pengi
lvl.1
Sweden
Offline
|
spikestabber Posted at 7-30 06:40
This almost seems like a power issue, the air unit will use slightly more power as it heats up. I have questions about the thin gauge power wires that feed the air unit, especially at lower voltages such as 7.4V. The newer firmware will likely consume a lot more CPU capacity thus consume more power. For a device that is known to draw 7W and likely peaks much higher the more you turn up the tx power, the stock power wires themselves seem questionable.
I have some updates regarding power, since I've been trying to isolate the issue. And I have an extra spare Air Unit, which I've saved for another build. But wanted to test.
Calling the Air Units AU1 (old and well used) and AU2 (new).
Have been running it both on 25mW and 500mW. Both gave the same error on the AU1.
Thought it might be some problem with the Air Unit, for example some moisture might have slipped in. So changed to the AU2. No difference, still problem. So switched back, so AU2 can be saved to the next build.
Have been running it from the Holybro Kakute F7 HDV which contains a 8V 2A regulator, so rebuilt the cable to take power directly from the cap on the input, since i'm running 4S. No difference, same error.
I also thought it might be a thermal issue, so I remounted the AU so that it has thermal connection via thermal tape to the frame, and better in the airflow. Now I can't even feel it getting warmer when flying. Only when I have it standing disarmed for a while.
I downgraded v01.00.04.00 just to see if the problem still was available, and in fact it was. (Using AU1 on 8V, however cooler mount). Upgraded back to v01.00.06.00
Also started using the recommended SanDisk 128G card, providing 81.7MB/s write and 89.7MB/s read using the Blackmagic speed test. Previously I used a Samsung EVO card with write speed 47.8MB/s and read speed of 90.0MB/s.
One note is that both Air Units was bought at the same time, so might be the same batch...
I'm happy to do some more research and testing, but at this point, I have not anything more to go on except the message "SD CARD Slow" in the goggles and an encrypted from the tool. So is there any way to get a more verbose message, since I'm lost.
I have one last theory, which I can't really test... I'm thinking if there is a mechanical issue, that it appears during high-G events, and the physical connection between the card and the board glitches. Which if it just happens sometimes, I guess it should be possible to recover from? But if so, I would at least expect a an error message not saying the "SD CARD Slow", but rather write error. And possibly trying to recover the connection and write a proper file end so what's recorded actually is valid. (might be able to open the air unit and jam a paper over the card to add extra pressure) |
|