Yesterday I finally made the decision to move on from the Pixhawk/Companion and upgrade to the Navigator/BlueOS. I followed the available guides, and the hardware integration went without issues. However, when powering on the system for the first time, the Pi/Navigator would not boot. The system would just beep the first 3 times but not the final 2 (even when waiting for 10-15 minutes).
I checked to see if everything was correctly assembled (it was) and eventually just tried a different battery. I finally got it to boot. Took the time to update the OS (1.3.1) and Ardupilot (4.5.1) and explore the new environment and, when rebooting, the system wouldn’t boot again. After some tests I found that from the 5 batteries we own, 2 can boot the system consistently (50 to 70 seconds boot time, though they might fail to boot occasionally, 1 fail every 8-10 tries), while the other 3 are less likely to boot the system (1 boot in every 4-6 tries).
Looking at the led pattern in the Pi I can now spot if the system will boot or not.
Trying to solve the issue I searched the forum for similar issues and found this and so i tried to disconnect the DVL and test again. With the DVL disconnected the system booted correctly every time (to this moment) independently of the battery used.
The DVL firmware is updated to the latest version 2.6.2. and is operational and working (from http://192.168.2.95/#/) even when the system doesn’t boot
Our DVL A50 was supplied directly from Waterlinked and integrated following their documentation.
The Pixhawk/Companion solution was running everything ok, without any major issues.
I don’t know how to fix this…Any ideas?
Our Rov is a heavy-configuration mod with the DVL A50 and Ping360 installed.
Hi @FredO -
Welcome to the forums!
Could those batteries have a low voltage compared to the others, insufficient to provide the system with power? Are they the same type, from Blue Robotics?
Checking the battery voltage and 5V power supply breakout on the Navigator with the system powered and sharing the values would be helpful. Downloading system logs from the gear icon in the center of the left menu and posting them here would also give us something to examine!
Hi Tony, thanks for the quick reply. I will be out of the office until Wednesday and will share all the information when I get back. For now I can say that one battery was fully charged for the upgrade process but it turned out to be one of the “non-bootable” batteries, all others were in storage charge (I don’t know the exact values of all batteries but I will look into it). All batteries have the same specs as those from Bluerobotics, but were produced by the company that developed the ROV.
Hi Tony, I made 5 boot tests using just two of the batteries and these are the results:
Battery 1 (14.56 V - Storage) system booted successfully 4 times (out of 5)
Battery 2 (16.22 V - Charged) system booted successfully 1 time (out of 5)
We always get 5.02 V for the main 5V connector on the Navigator even when the system doesn’t boot.
Also, probably not related but I was having some trouble downloading the System log files using the gear icon. I would get the following notifications:
“Filebrowser
FOLDER_FETCH_FAIL
Could not fetch folder system_logs: Request failed with status code 401
less than a minute ago (1 times)”
after a couple of reboots I was able to download it: system_logs.zip (8.8 MB)
Hi @FredO -
What Willian is saying is the 5V UBEC in your ROV doesn’t seem to reliably provide your system with voltage sufficient to boot up. You could check this with a multimeter- I would suspect it shows just under 5V when things are first plugged in.
You could try replacing it, or simply removing things that are also powered off of it - if using an ethernet switch this can be powered directly from the battery, for example. Otherwise, make sure both 5V connections from the UBEC are connected to the Navigator!
If you have an older revision ROV, it could be using an earlier version of this power supply, which were not as good…
Our 5V UBEC was probably from early versions (maybe 2019), but I had an extra UBEC 5v6a-r1 in the ROV spares (from 2023) and I replaced it. Also, both 5V connections from the UBEC are connected to the Navigator (as they were since the beginning) and there is nothing extra being powered from the Navigator. The Ping360, DVL A50 and ethernet switch are connected directly to the power terminal block.
Nothing changed, still the same behaviour.
update: disconnecting the Ping360 from the terminal block will make the system boot normally, similarly to what i tried with the DVL.