Powering a GSM modem with alkaline batteries


I am working on a open-source / open-hardware low-cost buoy for measuring ocean drift and waves close to the shore: GitHub - gauteh/sfy: A low-cost drifting wave buoy for near-shore deployments. for S&R and research. It uses a cellular modem for this, the blues notecard.

I would really like to stay on alkaline batteries since mixing lithium with water is not good, and these buoys will be used close to shore.

I was hoping I could use 6 D-cells (9V) in series with the S13V30F5. The problem is that the modem draws up to 2A when connecting to GSM, and this seems to be problematic once the batteries are not completely fresh. The MCU and IMU uses on the order of mA of current, so the 3A limit should be find.

I have also been testing the S7V8F3 to power the MCU host with 3V3 and then put 2x3 D-cells at 4.5V, with no extra regulator before the modem.

Both these solutions do not last as long as hoped before it seems that I experience voltage drops that causes all kinds of issues. In theory there should be enough energy for at least a small week (depending on how often I transmit and connect). Do you have any suggestions for how I could power this project using alkaline batteries? Are there any Pololu products that could be used? If there was a break-out for a charging super-cap circuit (say 5V, 5F) that could probably help a lot, then I could charge the caps before initiating the connection. Any thoughts or ideas hugely appreciated!

Best regards,

Hello, personally I always power up GSM modules with 5V,2A adapter only. Here are some guidelines about powering them up with batteries.

Thanks. It looks like NiMH batteries might be the best option if Lithium is ruled out. Will try with a super-cap + alkaline first.

Lipo battery pack+ voltage regulator can be a good option too. Like this project. This man made a GSM enabled purse. He has kept provision for a lipo pack in his PCB.

Thanks. The point is to avoid Lithium.

Hi, Gaute.

The voltage and charge on alkaline batteries drops pretty fast with high current draws, so I suspect the number or type is just not enough for your application. Monitoring the battery voltage while the system operates could help you understand whether that is the case and how much more power you need. I am not too familiar with the considerations for electronics that could become submerged, but maybe there is another battery type that would work, such as NiMH or lead-acid (which can be more suited to high current demands). Otherwise, you could try adding even more alkalines. If your input voltage or current requirements change, I would be happy to recommend a different regulator.