Custom board with XUF216-512-TQ128-C20A - xrun works, xflash doesn't (solved)

Technical discussions around xCORE processors (e.g. xcore-200 & xcore.ai).
dtns
Member++
Posts: 17
Joined: Sun Sep 01, 2024 6:05 pm

Post by dtns »

mmar wrote: Sun Dec 08, 2024 3:40 pm Areyou sure your IC is XUF and not XU ? When yes show your schematics
Yes, unless my PCBA house did a mistake.


schematics.png
Screenshot 2024-12-10 at 11.46.53.png
You do not have the required permissions to view the files attached to this post.
dtns
Member++
Posts: 17
Joined: Sun Sep 01, 2024 6:05 pm

Post by dtns »

Update: Flashing was successful by adding --spi-div=8 to xflash. Which is confusing for me.
Joe
Verified
Experienced Member
Posts: 83
Joined: Sun Dec 13, 2009 1:12 am

Post by Joe »

So all the reading works from the looks of it.

Looking at your layout you have a trace on the QSPI clock (X0D10) out to somewhere else on the board - this may well cause signal integrity issues - this is now a long stub on the clock line which will cause the clock signal the flash sees to be poor.

Similar story for the other QSPI flash signals - need to make sure whatever this connects to externally doesn't affect the internal QSPI flash.
QSPI.PNG
You do not have the required permissions to view the files attached to this post.
XMOS hardware grey beard.
mmar
XCore Addict
Posts: 134
Joined: Fri Jul 05, 2013 5:55 pm

Post by mmar »

dtns wrote: Tue Dec 10, 2024 12:02 pm Update: Flashing was successful by adding --spi-div=8 to xflash. Which is confusing for me.
From your image board show lines on every pin. But in real life some pins is restricted... And create long open antena wires is counterfail. Yuo can have trouble not only flash fw, but too on reset boot from flash on speed used in loader. Too in schematics image isnt clear used and unused pins ...
Joe
Verified
Experienced Member
Posts: 83
Joined: Sun Dec 13, 2009 1:12 am

Post by Joe »

An update on this. This is potentially a bug introduced into the 15.3.0 tools which we confirm is fixed by running with spi-div=8. It probably isn't related to your layout although avoiding any trace stubs on the spi clock line is still recommended.
XMOS hardware grey beard.