Support Tang onboard BL616 µC TC60k and TN20k / green LED fix

[vossstef]

new Features:

• Disable the green LED to save power and minimize light pollution. Now red LED in use.
• Support Tang Nano 20K onboard BL616 µC for the FPGA Companion firmware
• Support Tang Console 60K onboard BL616 µC including the WIFI modem for the FPGA Companion firmware

Note

• Only newer Nano 20K types marking : 3921 are supported
• Console 60K +TN20K require an extern powered (see ->) USB Hub + USB-A to USB-A cable + USB-A Power Supply Plug

Warning

Important Console 60K FPGA programming note
Recommended to use Gowin 1.9.11.03 programmer toolset.
You have mandatorily to press the reconfigure button on top of the console plastics lid before programming.
Timing of button press or better say release is critical in success of the FPGA programming !
You have to give extFlash Program Go/Start immediately after release of the button otherwise programming will fail !

Background
The FPGA bitstream will reconfigure the programming JTAG pins into a SPI Interface and by that further programming attempts are blocked. Hardware connections in between onboard BL616 and FPGA are very limited and there is no other way than make use of both JTAG and UART signals to compile a proper SPI Interface with interrupt capability. For Tang Nano 20K the situation is different and JTAG chain can be kept intact all the time.

onboard BL616 µC
The BL616 µC enters programming mode when you keep the BOOT button pressed during a power-up (bottom left corner, close to one of the USB-C ports). You always can go back to factory default by just loading the bl616_fpga_partner_Console.bin firmware at address $0.

---

This discussion was created from the release Support Tang onboard BL616 µC TC60k and TN20k / green LED fix.

[RetroSilicon]

Hi,
I'm not sure which firmware (for BL616) I should upload to the Tang Nano 20K to use its internal BL616.
Should I use the fpga_companion_NANO20K_onboard_bl616.bin file?
Thanks for your awesome work!
Chris

[RetroSilicon]

Thanks for pointing me to TangCore bl616, I'll definitely take a look.

Unfortunately, I don't have any 138k boards — I've been focusing on the 20K only, sorry about that.

By the way, did you try using the m0s debugger mentioned here (https://github.com/harbaum/MiSTeryNano/tree/main/firmware/m0s_debugger) ? I built it today for testing (just rx and tx wires between the m0s boards), but the screen stays black after connecting to the m0s debugger board (usbmodem) from my PC... Should I use 2000000 as baudrate ?

[vossstef]

yes , 2000000 as baudrate ?

i developed for the onboard bl616 "support" cases something.
https://github.com/vossstef/bl616debugdisplay

It's just reading. No pin left in between fpga and bl616 for giving keypress...

[RetroSilicon]

Hi, thanks a lot for the tool! It really helped me figure out where the enumeration code stops (I got the "Failed to set configuration, errorcode: -8" error). And the VT52-style display is awesome :)

By the way, which particular lines did you mean when you said you "copied a few lines from the TangCore bl616 firmware"? I'd like to check if I'm missing something on my side.

[vossstef]

@RetroSilicon
the video of the terminal is rather buggy and will fix in the future. Good enough for a little usb debug to figure out that usb hid don't have power....

initialization

// from fixcontroler.py: https://gist.github.com/adnanh/f60f069fc9185a48b73db9987b9e9108
struct usb_setup_packet xbox_init_packets[5] = {
{0x80, 0x06, 0x0302, 0x0409, 2}, // get string descriptor (SN30pro needs this)
{0x80, 0x06, 0x0302, 0x0409, 32}, // get string descriptor
{0xC1, 0x01, 0x0100, 0x0000, 20}, // control transfer 1 (a lot of pads need this)
{0xC1, 0x01, 0x0000, 0x0000, 8}, // control transfer 2
/* {0xC0, 0x01, 0x0100, 0x0000, 4}*/}; // skipped as 8bitdo wireless adapter hangs on this

// four data packets to EP2
uint8_t xbox_ep2_packets[4][3] = {{0x01, 0x03, 0x02}, {0x02, 0x08, 0x03},
{0x01, 0x03, 0x02}, {0x01, 0x03, 0x06}};

[harbaum]

I’ve almost finished two Tang Nano 20K cores
That sounds great. I just migrated this project into a separate organization on Github. I'll invite you there.