I’m running a Feather MO to send data from a remote observatory to a cloud service every 60 seconds. After each successful "send" the WDT is patted. If a “success” code isn’t received for 2.5 minutes or so the WDT resets the Feather. This system runs perfectly for 3 weeks or so and then stops. The WDT is still blinking but the Feather has stopped. Pulling the plug and restarting from cold sets everything going again. What could cause this problem? Does triggering the RST pin not rest everything, SRAM, heap etc, I thought it did? Thanks Michael
Boy, your question is very much a Feather MO question than a WatchDog Board question. For some reason, the Feather is not being reset every time. You might want to turn the power on and off with a USB PowerControl board controlled by the WatchDog board. That should get it every time.
The Power Control board is a great idea so I’m trying it.
I set up the Power Control board (0042-112116-001) with Lipobatin tied high with 10K resistor, the ArduinoReset that used to go from the WDT to the Feather RST is now connected to the Grove Control pin and the Feather runs. But the Feather and WDT are operating at 3.3v and when I saw the note on your site about Controlling the USB PowerControl with a 3.3V Line, I added the diode, switched the control line but now the board no longer powers up.
Could you please provide a more detailed schematic of connecting a 3.3v Arduino style device being monitored by a WDT board to a power supply provided through the PowerControl board. Thanks Michael
OK. You have LipoBatIN tied to what? 3.3V? That's not high enough.
I'd leave it float and connect Pin one (control) of the Grove connector to 3.3Vand pin two (enable) to 3.3V. Make sure you have a common ground. If you aren't sure, then connect pin 4 (Ground) to ground on the board you have the control signals coming from.
Thanks for the ideas. I tied pin 3 and 4 to JP2 on the Power Control board with 10K resistors and connected Grove pin 1 (ground) to the matching pin on the WDT. Lipbatin is floating. Arduino Reset on the WDT is now only connected to TP3 on the Power Control board via a diode and it seems to be working. Does this seem like a working solution? The problem is that the error I had been experiencing, where everything locked up after a few weeks of continuous use, is hard to simulate and I've no way of knowing for sure until it happens and the Power control board forces a complete reboot! Thanks Michael