The M2 rover is connected to a modified Sony Qx1 camera using C1 port pins Ground and Timemark.
M2 firmware is up to date, logging all constelations at 10Hz, except by the last flight (Campaign 2, Flight 2) when I changed to GPS only, still at 10Hz.
I have no clue about what is causing this problem, also have not identified a pattern…
It’s hard and time consuming to disassemble the drone, I am avoiding doing that, but the wiring is pretty simple: trigger feedback and ground from camera are connected to M2’s C1 port pins Timemark and ground. The rest of the pins/wires in the C1 port connector are not connected to anywhere.
I’ve noticed that even for the flights with no missing time marks, the number of expected positions in the Missing Summary is smaller than the number of positions in the _events.pos file. Do I get it right that the events collected on the ground were not included in the Missing Summary statistics?
Since the number of calculated positions is equal to the number of time marks in the raw data log, it seems that some time marks were initially not recorded by Reach. The first possible reason for it is that the camera wasn’t triggered at that time. To check it, please clarify if there are photos corresponding to the missing time marks and how the camera is triggered in the setup.
The second possible reason is that there are some issues with the connection between the Reach and the camera. If I get it right, the Time mark and GND pins are connected directly to the camera, and the others are not connected to any parts of the drone. However, as I can see from the setup photos, there are 5 wires attached to the Reach M2 C1 port. Do I understand correctly that all wires except for the connected to Time mark and GND are isolated from another side?
I forgot to tell, but yes: events recorded at ground level were discarted on the summary, those are recorded pre launch just to verify if trigger and M2 logging are ok, when this is done it’s possible to hear the shutter mechanism and check the last event record in the Reach App, I have never experienced a failure with event record on this pre-launch phase.
Those events / photos have no use for the survey, thus are discarted, and it’s really easy to separate from the rest by just looking at the altitude.
The events are really missing the exactly way as listed in the Missing Summary, I could provide all the photos if you wish. It’s not a trigger issue, for sure. The camera is a modified Sony Qx1, in which a logical 1-wire signal is extracted from the shutter/flash electronics and connected to the EMLID M2 for precise feedback, there is also a PWM input for trigger signal, which is connected to an EMLID Edge. This camera modification is on the market for several years now, proven to be correctly working with a few variations of RTK receivers and Flight Controllers.
Indeed only the Time mark and GND pins have use in this setup, the other 3 pins in the C1 port are not connected in the other end.
Thanks for the details! Since there are no missing photos, it looks like there might be some issues with the connection between Reach M2 and the camera.
May I ask you to tell me if the connectors that are not attached to the camera are isolated? It could also be of help if you provide me with a photo. During the flight, these connectors may touch each other and other drone parts. In case they are not isolated, it can influence the time mark transmission.
