Yellow traffic light goes out before boot is completely

[Tsail83]

So I have the MHat wired up with out power module exactly as the documentation says to. I have the same power source as my NMEA network. When on first boot the initial power on shows can network with steady yellow but before I get to the GUI the light fades out completely. 

Grok says this is due to a voltage drop and that I should be using a jumper wire from the GPIO 1 to the The transceiver next to the can module on the m hat. 

This seems drastic. Has a tone else seen this on MHat with pi5 open plotter off Sd card?

Thanks

[Sailoog]

Do not do that.
could you provide more details about your setup?

[Tsail83]

Do not do that.
could you provide more details about your setup?

Sure. The Pi is a Pi 5.The NMEA 2k system runs B&G Triton instruments. The connection for its power comes off the main panel to a bus board. I am also running the power directly to the pi from this same bus board. I have a small 10 inch screen, wireless mouse and keyboard connected to the Pi.  The screens power comes from an alternative source currently. 

I went with the MacArthur configuration that doesn't use the power module. The only wires connected to the hat are CAN L and CAN H. 

Thanks

[luu5]

So I have the MHat wired up with out power module exactly as the documentation says to. I have the same power source as my NMEA network. When on first boot the initial power on shows can network with steady yellow but before I get to the GUI the light fades out completely. 

I had similar issue, which for some reason has dissappeared. But when I turned my system up (Raspi5, AIS transceiver, Axiom9) CAN light flashed ok, but when AIS turned on, it caused NMEA2K to "reboot" and raspi can0 turned off. I could turn it on using terminal by first turning can0 off and then back on. And then it worked. I could also cause this by unplugging nmea2k cable off other devices.

Now for some reason, perhaps I did change something something in the setup, I have not experienced this for some time.

[Sailoog]

You should not power your raspberry/monitor from the CAN bus because it is too much. you have to power it directly from the battery just as your plotter is. and everything powered from the same battery bank.

[Tsail83]

You should not power your raspberry/monitor from the CAN bus because it is too much. you have to power it directly from the battery just as your plotter is. and everything powered from the same battery bank.

The pi is not powered from the can bus. The only input to the bus or Mhat in the CAN  L  CAN H wires. The NMEA has a 3amp blade fuse. Pi is powered off battery bus with converter. The NMEA is powered off same bus. Both connect to main panel which is hooked to house battery. 

AI says it's a boot power drop that causes the Crystal to get out of oscillation. I'm not putting in the jumper wire it suggested. 

I'm thinking about wiping and reinstalling open plotter onto a clean SD card to see if that would work. 

If the M hat has an issue I did buy a waveshare can hat as a back up.

I am kind of thinking that having the nmea powered from the board to a bar to the back bone may be an issue but my NMEA runs solid as is.