5

Take a typical CR2032 battery, if it is not discharging, what is the typical variation in voltage over a short period of time (eg a few seconds or minutes)?

Dave Tweed
  • 168,369
  • 17
  • 228
  • 393
John Smith
  • 249
  • 2
  • 8
  • Are you looking for what the error might be in the expected voltage? –  Feb 25 '19 at 22:52
  • 1
    It sounds like the OP literally wants to know how stable battery voltage is when open-circuited and sitting there doing nothing. Little pertubations and the like. What are you trying to do? – DKNguyen Feb 25 '19 at 22:53
  • 2
    Yup, I'm interested in the nearly open circuited changes. What I'm trying to do is use it as a voltage reference for an ADC which will be used to measure thermal effects of resistors. I have other voltage references but their noise is greater than the thermal effect of the resistor. I could use a part like the LTZ1000 which has 1.2µV peak-to-peak noise and 0.05ppm/°C, but it costs $100. For my usage, I don't care about the exact value of the voltage, only that it remains stable over a few minutes. – John Smith Feb 26 '19 at 00:03
  • 1
    for the sake of posterity, could you please change the question to use "noise" instead of "error"? the battery has no "error" relative to anything else, it's just a battery being a battery, there's only electrochemical noise. – Techydude Mar 09 '19 at 04:29

1 Answers1

9

Apparently you are asking about "ECN", Electrochemical Noise, which is the source of RMS noise from Li-Ion primary cells, and all other chemistries. Quick Google search and some browsing of search results leads to the following scholarly article, "A Method for Voltage Noise Measurement and Its Application to Primary Batteries".

They did several measurements, including CR2032 cells. The resulting noise is very small and needs special precautions to even measure it, a Faraday cage is a must, and uV-level special instrumentation was used to capture data. Here is one of their results, a typical waveform of noise, and its power spectrum:

enter image description here

By my eye-ball estimation, the RMS of noise of unloaded CR2032 cell is about 2-3 uV in 10^-2 to 5Hz range.

Ale..chenski
  • 38,845
  • 3
  • 38
  • 103
  • 2
    As the senior author of the manuscript, first, thanks for the plug :). Also, the black curve is really limited by the instrument, the battery noise is probably even lower than that unless the battery is disturbed in some way. – Burak Ulgut Dec 06 '19 at 12:52