Help debugging BL616 firmware / FPGA-Companion

[vanfanel]

Hello,

I'm going to try to add support for my USB keyboards (8BitDo, IQUNIX L80) to the FPGA Companion firmware. I use it with the M0S Dock (BL616 microcontroller).

I have built both the BL616 firmware in the past, so that's not a problem.

The problem is: How do I get debug traces from the M0S Dock to my GNU/Linux PC while having an USB keyboard connected to the USB-C port on the M0S Dock?

I have already tried connecting the TX/RX pins from the M0S Dock to my serial/USB adapter.. and I need to somehow power the M0S Dock, obviously, but... how to do that while the USB-C port is occupied by the USB keyboard?

[vossstef]

@vanfanel you need to connect a USB-UART 3.3Volt interface cable (aka FTDI cable) to the TX,RX (crossed) and GND pin of the M0S Dock. At the edge where USB-C is located. Baudrate is 2000000 8N1 no flow control. For the Board using the integrated BL616 i routed the TX and RX to pins and you could connect the UART cable to those. Anyway best approach is to use M0S Dock for development as it can be operated independantlog of the FPGA code.

There is already a very big change in progress that @RetroSilicon developed dealing with USB HID and GameController mapping.

Development Branch : https://github.com/MiSTle-Dev/FPGA-Companion/tree/bl616_retrosilicon_hid
SDK CherryUSB 1.5.3: https://github.com/MiSTle-Dev/bouffalo_sdk/tree/cherryupd3

We have a Discord channel. Maybe you want to join. https://discord.gg/T5M3fWAQ

Most expensive solution is https://ftdichip.com/products/ttl-232r-3v3/ but there are far cheaper type available everywhere. Support of 2M Baudrate is important

[darfpga]

@vanfanel M0S Dock can be powered directly from its 5V and GND pins on the header side. M0S dock schematic show there is only one fuse between VBUS and 5V. Make sure to use a single power source (the debug PC USB port is a good one). Anyway Be sure to disconnect 5V when reprogramming M0S dock from the USB -C cable connected to PC.

[vanfanel]

Thanks for the information, I will keep trying to communicate with the M0S via /dev/ttyUSB0.
But I remembered something: On HIDMan (https://github.com/rasteri/HIDman), in order to get those keyboards to work, I need to activate "Adv.Keyboard".
In the HIDMan docs, it seems to mean:

Adv.Keyboard and Adv.Mouse switch the keyboard and mouse respectively into "Native Mode". This may be required to access certain functions such as scroll wheels and extra buttons on mice.

Note : Some USB devices with both keyboard and mouse interfaces (e.g. mice with configurable buttons) require both Adv.Keyboard and Adv.Mouse to be enabled otherwise they behave unpredictably.

Does it ring a bell on what could be going on with these keyboards?

Also, I have just built the fpga-companion firmware using these branches:

https://github.com/harbaum/bouffalo_sdk.git -b cherryupd3
https://github.com/MiSTle-Dev/FPGA-Companion.git -b bl616_retrosilicon_hid

...But keyboard (in)compatibility remains the same.