both via the connections and airborne. If one really want the most correct reproduction - any switching unit should be kept of arms length of the DAM board. Now wires might be OK as the DAM is isolated but this isolation seem to have some limits - radiated HF however is quite hard to get rid of.

But on the other hand - if you like the little edgy sound which could be taken for resolution - go ahead and use a "computer environment".

In my experience, if you want to squeeze the most fidelity out of a DAM board, these should be your priorities:

- DC power quality. (Non switching regulators close to DAM)
- Lowest possible received EMI. (Use toslink. Anything switching - 50cm away at least)
- Vibration isolation. (Be inventive - experiment)


//
Click to expand...

According to the designer, the power supply should not be critical as well due to local regulators. Though I must admit I saw a great improvement from going frrom LM78xx to LM317. Later on, I may try modern low noise regulators to see if this still brings an improvement.

Regarding EMI from a RPI, this could be largely shielded by some metal plates. Furthermore, the I2S lines are indeed shielded so this should not cause to much noise either. Though I will keep my PC as source for convenience reasons.

I am not a big fan of optical sp/dif. Though this may work much better on DAM due to reclocking. In my situation I want to get rid of my RME soundcard. The ASIO drivers are rock solid, but kernel streaming does not work so good with RME drivers. And it turns out that kernel streaming sounds significantly better. With USB I could have proper kernel streaming support and also do experiments with 192 kHz and 384 kHz.

Do you have any suggestions for improvements in vibration isolation?

Fedde
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom