Just for debugging my crackle issue and out of curiosity I quickly hooked up my RPi and HifiBerry DAC+ to the DDDAC using the IS2 out of the HifiBerry. As far as I can tell, this is exactly the same hardware setup as used in Dicks article (see post 6022), and as sold as the "DDDAC 1794 NOS Berry Streamer Kit". I am assuming this should work ok with the DDDAC. Result: massive crackle!
I really think the problem is related to my DDDAC board, whereby the WaveIO-I2S output seems to able to drive the IS2-IN on the DDDAC, whereas the Kali and the HifiBerry are not. I wouldn't be surprised if this is related to my earlier 1/2 clock delay modification.
Before I buy a brand new board, maybe someone has a (blue) board in good condition to spare? I'd be happy to buy a used one in good condition!
I really think the problem is related to my DDDAC board, whereby the WaveIO-I2S output seems to able to drive the IS2-IN on the DDDAC, whereas the Kali and the HifiBerry are not. I wouldn't be surprised if this is related to my earlier 1/2 clock delay modification.
Before I buy a brand new board, maybe someone has a (blue) board in good condition to spare? I'd be happy to buy a used one in good condition!
Pictures of the Kali sitting in between the WaveIO and the DDDAC or pictures of the 1/2 clock delay mod?
Do you suspect that it could be a case of some sort of interference that your dac/red board is picking up from the external i2s signal sources, whether it is the digi+ or kali? Is it possible to try some isolation measures? Dont know maybe some shielding? Just thinking out loud.
Also, since now you are using an rpi, try the moode software.
Im sure you would have been careful, but just recheck the pin arrangement (pin out from digi+/kali to pin in of red mobo).
I have thought on these lines looking at the issues other diyers are facing with rpi based setups. An alternative is something like an Auralic Aries Mini with its own linear power supply.
Do you think it's a worthy consideration? It has an inbuilt dac which will be redundant here but otherwise as a streamer, will it work well with THE DDDAC? I can use the usb out from the aries mini to feed the DDDAC.
One more question, if i use the aries via usb out to DDDAC and make it also detect say my secondary amp having built in Bluetooth, can i use the aries to stream diff content to these diff devices at the same time?
Eager to hear impressions on the aries by fellows who have used it.
Likely both, the start with the Kali and its input/output connections.
Greg in Mississippi
Hi mbrennwa,
I'm not sure, if this helps, but I had also problems with a instable playback. Not sure if I would call it crackle, but Kali was not going into stable lock status, so many many interruptions, no stable playback. I had this issue when kali was connected to a IL715 isolator chip (my setup was ODROID C2 -> IL715 -> Kali -> DDDAC). The same chip is installed on WAVE IO Board. I could solve my problems by installing resistors (I used 100ohm) into the single I2S lines from IL715 to KALI.
I think it is worth a try to add 100ohm resistors on your setup (at the single cables between WaveIO and KALI).
Hope this helps...
Michael
I'm not sure, if this helps, but I had also problems with a instable playback. Not sure if I would call it crackle, but Kali was not going into stable lock status, so many many interruptions, no stable playback. I had this issue when kali was connected to a IL715 isolator chip (my setup was ODROID C2 -> IL715 -> Kali -> DDDAC). The same chip is installed on WAVE IO Board. I could solve my problems by installing resistors (I used 100ohm) into the single I2S lines from IL715 to KALI.
I think it is worth a try to add 100ohm resistors on your setup (at the single cables between WaveIO and KALI).
Hope this helps...
Michael
Thanks, that's useful information! Just found your posts in the Kali thread...
How did you find out about the 100Ω resistors? How do they work?
Did you use the DDDAC 5V supply to power the Kali?
the series resistors in the i2s lines were just a try and I was happy to solve my problems with that try. You can find resistors like this in many applications. Even in front of the 1794 chips on the DDDAC boards these resistors are installed. More information you may find here:
http://www.diyaudio.com/forums/digital-line-level/212002-i2s-digital-signal-path-line-resistors.html
I think 33ohm would be better, but I didn't had one on my hand, so I tried with 100ohm and it was running.
regarding the PSU. I use choke psu lclc setup with lundhal 1694 for feeding my DDDAC Boards with Tent shunts. I also feed the kali with it. But between the 13V from choke psu and Kali, I have installed a cascade of LDOs (one UA7808 and after that one UA7805). This sounds best in my case. Kali really needs a good power supply to sound best. The 7805 without the 7808 in front sounds worse than with both regulators...
