Silicon Junction

I wanted a reliable way to measure the true capacity of any rechargeable battery, rather than relying solely on manufacturer specifications. The most accurate way to do this is to fully discharge a cell under controlled conditions and measure how much energy it can actually deliver.

To achieve that, I decided to design and build a simple, reusable battery capacity test circuit. The goal was a setup that could discharge a battery at a constant current, making the results repeatable and comparable across different battery types and brands.

To validate the circuit, I used a set of inexpensive rechargeable batteries I had recently purchased from Aldi as pilot test cells. Their low cost made them ideal candidates for initial testing without risking more expensive batteries.

Discharging a battery at a constant current requires active control, so I built a straightforward circuit using an LM358 op-amp to regulate the load current. This circuit allows the discharge current to remain stable as the battery voltage drops, enabling accurate capacity measurements.

Parts you’ll need

• TIP31C (AliExpress)
• Assorted Resistors (AliExpress | Amazon)
• LM358 (AliExpress)
• ADS1115 A/D (AliExpress)

Here the LM358 is used to compare a reference voltage (V2) against an measured voltage (R1):

• If voltage at R1 is greater than V2,  the LM358 output voltage will decrease.
• If voltage R1 is lower than V2, the LM358 output voltage will increase.

V2 can be provided by the center pin of a potentiometer across the rails, or your laboratory power supply, if you have one.

This setup will mean that 0.1V will be maintained across R1, from ohm’s law, we can calculate there must be 0.1A  of current flowing through R1 as well.

I connected VBat and IBat to a raspberry pi using an adafruit ADS1115 analogue to digital converter and logged the voltage and current every 30 seconds. 

Once I had gathered enough data I was able to produce a graph of the Activ Energy cell’s discharge curve:

Well, it looks like I didn’t get my full 2100mAh. It would be more accurate to rate this cell at 1600mAh. In this instance, it seems I got what I paid for.

Admittedly, this is only a sample size of one; so it’s possible that I was unlucky and received a bad cell. In the future I’ll test a cell from another batch, perhaps Aldi will be able to redeem themselves in that test.