simulate this circuit – Schematic created using CircuitLab

I'm measuring current consumption of several different handheld devices (calculators, flashlights, etc), at the [lithium] battery. Using a simple Uni-T digital multimeter set to mA, I'm measuring current separately between '+', and separately between '-'. I see some devices which show slightly different readings between '+' and between '-'.

One no-name calculator, for example. When in use, I'm seeing a small difference of -53.2mA at '+' vs -52.8 mA at '-'. After about 5 minutes of inactivity, it automatically turns the display off, and for almost exactly 120 seconds measures -1.57mA drain when measured between '+', and -0.08mA drain when measured between '-' of the battery, then both readings drop to about 0.0 mA. I am assuming it goes to some sort of a deeper sleep mode. But what is the most likely explanation for the discrepancy in measurements?

  • 3
    \$\begingroup\$ A schematic is needed to clear the ambiguity of your words. \$\endgroup\$
    – Andy aka
    Commented Dec 8, 2020 at 14:30
  • \$\begingroup\$ So, you put your ammeter in series with the batteries, only one time at the + pole, and the other time at the - pole? Are you using the same instrument at different times? Did you retake measurements several times to confirm the discrepancy is consistent? Does the current always flow in the same sense through the instrument? \$\endgroup\$ Commented Dec 8, 2020 at 15:11
  • \$\begingroup\$ (1.) The same instrument (2.) Repeated 5 times or more. (3.) Some devices include charging, then the current is positive. Otherwise, the battery is draining (the current is negative). (4.) The measurement is done by splitting the + wire of the battery in half, and putting the multimeter in-between, while - remains connected straight to the device. Then visa-versa. The measurements of + and - are not done at the same time. \$\endgroup\$
    – wolf9000
    Commented Dec 8, 2020 at 15:12
  • \$\begingroup\$ I do not have the schematics on the device side since the devices are just a bunch of random end user devices I have lying around. I am asking whether any blackbox device can have this behavior ([protection] diodes [leakage], etc) when battery drain current is being measured, or is there an e.g. fundamental difference between measuring current via + or - on the battery. My assumption, and everything that I have read so far, says that there should not be any difference. It seems to be device-specific. On most devices, the current is equal. \$\endgroup\$
    – wolf9000
    Commented Dec 8, 2020 at 15:21
  • \$\begingroup\$ But does the current always flow in the same direction in the ammeter when you measure at the + and at the - pole? I mean, have you tried inverting the ammeter's terminals when you are measuring current at the same pole? Same measurement? \$\endgroup\$ Commented Dec 8, 2020 at 15:31

2 Answers 2


In theory, the results should be exactly the same.

But there are a host of confounding issues, such as:-

  • The state of charge of the battery
  • The temperature
  • The contact resistance at the battery and ammeter terminals
  • What the calculator is doing at the time (maybe even if there's something in memory)
  • \$\begingroup\$ He's repeated measures several times and he's got consistent results, so charge, temperature and operational status are not a factor. I had thought of contact resistance, for example an oxided battery holder contact, but he's cut the wires so those contact are not changed by the measure... \$\endgroup\$ Commented Dec 8, 2020 at 20:35
  • \$\begingroup\$ ...unless the device is upside down and the keys gets pressed in different points when the measurement point is changed (all of this assuming the solar cell was disconnected, otherwise shadows come into play). \$\endgroup\$ Commented Dec 8, 2020 at 20:45

measuring in a forward direction can be different from the value when measuring in the backward direction due to a zero-adjustment error.

ever tried swapping the connecting terminals around to make the "+" a "-" and vice versa?

PS: the typical error for a not-that-perfect calibrated device might be in the 1-2% plus some 1-2 digit range. just that your display can show you that much more digits does not mean the absolute value is more accurate. (but doing relative measurments or tendency observations might still work - unless you have some warm up drift in the DMM. often its up to you to decide if your measuring approach is solid.)

  • 1
    \$\begingroup\$ It is interesting that out of almost a dozen various devices, currently actually only one shows this behavior, and quite consistently. \$\endgroup\$
    – wolf9000
    Commented Dec 9, 2020 at 2:42
  • 1
    \$\begingroup\$ @wolf9000 can you take the batteries out of the equation and supply the device with an external voltage supply? Are we talking button batteries - as I would expect in a calculator - or sealed lithium batteries? Because if we were to believe the readings, maybe there's a leakage that introduces a 'third' terminal between the + and - (and the circuit will close through the pcb or metallic chassis). I doubt it, but just to rule out something is happening between the poles... \$\endgroup\$ Commented Dec 9, 2020 at 3:47
  • \$\begingroup\$ there might be the chance that having some wires close to certain circuits and parts an effect is created that impacts on power consumption of the device. coil based amperemeters would definitely have a magnetic field. and such fields can impact on e.g. feritic beads as well as on air coils and on transformers. - reminds me of the teacher that once had built an oscillator and it ran only when he attached a measurement device. in the end he used a resistor of similar value as the meter - and his oscillator continued to work. - thats the impact of doing measurments. \$\endgroup\$ Commented Dec 10, 2020 at 0:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.