So (just to summarize) the Pi 5 is producing timings very close to those requested, with a DSI clock exactly 24/2 times the pixel clock. But you note it might be in LP11 state for too short a time (the driver has tables of timeouts to try to prevent that, but perhaps there is a mistake somewhere).
The CM4 is producing something quite different from the requested timings, for reasons to do with its clock architecture, but does have the advantage of actually working!
All I can suggest at this point is to try to adjust the pixel clock and/or HFP to try to match what CM4 is doing, to see if that works.
The CM4 is producing something quite different from the requested timings, for reasons to do with its clock architecture, but does have the advantage of actually working!
All I can suggest at this point is to try to adjust the pixel clock and/or HFP to try to match what CM4 is doing, to see if that works.
Statistics: Posted by njh — Wed Jul 10, 2024 9:25 am
