Actual clock frequency isn't an issue on Pi0-4 as the receiver is very forgiving. Pi5 is a little fussier.No specs, nothing. Measuring clocks was my first idea, but my oscilloscope is only 200 MHz.
200MHz bandwidth should still be enough to see LP-HS transitions on the lanes. https://download.tek.com/document/61W_25772_0_HR.pdf figure 3 shows you how the lane will shift from 0/1.2V LP signalling to 100mV/300mV HS signalling. That will at least tell you which lanes are active.
Unlikely.If I activate raspberry pi camera via libcamera-hello, but instead of CSI-2 data lines from camera I put mine instead, will it work? I mean will I be able to see some noise at least?
Image sensors require quite a lot of setup in order to actually start producing MIPI data. A quick look at the register writes in the imx708 driver would give you a very rough idea.
You'll also find that most kernel image sensor drivers check that they can talk to the sensor over I2C as part of their probe sequence, so having something else connected would mean the driver doesn't probe.
What is this device? Some image sensor that you've recovered from a phone? In which case you've also got to sort the relevant power rails and clocks, so you'll need a pinout of any connector. And note that the D-PHY lanes have to be connected in the right order and right polarity too.
Statistics: Posted by 6by9 — Thu Mar 27, 2025 12:45 pm