T114 (MeshCore repeater) quits solar charging

Hello,

I am having an issue where my T114 setup with solar and LiPo batteries will begin charging, but then after some time seems to quit solar charging before full. I’m running the latest (v1.14.1) MeshCore repeater FW. Here is what I am experiencing:

T114 connected to 3x 4900mAh 21650 batteries in parallel via on-board battery connector. 5W solar panel connected to the on-board solar connector.

  1. Physically reboot the T114 with pushbutton.
  2. Verify I can reach it from a MeshCore companion. Battery reports 30% charge and multimeter readout of battery confirms roughly 3.5V charge.
  3. Orange charging LED is lit and the panel is in indirect sunlight. Verified with multimeter that voltage is 5.2V and charging current is 0.5A (reasonable given not direct sun).
  4. Leave this setup all day in sunlight (varying cloudy/indirect sun/direct sun). Over 8 hours of daylight.
  5. Similar or better sunlight, open the enclosure and the orange LED is now off and the unit is not charging. No meaningful charge has been added to the batteries. Slight overall lost charge during the day.
  6. The device is still on and reachable via MeshCore companion.
  7. I can initiate solar charging again by manual power cycle, which returns us to step 1.

I would expect solar charging to continue indefinitely without intervention, but it seems like after some time it is disabled for an unknown reason. Does anyone know what is happening here?

Thank you in advance.

Yes, there is no solar charger on the board, only a LiPo charger that won’t cope too well with variations in input and is known to go in to shutdown/protection mode when used like this.

You need to put something that can deal with the solar panel volts vs amps issues - a board between your panel & the “solar” input.

Something based on a CN3065 may be able to do the job with all those batteries. But I’d go for an Adafruit BQ solar charger - or even one of the older MCP73871 based boards - loads around at a very reasonable price.

Thank you for the help. I assume the MCP73871 does not require a minimum battery charge to operate because it takes its power from the solar input. I also assume this means that, from a dead battery (T114 doesn’t power on) the MCP73871 should be able to solar charge the battery back to a sufficient voltage to power on the T114. Is this your experience?

My goal is to have a system that operates only on solar without physical intervention, which would require recovery after, say, a month of insufficient sun for continuous operation.

My experience is never assume, read the docs. But there is some underlying expectations when using a well exercised design from Adafruit implementing a chip from Microchip, a very experienced company, that has been copied a lot by the Far Eastern companies that good things can happen. There is also the investment issue - you try something that costs $5, it doesn’t work, your time was worth more.

As it happens, yes, the chip will be powered by the solar panel when the panel is providing sufficient power, it will protect the battery from being over discharged or over charged and it will use the solar over the battery to the output.

However the whole brown-out, low power reset with the T114 module is a bit of an unknown and the track record of the engineering on the MeshSolar is pretty dismal which appears to have exactly those sorts of problems.

An additional power-reset-voltage-detection board could be added to make this hardware dependent but in theory something could be done with the firmware programming.

The bottom line is to try it, because reality is everything.

1 Like

PS, you’ll have to connect the LiPo to the solar board, not the T114 and pipe the output from the board to the T114 VDD_5V edge.

If you need a power detect reset, I’d go with a TPS3710

1 Like