Not giving up on the DIY; I ordered the RPI/Hifiberry Digi+ kit from Audio Creative. Very curious to see what that can do for me ... 🙂
This thread also mentions that I2S wire connections should be made with signal-GND-signal-GND-signal-GND flat cables. I was a bit lazy in this regard, but I tried to keep the wires as short as possible. Anyway, how did you go about this with your O2 / Kali setup? The Kali connector (40 pin) does not have signal/GND pin pairs next to each other (the WaveIO and the DDDAC mainboard do).
I guess I'll have to find better cables and connectors plus a few resistors... in the meantime I'll enjoy the WaveIO directly into the DDDAC. Will report back once I get this done (with photos if necessary).
My setup at the moment is ODROID C2 -> KALI -> DDDAC. I use pin 39 on KALI for ground connection. It is important, that you connect the ground through the complete chain. In your case: WAVE IO ground -> input Pin 39 of Kali -> output Pin 39 of Kali -> DDDAC.
I don't think that ground signal pairs are really necessary. It might help, but I managed to get a stable setup with only one ground line. This was the case for the test with the isolator chip wihtin the chain (between ODROID an KALI) and also now without it.
For me, there was no audible improvement caused by the isolator. That's why I'm now running without it. I guess the KALI does also a good job with 'isolating'. The series resistors in the I2S lines are not necesary in my current setup without isolator chip.
If you do not really need USB connection, you should try this chain (ODROID C2 + KALI + DDDAC). It sounds amazing.
By the way. If you have the crackling issue, what is the lock LED on KALI doing. Is the lock LED stable?
We can also discuss this issue in Kali thread, to get not too offtopic...
just a pet on my own shoulder 😀
Nice Testimony from Paulus from the Netherlands 🙂
Please note his comment on the "Air Cables" I made them also and they are really great for a small DIY budget 🙂
Link to Testimonies on my website
Nice Testimony from Paulus from the Netherlands 🙂
Please note his comment on the "Air Cables" I made them also and they are really great for a small DIY budget 🙂
Link to Testimonies on my website
There shouldn't be any issues using an RPI-based setup. I use an RPI into Ian's FIFO/isolator/re-clocker, into my DDDAC and it has been stable for months. Prior to that I used a BBB with Acko S03 and that was also stable.
How was the sq as compared to a Pc based setup? I will try the kali based setup in a few days.
Ok, after reading the thread on I2S cable connections I gave the resistors a try. I installed 33Ω resistors in all three I2S lines on the WaveIO output and also on the Kali output. In addition, I was more careful to use dedicated signal return (GND) wires for all three I2S signal wires.
The result was much worse than before. The crackle/noise was so strong that it was hard to hear the music. But this also revealed that the noise happens on one speaker only; the other channel is actually fine! It seems I didn't realise this before, when the noise was not as strong. Again, this points to a problem with my DDDAC (e.g., the 1/2 clock delay mod) rather than to a problem with the WaveIO/Kali combo, where the left and right audio data travel the same path.
To further debug this, I removed the I2S resistors on the Kali output. Now the crackle/noise was (almost) gone! It seems I2S signal was degraded by the resistors, which again seems consistent with my suspicion of a problem with my DDDAC 1/2 clock delay, which is the first circuitry that receives the signal from the Kali. Nevertheless, improving the signal returns obviously helped a bit.
Apart from my (minor!) noise/crackle issue, I can confirm that adding the Kali is definitely worth it! The <3 ps jitter spec of the Kali really does seem to matter, as the sound improvement over the stock WaveIO I2S output is substantial. I was listening to a live recording with someone coughing in the audience. This sound made me jump, because it sounded like someone was coughing in my living room although I though I was home alone...
I guess the next step to get rid of the residual noise/crackle would be to try a blue DDDAC board that has a properly working 1/2 clock delay. I'd be happy to take offers for a used board 🙂
P.S. michl2604: the Kali has no problems with locking to the I2S signal coming from the WaveIO.
mbrenwa
You could look at the driver settings of the WIO in your pc and change the freq. settings from 32 to 24bits. It solved this same proplem in my setup.
You could look at the driver settings of the WIO in your pc and change the freq. settings from 32 to 24bits. It solved this same proplem in my setup.
Go to search, choose 'TUSBAUDIO controle panel' > tab 'Format' and choose setting: 2 channels, 24 bits
good luck!