• Made myself a Bangle JS Hardware Exploration Work Bench... ;-), where I added some wires to some power related points in the circuitry:

    • Battery + (B+) - 3.7V nominal LIPO
    • Main + (+) - 3.3V nominal LDO Reg output
    • Status (Pin 1 of 65b924) - neg / low when charging
    • Piezo Speaker (which is actually + / 3.3V nominal)
    • Piezo Speaker + (which is steady 3.3V nominal when off)

    With a Pico connected to the broken out signals I can monitor with a small JS program what is going on, displaying a live graph and even log the values in a file, latter two items with @JumJum 's Espruino built-in Testing feature (a next step is to use an Espruino BLE device such as PixlJS and talk to an agent on BangleJS to control BangleJS 'from the inside'.

    To be able to measure the full range of possible voltages, voltages are measured with a resistor voltage divider with two equal resistors. I'd have liked to use 1M resistors, but had only a bunch of 820K ones... I measured them to pair them up and with my cheap Multi Meter I got two pairs with the same K... 8123+812 and 816+816; with that the error rate is +-1K on 800+K (or +- 2K on 1600K) which is at worst < 0.25%. The high resistor values show when the meter is co-connected or not connected when calibrating (getting the values for uBattF and uMainF factors). Very strictly and seriously speaking and calculating, I would have to take into account the inner resistance / load of the Multi Meter and the Pico Pins and the additional load of the voltage dividers... and solidly solder all connections and use high-conductive wires.

    The stills and clips show 1 value taken every second, which is of course a bit too fine time grid. Even every minute is still a lot of data. I guess taking a value every half an hour and may be even an hour is good enough.

    To have a decent change of the lines in the graph, the retrieved values are adjusted by rebasing - -3V - and multiplying by 1000, which then shows around 320 for uMain of 3.3V and around 1200 for uBattery of 4.2V. Charging/not charging(anymore) is scaled into as lines with level of 0 and 100 and 0 again.


    4 Attachments

  • Looking for some suggestions what to have installed for a minimum to establish first Battery test base line.

  • Wow, this looks great! I'd say the first basic test would be just 'Install default apps' from the App Loader.

    The pedometer and weather widgets were mentioned on http://forum.espruino.com/conversations/­350267/ so could be worth a try I guess...

  • Make a setInterval loop that calls Date.now(), keep the cpu busy. See how it compares.

  • Post a reply
    • Bold
    • Italics
    • Link
    • Image
    • List
    • Quote
    • code
    • Preview
About

Bangle JS Battery Voltage, Main Voltage, and Charging...

Posted by Avatar for allObjects @allObjects

Actions