The card can also be potentially adopted to work with other sources such as Analogue Consoles. You read that correctly. So yes, this is a cheaper DAC potentially. The MCU has firmware that can be user-upgradable if necessary.
There are switches to turn on video LPF and also swap from RGB output and YPbPr output. I have direct access to time tables and I could theoretically plop in a custom video mode for YPbPr. I may experiment with that.
So I have finished the DAC card. It will convert YUV422, YUV444 and RGB colorspace digital video sources into analog video. The IC works with misterFPGA and direct video mode. It also works with the video from my scaler board. It will be sent to get made at same time as digital video board
Better.
UI WIP
Idk if i would use that term, but I am capturing screen data into FPGA's SDRAM and scaling an image output.
With audio.
At this time I can serialize audio into the video stream. Sometimes tomorrow Ill get controller working and might dig around for capture card.
After finding audio board and fixing backwards clock buffer (this broke a PCM and a buffer), and fixing FMT0 and FMT1 resistors to pullups, I got digital audio into the FPGA. Then I had to make an async fifo that was 48 wide for my PCM words.
Again, this took several hours for me to realize this.
So I re-implemented the hsync vsync de generation organically in the softcore but still generate coordinates externally. The hdmi softcore was broken for hours when it was supposed to carry valid audio.
NGPC update:
It seems when you tear out the timing module from HDMI softcore, you may misalign everything when generating audio.
....
Shielded flex helps.
I think washed out color is electrical issue...
Two very minor misalignments here but I mostly have it nailed.
I have since fixed the first scanline being messed up.
It turns out issuing 3 instructions in 1 cycle depending on each other made the BRAM select double toggle.
All thats left is fixing the initial garbage values popping out. Suspect SDRAM misses personally...
Digital video..some FFC EMI trouble and off by 1 issues. Very alpha build of 1440x240 for NGPC Digital Video.
2/2
I will order new board in a few weeks that is better with EMI and SI and shorten bridge FFC. Ill also begin designing the other subcarrier boards for other handhelds. If theres anything you want me to support let me know in comments please. Ill go buy one of them and study it.
1/2
2026 going to be lit 🔥
Framebuffer project totally works and this is NGPC working in 1440x240. NGPC must be triple buffered because the clocks turn off in blanking just like GBC, GBA, and launch model DS. Will test 640x480 at some point today, but fairly confident it will take less than 1 hr.
Its a signal integrity nightmare but ironically works. Verified pixel accumulator logic and frame indexing/arbiting.
Will have purpose built PCB in a few weeks for NGPC interfacing so this wiring nightmare go away
It seems write ptr is inaccurate and also miss writing one entire screen to region; duplicate data from only 1 screen is written to sdram. So readout is distorted.
Nintendo DS Digital Video Progress
Is there anyone even using this platform?
Hi
