Battery charger on v1.0 and v1.1 drops vcc below 3v, breaks sensors unless battery connected

[lars-t-hansen]

It was working well for several days but now it provides a few readings and then stops working, with CO2=400 (base level) and TVOC=0. This is weird.

[lars-t-hansen]

This is still broken. I thought I fixed it by tweaking the initialization sequence but it's back to the way it was.

[lars-t-hansen]

There is some kind of connection with running on USB power (whether USB battery or other source), TBD. On some of the devices at least, the air sensor seems to work OK when the device is only on true battery power.

[lars-t-hansen]

I would say that on all the v1.1 devices, starting up on battery power alone fixes the issue.

If the device starts up on USB power then adding battery power does not make the problem go away (maybe not surprising; nothing is done to re-initialize the device in that situation).

If the device starts up on battery power then adding USB power does not destroy the readings, they remain OK.

It could look like the AIR sensor is not being initialized / set up properly when only USB power is connected, and of course since we do nothing to initialize it later if battery power is connected then it remains uninitialized.

Whether this is because of some stray current from USB that disturbs the sensor, or if it is because the sensor is not powered properly when there is only USB power, is unknown.

[lars-t-hansen]

In contrast, the air sensor on the v1.0.0 device, which used to work, seems busted in all configurations now.

[lars-t-hansen]

Gunnar says "dirty power due to battery charger circuit, which pulls the on-board power below 3V fairly frequently, sometimes below 2.7V". HW 1.2.0 will fix this.

[lars-t-hansen]

The workaround is to use the battery, and that's the only thing that will help on HW 1.1.