• The Vendor's Bazaar forum is for commercial offers and transactions. Only unmoderated members can post here.

    diyAudio provides this forum 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.

USB UAC2+HID Multichannel input/output interface York

I'll send you a new version of the config tool with new feature: set all settings to default.
Hi Marcin,

I've added a new button to the about tab to reset the settings:
1736189502967.png


After click it would set default setting which is I2S 2 channel output with on-board 512fs clocking. Try checking I2S right after that, no need to change any settings. I've checked on my side, if it does not work - then something happened to the hardware.
Link to the new config tool: https://york.eclipsevl.org/config_tool/york_config_tool_06_jan_2025.zip
 
HNY @eclipsevl,

Ihaven't readed it all, so sorry if already asked : I'd like to know if the tool provided to prg it through USB makes the board compatible for AD1862 ?
Happy New Year! 🙂 Sorry for delay with the answer.
Yes, it is can be configured for AD1862, the configuration is exactly same as for PCM63 which is tested in real hardware (NOS, 2 channel, 20 bit)

Futur option with an uf-l/output header for TDA1541A simultaeous without stop clock ?
I think it should already work, configured as NOS 16 bit 2 channels. If you are interested, I can run an experiment.

I like very much your idea of outside recloking. Should permit with few socket optionsto try some differentXtal and why not external MCLK via ufl or horizontal SMA plug.
Ok, I'll consider this in next revision

Hi @Tesla Audio ,
Is possible to configure york board in multichannel mode with right-left i2s for Miro's AD1862 without shift registers. Idea is to make multichannel DAC
Unfortunately, no. There is not enough serial data output pins in the MCU.
But how many channels you are talking about?
 
  • Like
Reactions: diyiggy
Hi Marcin,

I've added a new button to the about tab to reset the settings:
View attachment 1403963

After click it would set default setting which is I2S 2 channel output with on-board 512fs clocking. Try checking I2S right after that, no need to change any settings. I've checked on my side, if it does not work - then something happened to the hardware.
Link to the new config tool: https://york.eclipsevl.org/config_tool/york_config_tool_06_jan_2025.zip
Hi Vladislav,

Reset fixed the issue of wck.


On windows wck is always set as the output. If I set 48 it's 48 if 96 it's 96 etc. It doesn't change depending on file.

Now on aurender it works and change wck to file frequency. But there is different issue that you can see on movie file. I found certain songs that jam the card. In the file you will see ZZ top la grange 24bit 192khz to gus gus over 16bit 44.1khz. Then I need to stop because if I will hit last song again it does the same. Then I play 16bit 44.1khz file and I'm able to go to gus gus over. But then I switch to justice genesis 24bit 44khz and the same it pops. Justice album is even better because I'm not able to play this album from beginning I need to first start DANCE from the same Justice album and then choose genesis song.



Plug pin still not work after reset it drops from 5V to 1V when isolated side is powered with 5V.
 
Now on aurender it works and change wck to file frequency. But there is different issue that you can see on movie file. I found certain songs that jam the card.
Thank you for the video! That was something I've never seen before but after many cycles of sample rate changes on Windows I managed to reproduce the issue. In my case it was rare occurance but I guess it also depends on the driver/OS.

I have uploaded new firmware to the server, it should fix the issue. Please try.

Plug pin still not work after reset it drops from 5V to 1V when isolated side is powered with 5V.
The 'plug' pin is unfortunately not controlled by MCU, it is just pulled up by 10k resistor to isolated power supply. But is still should be 5 or 3.3v unless some low resistance load is connected to it.
Reason fur such decision is use of ISO7741 isolators: by default if one side of isolator is not supplied, the second side would output logic high. Which is not suitable for 'plug' signalpurpose. Instead, the 'mute' pin (right next to 'plug') is controlled by MCU.

Let me think about the best solution for this.
 
If possibly confirm, the simultaneous splited data mode for TDA1541A is feasible according this : https://www.diyaudio.com/community/...ate-nos-dac-using-tda1541a.79452/post-7837119 timing (and also w/o stopclock as said in PM) I think it will interrest lot of fellows and at least I get one to support your work ! 👍

on a least importance : on the reclocker board, I think people migth like a double foot print inside one another : smd for NDKA, and hybrid smd/tht bigger one for Accusilon and CHD-957 if compatibe with your shematic.
 
Last edited:
  • Like
Reactions: eclipsevl
Thank you for the video! That was something I've never seen before but after many cycles of sample rate changes on Windows I managed to reproduce the issue. In my case it was rare occurance but I guess it also depends on the driver/OS.

I have uploaded new firmware to the server, it should fix the issue. Please try.
Did update. Still have this issue. When it jam I really need to find some file to unlock it or power boot pico.

On windows 11 I'm not able to get this problem as it's not bit perfect and play everything in 96khz as output is set.

