ESP8266 Community Forum

Hi, I'm powering a full breadboard with the NodeMCU devkit board, display, etc and the LiPo battery works fine when I connect it directly to VIN and GND of this ESP12E board. I'd like to measure actual draw next to evaluate effectiveness of a deep sleep and estimate battery lifetime. However, the esp12e doesn't come up, only a faint blinking led.

To confirm - broke the VCC line from the battery to serially plug in the ammeter. It does show around 140mA draw, but the board never proceeds to boot with the meter inline

The battery is good, measure it at 3.88V still.

seems like you're wiring it properly, battery+ to multimeter (red), multimeter (black) to vcc breadboard rail, so the current flows from the battery, serially through the multimeter, then battery- going to gnd breadboard rail.

have you tried it with other power supplies - like 2-3 AA's or a 18650? maybe the lipo doesn't like the interruption?

i must say i've not had much luck with multimeters and esp8266's, seem to have fried a few for no apparent reason.

I've had trouble with the ESP booting while a multimeter is connected and think i remember the 140mA was one of the symptoms. From what i could tell its from the multimeter not letting quite enough current flow for the initial burst as the ESP starts - a lot of cheaper multimeters have 200mA as the top of one of the ranges and thats not quite enough. I found switching to the highest current setting on the multimeter if you have one higher than 200mA, or just shorting the probes together for startup and then separating them after its had the initial inrush of current can work.

If you hook a multimeter in series with the supply to measure current, you are basically putting a resistor in series. This results in a voltage drop across the multimeter ("burden voltage"). Especially cheap multimeter can have a rather high burden voltage, which could cause the ESP to brown out.
The burden voltage will decrease on the higher measurement ranges (since the multimeter will use a lower value shunt resistor), but of course you will loose precision.
There are solutions to this problem, like the µCurrent adapter or a multimeter with lower burden voltage (the Mooshimeter advertises an exceptionally low burden voltage for example). Those are not cheap however, so unless you really need a better solution, I would recommend using the workarounds suggested by torntrousers.

PS: the max. supply voltage of the ESP is 3.6V, so I would not recommend running it of an unregulated LiPo.