A NOS 192/24 DAC with the PCM1794 (and WaveIO USB input)

Jump to Latest
#7,341
smooth dancer said:
I think it is important to separate 5v from waveio and 5v out from mainboard. If not, waveio will be on the same side of the fence as fifo’s isolated side. I only connect the i2s lines +gnd from J7 - i2s mainboard. J7 have dedicated 5v psu.

Doede, can you confirm if my thoughts about 5v between waveio and 5v on mainboard are correct ? As I see it, only i2s lines + gnd, from J7, shall be connected to mainboard, and fifo powered by a clean psu (J-5 header).
Click to expand...

we are aligned that waveio and Mainboard should be separated by the IAN...
Now the 5 Volt for the FIFO (J5) can be separate, or coming from the LF50 (10 pin header) from the mainboard (must be cooled than)
 
#7,347
A choke block higher frequencies, and makes life easier for the other components to fight ripple in a psu.
A choke need a bleeder resistor (after the choke) for optimal functionality.
I put chokes in my psu because I want to maximise where I can, and since it is diy it’s not expencive and just pure fun.

Doedes stock psu is very good too. I have’t been doing a comparason, but a friend of mine use stock Doede psu’s on his dddac, and it sounds very good.

Doede’s stock psu deliver enough current if you don’t stack more than 4 boards.
 
Last edited:
#7,349
What value for a choke ? It depends on rest of the psu design. Smaller chokes cost less and is easier to implementing in a given cabinet size. I use the Lundahl 1694. You can order directly from Lundahl, or maybe there is a dealer in the UK ?

I know eduard is a fan of big chokes, but how big is needed in a dac ?
 
#7,351
Sorry, no diagram. It’s been a long time since I put it together, but do as I did, use Duncan’s psu designer, and you will end up with a good psu in the end.

From my memory; I use R-core trafo’s (300va) Toroidal is nice too, and easier to buy. The trafo’s don’t need to be so big. (I have my R- core already in house).
Sjøstrøm rectifier bridges, Mundorf M-Lytic caps ( 20000uf-10000uf), and wirewound (bleeder)resistors.

When you play with Duncan’s psu designer you must try to end up with around 20v before dddac’s 12v psu, and 12-15v on dddac’s 5v psu. 7-10v for regulation is ok without the regulator on dddac psu get’s to hot.
 
#7,353
Ian FIFO problem - and.... solved

Need to share this right now... :wrench: :wrench: :wrench:

The Ian was NOT working till for 10 minutes ago. I was very frustrated yesterday evening as the FIFO status LED stayed red (empty buffer). I had to sleep a night over this - ask a few fellow DIY and even thought the board could be defective (already ordered a new one 😀 this one will stay for a RoPIeee server next to build 😛 )

today I had the smal lightbulb flash and tried with a lower voltage on J3, the connector for the power supply input for the not-isolated INPUT side of things and where all the controlling is happening. My idea was that the PI works with 3,3 V logic and maybe this was the issue, that the 5 Volt I am taking from the Waveio was too much and the 5V logic I2S signals were not working correctly. Well indeed...……

NOW it all works perfectly well between roughly 3 and 4 volt on J3 (J5 stays 5 volt from the mainboard at the isolated side to be 100% clear)

Now to avoid an extra 3,3 Volt power supply just for the controlling side of the Ian FIFO I still take the 5 Volt from the Waveio but with a 18 Ohm / 1 Watt resistor in series to arrive at 3,34 Volt at J3 (The current draw is pretty constant between 85mA and 90mA depending on FS 44.1 - 192kHz)

As this makes the earlier "recommendations" basically wrong, I wanted to post this ASAP... Listening tests I will do tomorrow 😎

FYI:

J3 - 3,34 Volt draws 87mA in average - no problem for the 5Volt supply for the Waveio

J5 - draws (only) 24 mA. This comes from the LF50 on the DDDAC Mainboard and hence does NOT require any extra cooling (I am trying right now as I type this - stays below hand warm)

sorry for wrong Infos based on "logical" thinking, which turned out to be not so logical 😱

For tomorrow….. Happy new year evening to all of you and make sure after the fireworks you still have your fingers to do all this nice DIY 😀
 
#7,354
Ian FiFo test rig dddac

here is a pic from my Ian FiFo test rig
 

Attachments

  • Ian FiFo test rig dddac.JPG
    Ian FiFo test rig dddac.JPG
    783.5 KB · Views: 328
#7,355
Does the IAN FiFo introduce a delay in the sound output, similar as the Kali? The data buffering in the Kali resulted in a small delay of maybe 0.5 seconds (or less?). This is completely irrelevant for normal music listening, but if you also want to use it for movies, the delay is quite annoying. It would be interesting to know if the IAN board does the same, and how much the delay is.
 
#7,356
dddac said:
Need to share this right now... :wrench: :wrench: :wrench:

The Ian was NOT working till for 10 minutes ago. I was very frustrated yesterday evening as the FIFO status LED stayed red (empty buffer). I had to sleep a night over this - ask a few fellow DIY and even thought the board could be defective (already ordered a new one 😀 this one will stay for a RoPIeee server next to build 😛 )

today I had the smal lightbulb flash and tried with a lower voltage on J3, the connector for the power supply input for the not-isolated INPUT side of things and where all the controlling is happening. My idea was that the PI works with 3,3 V logic and maybe this was the issue, that the 5 Volt I am taking from the Waveio was too much and the 5V logic I2S signals were not working correctly. Well indeed...……

NOW it all works perfectly well between roughly 3 and 4 volt on J3 (J5 stays 5 volt from the mainboard at the isolated side to be 100% clear)

Now to avoid an extra 3,3 Volt power supply just for the controlling side of the Ian FIFO I still take the 5 Volt from the Waveio but with a 18 Ohm / 1 Watt resistor in series to arrive at 3,34 Volt at J3 (The current draw is pretty constant between 85mA and 90mA depending on FS 44.1 - 192kHz)

As this makes the earlier "recommendations" basically wrong, I wanted to post this ASAP... Listening tests I will do tomorrow 😎

FYI:

J3 - 3,34 Volt draws 87mA in average - no problem for the 5Volt supply for the Waveio

J5 - draws (only) 24 mA. This comes from the LF50 on the DDDAC Mainboard and hence does NOT require any extra cooling (I am trying right now as I type this - stays below hand warm)

sorry for wrong Infos based on "logical" thinking, which turned out to be not so logical 😱

For tomorrow….. Happy new year evening to all of you and make sure after the fireworks you still have your fingers to do all this nice DIY 😀
Click to expand...

This is very strange. be interesting to read if the others that have built the IAN have the same issue Doede.

Happy New Year too! 😀 :cheers:
 
#7,358
Me again 🙂

Ref the PIN numbers on the 40pin connectors for the IAN. I thought the numbers were as in my diagram but looking at peoples pictures I see a wire attatched to PIN 40 as expected but then counting back does not align, what am I missing...if anything and its just me?

Dont want to make mistakes

Thanks
 

Attachments

  • IMG_6324.jpg
    IMG_6324.jpg
    538.6 KB · Views: 279
#7,360
mbrennwa said:
Does the IAN FiFo introduce a delay in the sound output, similar as the Kali? The data buffering in the Kali resulted in a small delay of maybe 0.5 seconds (or less?). This is completely irrelevant for normal music listening, but if you also want to use it for movies, the delay is quite annoying. It would be interesting to know if the IAN board does the same, and how much the delay is.
Click to expand...
His previous version did and any buffered solution will. It was around the same as Kali ie 0.5s and annoying with dialogue
 
You must log in or register to reply here.
Top Bottom