# Can a battery's open circuit voltage be measured without actually disconnecting the load?

Is there:

• a specific tool for it or
• a method for its calculation which includes the terminal voltage I'm actually in need of a method which would not require disrupting the circuit by removing the load for this calculation.
• Are you able to measure the drawn current, too? Sep 8 at 5:55
• If you do not disconnect all the loads then what you measure will not be the open-circuit voltage. Sep 8 at 6:02
• Do you know the load and can it be adjusted? I think if you can define 2 different load values, and measure the voltage across load at those 2 times, you can back calculate to get the OCV. Sep 8 at 6:08
• Here is a paper about a model of a lead acid battery: semanticscholar.org/paper/… Sep 8 at 7:10
• Here is a paper about a model of a lithium ion battery: intechopen.com/chapters/78501 Sep 8 at 7:13

Can a battery's open circuit voltage be measured without actually disconnecting the load?

It depends on what you mean by "disconnecting the load".

For instance, if you can temporarily provide load current from a charged capacitor, you can open circuit the battery and measure its voltage in a matter of a few milliseconds or less. This just needs a fairly trivial MOSFET switch and simple control circuit.

Over the small time period that the battery is disconnected from the load, there needs to be sufficient capacitance to "hold-up" the load's supply voltage. If that is acceptable then yes, it can be done.

So, if by not "disconnecting the load" you mean keeping the load powered then no problem. The capacitance value needs to be worked out based on load current and supply voltage range but, given that the normal supply is a battery (which will droop), you have a fair amount of flexibility in optimizing the capacitor value.

As an example, a 10,000 μF capacitor's terminal voltage will "droop" by only 100 mV when supplying a load current of 10 amps for 1 millisecond.

Here's the equivalent circuit of a battery-fed system: simulate this circuit – Schematic created using CircuitLab

Without a load (or with a load which is high enough that $$\R_S\$$ can be neglected e.g. the internal resistance of a voltmeter), the voltage across A and B nodes will be equal to $$\V_{oc}\$$, open circuit voltage.

$$V_{oc}=I_L R_S+V_L$$

$$\I_L\$$ and $$\V_L\$$ can be measured easily. But if $$\R_S\$$ is an unknown then no, the open circuit voltage can't be measured without disconnecting the load. Plus, even if $$\R_S\$$ is known, it won't stay constant as it changes with temperature and, for some types and chemicals, increases as the battery discharges. So the battery "must" stay at the same conditions when loaded.

If the load current is so low that the drop across the internal resistance can be neglected and the battery temperature don't rise then the voltage across the load can be assumed as the open circuit voltage.

• This is the "simple model." Often it may be good enough. But it is definitely not an accurate model and does not explain voltage recovery that can easily be observed on any battery over a period of minutes after removing a load. Sep 8 at 7:16
• @mkeith true. None of the battery models represent chemistry- and technology-related properties, or dynamic behaviour. Often some designers assume the Rs as a fixed resistor but it's a combination of fixed resistors, thermistor(s) and some other types of resistors. For the sake of simplicity in the answer, I didn't take lots of details into account. Thanks for your comment. Sep 8 at 10:52
• Some functional battery models are designed to mimic battery behavior very accurately for monitoring state of charge. But they are generally more complex models. You can see the links in the comment section. I do not know if these models are used by battery fuel gauges and such. But it seems possible that they might be. Sep 9 at 4:26

Yes, it is possible to measure the unloaded voltage output of a battery while that battery is supplying a load, IF one can accurately model the battery as having an internal resistance, and the load as being a resistance as well.

To do this, one puts an AC exciting signal across the battery terminals. It can be small (and harmless to the load, one hopes), but an AC voltage and current measurement will tell you the parallel combination of the internal resistance, and of the load resistance, because an 'unloaded battery' has no AC voltage drop.

If you have a way to know the load resistance, such as voltage and current measurements before attaching the AC test signal, you can solve for the internal resistance and for the battery's unloaded terminal voltage, using the AC and DC measurements to give two equations and solve for those two unknowns.

Low-frequency AC is best, because stray conductances due to capacitance, or impedance due to inductance, is minimal at low frequency.