I did. I will look again. No such thing is reported.
I changed hwdef to make pwm5 and RSSI pads uart4 on serial1. It compiled and flashed ok. I think I need to first test my new UART to see if it works properly.
If nothing is reported then that implies that the uart cannot be found or opened at all. There is a debug flag at the top of AP_RunCam.cpp that you can set to 1 to get extra output if you recompile.
One thought - make sure that HAL_MINIMIZE_FEATURES is not set to 1 for your FC - I can’t tell based on the information you have given, but it’s possible. If that is the case then define HAL_RUNCAM_ENABLED 1 in hwdef.dat
However, I have tested my newly created serial port and it is not working for sure. I have created defined UART4 as RSS and PWM5 on the OmnibusF4 Pro by editing the hwdef file. So I have a UART problem.
So the ordering I understand, but the pins are dictated by the physical hardware - not sure I understand how those can be re-assigned - maybe it’s just that the hwdef.dat is wrong. The best way to figure out the correct pin assignments is to look at the betaflight code. I have a omnibus 4 pro so can have a play if you can’t get this to work.
After Henry W explained that the hwdef file had an expected default uart order then I had the missing link of knowledge. Now i have a working and tested extra UART on my OmnibusF4 Pro.
Where do I find HAL_MINIMIZE_FEATURES and HAL_RUNCAM_ENABLED. Still no Runcam mavlink message on MP even though I have a tested serial2 at 115 and protocol 26. I suspect the feature was not included in my compile.
I am getting there – so much to learn before I could test this, sorry.
I was interesting that my Caddx Turtle v2 had flashing serial leds with mavlink 2 at 115. But registers nothing when serial2 was set to protocol 26.
The f4 pro does not have HAL_MINIMIZE_FEATURES enabled so you should be good there. I suggest switching on the debug flag and seeing what output you get. If HAL_MINIMIZE_FEATURES is not set then HAL_RUNCAM_ENABLED will be enabled by default so you should be good.
You are sensibly anticipating the git noob questions before they are even asked. I just git cloned your fork and set up a new eclipse project. Inefficient I know…but it is compiling.
I have cloned your runcam branch in Eclipse, compiled, flashed, and am now getting Runcam initialized in my Mavlink messages from my Nano Talon. This would have taken a more practiced Git/Eclipse/Ardupilot person 20 minutes. Me not so much.
Anyhow will test tomorrow or Saturday when I get a chance.
@andyp1per I have assigned channel 10 (RC10) to function 78 and channel 9 (RC9) to function 79. The switches do nothing mind you. I cannot enter the menu either. But when i do enter the menu it seems that the FC is out putting endlessly a left button selection, so it makes the control pad useless until I disconnect the Caddx serial port. So it is receiving serial commands.
I am puzzled by the mechanism whereby we are using the RC Yaw or Roll. This will not work when in flight will it? What am I missing there?
So to summarize I am getting the mavlink message affirming the that Runcam is present. Some sort of serial communication is happening but it is only a recursive left toggle of the 5 position switch which I have to disconnect the serial port to stop. Two RC channels assigned to a switch and RC function 78 and 79 seem to have no effect on this behavior.
It is a start! I have a spare runcam split 2S. Sitting in a box! Send me your address and I will send it to you if you like. My donation to the cause of getting this working in Ardupilot.