I’ve 3 x WittyPi4 L3V7(latest firmware version 7) in combination with a pi zero 2w and USB Hat and a 18650 battery cell,
all 3 setups worked fine for random time periods within the Cell power capability but they randomly stop working and Ineed to toggle them back on. When I try to see in the log what’s happening I don’t read anything as a statement why the board was hard powered off.
On top of that I had to reflash the firmware again on 2 of the boards as the firmware “disappeared” as I only had the blinking LED working that a battery cell is connected to the board.
A) does anyone has a clou why the board is hard powering off the pi zero 2w. Or can I get better logging features to track what happens on the board level of the WittyPi?
B) why do I have boards that I need to reflash the firmware again them?
thanks
A) does anyone has a clou why the board is hard powering off the pi zero 2w. Or can I get better logging features to track what happens on the board level of the WittyPi?
The board is powered with 18650 battery, and it is possible the battery was running out and the voltage dropped too much to keep the MCU running. If your Pi was at ON state, the voltage could drop too fast to let the Pi actually log this event; If your Pi was at OFF state, it can not log this event at all (only Witty Pi 5 HAT+ can write log when Pi is off).
We call the time window for Raspberry Pi running as a "session". Each session starts with:
[xxxx-xx-xx xx:xx:xx] Witty Pi daemon (v4.21) is started.
and it ends with:
[20??-??-?? ??:??:??] Halting all processes and then shutdown Raspberry Pi...
Sometimes you may find some clues in the log. If the last working session is not complete, usually that was because the battery suddenly ran out and the whole device is off.
Also the battery voltage is printed in the log for every session, if the battery voltage in the last working session is already quite low (e.g. 3.3 or lower), there is very big chance that the battery ran out very quickly.
B) why do I have boards that I need to reflash the firmware again them?
Firmware should not disappear. You can see the LED blinking, that means the firmware is still there. Maybe it entered an unexpected state and it dosn't behave normally. It is hard to tell what happened without further diagnosing in such state.
Thanks for the feedback.
A) Is there a possibility to expand the logging feature, e.g. to the input statement? When reading the user manual I don see any minimum voltage input value hence we’ve set this to 2.9V-3V to shutdown the board.
but so far we don’t come to that point as the board “switch off” itself before it make a log on why it’s doing this.
So the logs aren’t helpful for troubleshooting and identifying the root cause. Even with 2 new 18650 (in parallel connected) cells this happens without a proper root cause.
B) isn’t it strange that this happens 2 boards on separate setups, at 2 different people?
on the forum there are also others that face the same and after reflashing the firmware, the board is again useful. It’s hard to explain this towards my customers, that this can happen.
For Witty Pi 4 (all models), logging is done by the software running on Raspberry Pi. If the Pi is off, there is no log.
Witty Pi 5 HAT+ can write log by itself, because it emulates a USB drive as its own data storage. Witty Pi 4 does not have these facilities. Witty Pi 4 also have used all IOs on its MCU (ATtiny841), and you can not output the log via UART interface (the pins are occupied already). It is possible to provide the log via I2C interface though, but that also need to modify the firmware and you will need to implement a device to read those logs.
As a product exists in the market for years, it is not strange that two users meet similar (not working) situation. We have tried to include as many scenarios as possible in our test cases, but the real world is complicated and there are always cases that can not be covered.
We can only fix a problem when it is understandable, or at least it can be reproduced on our side so we can eventually understand it.
You may think the device randomly stopped, but it did not -- there was always a reason. If the white LED was still blinking when you saw it stopped working, that means the MCU was powered and firmware was running at that moment. But it didn't say anything about what happened before: had the power supply ever disconnected? Did Witty Pi ever try to boot up your Pi when the scheduled time was due?
The users actually have the biggest chance to find out what happened, because they can monitor the voltage input to their device, they can record video on the device for long time, while we can not, and we can only see the logs.