Measuring current on a low power IOT device

[Maxlor]

This is what I mean:



Yes, the batteries made for this purpose work perfectly well at high discharge rates, but there's a substantial voltage drop which is much larger than the voltage dropped caused by the battery running empty. Also, in measurements I made for another project, I found that a LiPo takes a moment to recover from a high current peak to where the voltage should be for the lighter load.

Note also that a battery's internal resistance is not constant but changes with temperature (so indirectly with load) and state of charge.

This makes me think that there's going to be a lot of noise and error when trying to use the battery as an integrator when measuring very variable power consumption.

[mzzj]

I wonder if supercaps might work in this scenario though. They also have high internal resistance (maybe use several in parallel), but their logarithmic discharge curve makes it easier to accurately measure an energy delta.

There is plenty of super/ultracapacitors that have low enough internal resistance. Cooper-bussman Powerstor, maxwell ultracapacitors etc etc.
Problem is that they have some leakage current, 1F ultracap might specify 10uA leakage _after_ 72h applied voltage. Plentifull voltage derating should help with the leakage, so connect 3 2.7v ultracaps in series for 4.2v li-ion circuit.
http://www.edn.com/design/components-and-packaging/4374932/2/Coupling-a-supercapacitor-with-a-small-energy-harvesting-source-

[electronic_eel]

This is an example how you could accurately measure low power modes and short spikes during transmissions etc.:

http://www.dorkbotpdx.org/blog/paul/measuring_microamps_milliamps_at_3_mhz_bandwidth

[Rufus]

There can be 400mA spikes that last 4.7ms.

0.4A for 4.7ms will cause a 10,000 uF capacitor to droop by about 190mV. If a pulse like that happens once an hour it is an average of about 0.5uA - possibly insignificant.

Like I already said power the thing from a big capacitor. Feed the capacitor with a constant current source which needn't be more complicated than maybe a 100k resistor and a bench supply that can go up to 30v. Monitor the capacitor voltage and adjust the current source to maintain 3.7v or whatever you think your average battery voltage is. Whatever you have to adjust the current source to is your average current consumption at 3.7 or whatever volts. Remember to consider leakage from the capacitor and what you are monitoring it with.

I have a quality but ancient (date code looks like 1977) 46,000uF 20v capacitor in my junk box. I reformed it by sticking 20v on it for a couple of hours then measured its leakage at 3.7v. The leakage was -ve because of dielectric absorption and/or perhaps some chemical battery effect. After about 4 hours at 3.7v the leakage is holding at around -1uA - weird.

[Maxlor]

0.4A for 4.7ms will cause a 10,000 uF capacitor to droop by about 190mV. If a pulse like that happens once an hour it is an average of about 0.5uA - possibly insignificant.

I'm afraid not. That's the worst case specified pulse that happens during WiFi communication when reception is bad. It happens many times, I'm not sure how often that pulse would appear, I'm guessing every few dozen milliseconds. Lets say, 30 per second, and the WiFi is up for a few seconds at a time.

I'm going to need a larger capacitor

I like the idea though, using a quite large capacitor as an averaging or integrating element sounds like it's going to be more accurate (for measuring overall consumption, not instant current.)

[Rufus]

I'm afraid not. That's the worst case specified pulse that happens during WiFi communication when reception is bad. It happens many times, I'm not sure how often that pulse would appear, I'm guessing every few dozen milliseconds. Lets say, 30 per second, and the WiFi is up for a few seconds at a time.

Then you are kinda screwed trying to accurately predict battery life because the consumption isn't predictable.

[leppie]

I am in the same boat

I was thinking of perhaps trying a coulomb counter.

I had good results with ltc2943.

Cool  Will have a look.

[doctormord]

I'm afraid not. That's the worst case specified pulse that happens during WiFi communication when reception is bad. It happens many times, I'm not sure how often that pulse would appear, I'm guessing every few dozen milliseconds. Lets say, 30 per second, and the WiFi is up for a few seconds at a time.

Then you are kinda screwed trying to accurately predict battery life because the consumption isn't predictable.

It is, with the solution i posted.

[EEEnthusiast]

Not sure if someone is still looking at this thread. You may want to consider a power profiler tool like Qoitech Otii or the ZS-2102-A from ZSCircuits. There is another one called as JouleScope made for the same purpose. High dynamic range, high bandwidth current measurements.

[fcb]

The Otii is really slow (<5KHz sample rate), expensive and they want alot for the SW.

[thm_w]

Not sure if someone is still looking at this thread. You may want to consider a power profiler tool like Qoitech Otii or the ZS-2102-A from ZSCircuits. There is another one called as JouleScope made for the same purpose. High dynamic range, high bandwidth current measurements.

You posted in another thread but forgot to note it here: "There is a similar product already out on the market. ZS-2102-A from ZSCircuits. (Disclaimer : I'm the designer for this product)"

Why not instead make a thread to discuss your product. But be careful not to be a pure advertisement, otherwise people may not like it.

[EEEnthusiast]

Not sure if someone is still looking at this thread. You may want to consider a power profiler tool like Qoitech Otii or the ZS-2102-A from ZSCircuits. There is another one called as JouleScope made for the same purpose. High dynamic range, high bandwidth current measurements.

You posted in another thread but forgot to note it here: "There is a similar product already out on the market. ZS-2102-A from ZSCircuits. (Disclaimer : I'm the designer for this product)"

Why not instead make a thread to discuss your product. But be careful not to be a pure advertisement, otherwise people may not like it.

I have opened a new thread for this. Thanks
https://www.eevblog.com/forum/testgear/iot-power-profiler-and-battery-life-estimation/