Hi Brice, I got stuck with the Ethernet interface last year which apparently produced some audible "clicks". @pcdimmer advised me to wait until he had re-viewed the Ethernet circuit design. Otherwise I got quite far. Had already programmed a simple UI running on a webserver that would allow several users to remote control their personal mix via a mobile phone. However, meanwhile a band member decided to invest €3000 in a Midas M32 with digital stage box which provides the same functionality for us, so I didn't follow up anymore. Neverhteless, I'm glad to read there is progress and will wait for the article and/or the youtube video to come out and then hopefully bring this further. The band might still use it as stripped down solution for the rehearsal room. If I make progress that may be useful for others I will post it here. @pcdimmer: Keep up the great work !
So far, I've got AES3 transmitter and receiver running on an RP2040 pico board with the following functionality
- Transmits and receives 48kHz sampled AES3 frames between PIO's (directly connected), sending sine waves for each channel
- Produces and detects sync and X/Y/Z (aka B/M/W) preambles
- Puts received samples on an RP2040 PWM channel (for low quality analogue output)
- Verified data format is OK with S/PDIF decoder on PulseView logic analyser
- Verified stereo consumer IEC958 (S/PDIF) is decoded by a random cheap S/PDIF DAC
- Transmits 8 channels 48kHz (i.e. Ultranet format) on direct pin-to-pin connection to itself
@pcdimmer do you have a way to capture the data frames that go with the Ultranet signal from your brilliant FPGA implementation, so I can reproduce those? My next step is to try to connect the prototype up as an Ultranet receiver, so any information would be welcome. I mean the 24-byte user data and channel status data that is transmitted 1-bit per frame. At present I'm just sending the minimal AES3 (professional bit set, everything else cleared, no CRC) or minimal S/PDIF (professional bit cleared, copy permitted, no additional data) but I'm not sure if Ultranet will generate/require something more than that?
With PIO code for other multichannel digital audio (e.g. ADAT lightpipe), it looks like the RP2040 could be the heart of a useful 'universal translator'. The PIO blocks are very powerful!
If there's interest, I'm expecting to upload to github if I can get it working as an Ultranet device.
Hi, I've been working on a modification of @pcdimmer's FPGA design, my project is to make a simplified version of the P16M for personal monitor mix in our church band. We already have some P16M units but they are getting to be very expensive to buy, and are also a bit complicated (too many knobs and buttons) for some of our users.
I have passed on many thanks to @pcdimmer for doing a lot of the hard work to get the basic blocks working, on top of his design I have added a simple user interface using a 320x240 LCD screen, 4 buttons and a rotary encoder. This just runs off some spare pins on the Arduino Vidor board. The audio is output using a UDA1334 stereo codec board. I 3d-printed a box for my prototype version.
I have made some modifications to the FPGA configuration to give basic audio metering on each of the 16 channels as this is useful for initial line checks. Also I have added a sync line from the incoming ultranet to the output codec as I was getting some crackling which I think was caused by the output free running vs the input.
It all works great, except one weird problem. Using an XR-18 mixer, the 16 channels are in the correct order. On an X32 mixer, ultranet channels 3-8 appear on my unit as channels 1-6, then ultranet 1-2 appears at channels 7-8, Same for 9-16. So the X32 obviously does something different in the frame encoding, but I can't work out what it is yet.
The RP2040 approach also sounds very interesting as the main drawback of the Arduino Vidor board is it's quite expensive at around £70-£80. I'd be interested to see the code for that. My fork of the original code is at
https://github.com/congoblue/UltranetReceiver
I have passed on many thanks to @pcdimmer for doing a lot of the hard work to get the basic blocks working, on top of his design I have added a simple user interface using a 320x240 LCD screen, 4 buttons and a rotary encoder. This just runs off some spare pins on the Arduino Vidor board. The audio is output using a UDA1334 stereo codec board. I 3d-printed a box for my prototype version.
I have made some modifications to the FPGA configuration to give basic audio metering on each of the 16 channels as this is useful for initial line checks. Also I have added a sync line from the incoming ultranet to the output codec as I was getting some crackling which I think was caused by the output free running vs the input.
It all works great, except one weird problem. Using an XR-18 mixer, the 16 channels are in the correct order. On an X32 mixer, ultranet channels 3-8 appear on my unit as channels 1-6, then ultranet 1-2 appears at channels 7-8, Same for 9-16. So the X32 obviously does something different in the frame encoding, but I can't work out what it is yet.
The RP2040 approach also sounds very interesting as the main drawback of the Arduino Vidor board is it's quite expensive at around £70-£80. I'd be interested to see the code for that. My fork of the original code is at
https://github.com/congoblue/UltranetReceiver