Moderators: Mug UK, Zorro 2, spiny, Greenious, Sorgelig, Moderator Team
My reason for posting here is to ask for help converting the Quartus project from being just 'DE10-Nano compatible' into something 'MiSTer compatible' - before I get too far into the weeds, and write any more that will just need to be rewritten / refactored later anyway... I don't have any familiarity with the MiSTer 'wrappers' / ecosystem / pinouts, and the learning curve is a little intimidating!
YesuXe wrote:I need serial RX / TX pins, are there specific preferred GPIO pins to use for serial with MiSTer?
Code: Select all
input [6:0] USER_IN,
output [6:0] USER_OUT,
you set 1 to output port permanently, so you can always read corresponding input port. For output simply 0/1 is enough. The output is hard 0 and pull-up 1. It's important to keep USER I/O the way it is, so if other device will be connected to the same port, then it won't compete by logic levels.uXe wrote:With the USER_IO - the pins are bi-directional, but I just want simple output for TX and input for RX, how would you suggest implementing this?
(PS. THAT MUSIC IS NOT COMING FROM THE ARDUBOY - IT'S YOUTUBE FILLER!)
Not hard, for you.Sorgelig wrote:Sketches are definitely possible to load from files. It's not hard.
Hours of scouring through existing projects to figure out and understand how it works and then cobble something together, for me. 
Code: Select all
assign sd_rd = (sd_ack) ? 0 : sd_rd_in;
assign sd_wr = (sd_ack) ? 0 : sd_wr_in;
Code: Select all
while (~(*simple_in & SD_ACK_MASK)) {*simple_out |= SD_RD_MASK;}
*simple_out &= ~(SD_RD_MASK); // sd_rd LOW
while (~(*simple_in & SD_ACK_MASK)) {*simple_out |= SD_WR_MASK;}
*simple_out &= ~(SD_WR_MASK); // sd_wr LOWthis won't work! When sector finishes to transfer , sd_ack will go low, and then your sd_rd/sd_wr will get original values, probably 1 and ARM will send the sector again and again.uXe wrote:I have this code in Arduboy.sv to make doubly sure rd / wr immediately go to zero if sd_ack is HIGH:
assign sd_rd = (sd_ack) ? 0 : sd_rd_in;
assign sd_wr = (sd_ack) ? 0 : sd_wr_in;
I was already planning on checking out MiSTer, so nobody to blame but myself... Great! Looks like that bug I mentioned with the pixels being shifted over by one did not show up for you, good to know! (Maybe something with my setup then? I have a heatsink but no fan, so maybe running a bit hot and effecting the timing somehow?NML32 wrote:Thank you uXe, I was able to load Circuit Dude.
) EDIT: or to do with using HDMI vs analog?? Fair enough - not intending to re-invent anything though, this was just the quickest way to get a working demonstration off the ground in the limited 'spare' time I have!Sorgelig wrote:I don't understand why you need to re-invent the wheel.
I can imagine other use cases in the computer cores etc. where being able to send in a block of data over the internal UART connection could be a useful option though?