Hi there, I’m the developer of the ezsound-6x8 soundcard for the RPi5, and I noticed you’ve posted a lot on the Zynthian forum about multi-channel cards. I wanted to let you know that my soundcard is now live on CrowdSupply, and you can subscribe to updates at the campaign page. I didn’t want to post an “advertisement” on the Zythian forum, but it seems there’s quite a few people who might be interested… what’s the best way to let them know about it?
Hi @gordoste.
I would defer this to @jofemodo who is boss around here. (I invited him to this topic.) I suspect he will be okay with mentioning this project but it is his decision. (I have been burned before by other forums imposing harsh constraints on this kind of activity so you are wise to ask.)
Some observations:
- Why 6x8 and not 8x8? I guess it is the DAC/ADC you chose. We have done some work with individual 2-channel DAC/ADC to create 8x8 interfaces.
- The description says it is only 96khz but some users may want a lower rate, e.g. zynthian runs at 48kHz to ease processing power. (Also, the correct units is Hz, not hz and to be further pedantic, it is really 96000 frames per second (fps)
). Ah! Just re-read and see it says, “up to 96 khz” but later says, “only 96khz”. Maybe that bit needs rewording to make it clear it is a maximum and not a fixed value. (It caught me out!) - " * Drivers already included with Raspberry Pi OS" - I haven’t checked recently. I had to write a driver to get a card working with all 4 I2S. Has this now been implemented in ALSA?
- Are you considering alternative physical presentation, e.g. TS jacks, balanced (differential) i/o, etc. (I know that may be a step too far for an initial release but it is something that some of our users appreciate in the zynthian soundcard.)
- I see what looks like a power inlet. Does your design depend on a separate PSU and how clean does that PSU need to be?
Good luck. It looks like a cool project / product.
- You’re right - 6x8 because I selected the codec with Linux drivers already available.
- The driver sets the codec to run at 96kHz, however you can open the audio stream at other rates and ALSA automatically converts it. So Zynthian would work fine. If you really want the codec to run a different rate, it’s just a matter of changing the devicetree overlay. I did it this way to make life easy for users.
- You don’t necessarily need a driver to get all 4 I2S channels working, you just need to configure the devicetree properly. The fact that my codec already has a driver available, plus the ALSA SoC framework means that everything I needed to do was done in the devicetree. However, there was a shortcoming that needed a kernel patch (it wasn’t possible to have the CPU as I2S clock consumer with multiple codecs attached to the same clock line and one of them acting as clock producer).
- Alternative input/output options are certainly possible - the HAT with the codec sends and receives differential signals and the existing I/O boards convert this to/from single-ended. Of course, you could easily substitute balanced input/output instead.
- The audio circuitry is isolated from the Pi deliberately - the Pi generates plenty of high frequency noise which isn’t friendly to ADCs/DACs. For best performance, you want a separate power supply (9-15V), but you could just run it off the Pi’s supply with a boost converter to get up to the right voltage. It doesn’t need to be very clean as there is a fair bit of filtering. The difference won’t be obviously audible, but it will show up in measurements.
1 Like
From my POV, it’s perfectly OK to inform about the project, including URLs, etc. Also it’s fine to discuss about the technical details. I’m pretty sure you will find users interested on knowing more bout it 
I will open this thread to the public, OK?
Regards,
Yes, that’s definitely OK