On android phone also I can't have this issue as it doesn't play 44.1khz. It set 44.1 file as 48khz.

I need to check on raspberry pi but I need to install some software on it.
 
The 'plug' pin is unfortunately not controlled by MCU, it is just pulled up by 10k resistor to isolated power supply. But is still should be 5 or 3.3v unless some low resistance load is connected to it.
Reason fur such decision is use of ISO7741 isolators: by default if one side of isolator is not supplied, the second side would output logic high. Which is not suitable for 'plug' signalpurpose. Instead, the 'mute' pin (right next to 'plug') is controlled by MCU.

Let me think about the best solution for this.
I want to hook everything nice and neat so I wan't to order long pin header female just to put reclock board on top of pico not on the bottom. Because of this plug pin issue I need to place 5V on the dac board pin from power supply. As temporary solution I guess you don't advise to bypass this 10k resistor to get 5V straight from isolated side PSU? If no do reclock board need this pin to be connected with york board?
 
Hi Marcin,

thank you so much, it was very helpful! The waveforms on the video gave me great hint to where the issue is.
Also I was able to reproduce it on my side with Qobuz.

I do not know what happens exactly but identified the place: it is function that recognizes DoP stream. For some reason it gives false positive for DSD stream for a moment. It is also visible on your video, for a moment wck (blue waveform) turns into some high-frequency pulses, the dsd_on pin also goes high for a moment in my experiment.

For now, I have disabled DoP recognition and uploaded new firmware. It should resolve the issue, please try.
Meanwhile I'll try to debug what exactly goes wrong.

I want to hook everything nice and neat so I wan't to order long pin header female just to put reclock board on top of pico not on the bottom. Because of this plug pin issue I need to place 5V on the dac board pin from power supply. As temporary solution I guess you don't advise to bypass this 10k resistor to get 5V straight from isolated side PSU? If no do reclock board need this pin to be connected with york board?
As I understood from previous messages, you want this pin to be functional (i.e. low when USB is not plugged in and high when plugged in), not just tied to supply, correct?
 
Last edited:
  • Like
Reactions: MarcinGD
Hi Marcin,

thank you so much, it was very helpful! The waveforms on the video gave me great hint to where the issue is.
Also I was able to reproduce it on my side with Qobuz.

I do not know what happens exactly but identified the place: it is function that recognizes DoP stream. For some reason it gives false positive for DSD stream for a moment. It is also visible on your video, for a moment wck (blue waveform) turns into some high-frequency pulses, the dsd_on pin also goes high for a moment in my experiment.

For now, I have disabled DoP recognition and uploaded new firmware. It should resolve the issue, please try.
Meanwhile I'll try to debug what exactly goes wrong.
Hi Vladislav,

Thanks. Now I can enjoy great sound quality from this card 🙂

As I understood from previous messages, you want this pin to be functional (i.e. low when USB is not plugged in and high when plugged in), not just tied to supply, correct?
Yes that's correct.
 
  • Like
Reactions: eclipsevl
Ceramic caps and ferrite beads?
For the oscillators, yes. The schematic was posted in this thread.
Only one the oscillators is enabled at once (if they have OE functionality).

Shared IC ground bounce?
Not sure if I understood it correctly. If you are referring to ground layout, the bottom layer of this board is a ground plane and all connections are done directly to it with via's.
 
Sorry to hear that. Not too long ago I found that ferrites, ceramic caps, etc., are bad for good clock performance. I know well what the app notes say, of course. I started out believing them.

The ground bounce you will have is from different signals switching at different times and all drawing current through the same IC ground bond wire. Well known problem if not using single (1G) flip-flop ICs.

Sometime mid last year I started working on a clock board design to share with the forum. This was based on some years of research and testing at my place. Mostly we did listening tests as we had the right crew to do it. Since then other people here and there around the world have built or modified clock boards using our findings, and they are very happy with the results. Much better sound stage precision and depth among other benefits. More results from other people should be posted over time. Some of the early comments are in the MarcelvdG RTZ dac thread.

Thread on the whole clocking and reclocking thing starts at: https://www.diyaudio.com/community/threads/general-purpose-dac-clock-board.413001/ Somewhere in there I had a reclocker board I got from Andrea Mori. I started to suspect there might be problems so started doing some experiments to see. Turned out there were problems so I designed a new reclocker that works and sounds better according to everyone who has started trying it, or else trying derivative designs. Link to my reclocker is in first thread post linked above.

Good luck with whatever you decide to do. I would suggest to maybe try a few experiments though. Also in that thread is a link to an article on why not to use ferrites in my clock board thread. That information is in addition to Bruno Putzeys' findings on what ferrites can do. The whole history of ferrites and why they are still used today is an interesting story for this EE.

Mark
 
Last edited:
  • Like
Reactions: eclipsevl