Recently, I became curious about LiPo characteristics and quirks. What causes the drop-off in responsiveness over time? Why does power seem to fall off a cliff at the end of capacity? Why do throttle punch outs cause my flight controller to brown out when the battery is low? What makes one battery better than another, and is there a way to measure that and know for sure?
There’s a lot of conventional wisdom and theory out there that speaks to these questions. Battery ratings (capacity, C-rating, etc) attempt to standardize electrical/chemical properties (internal resistance, voltage sag, etc) but I feel they end up confusing matters more. For example, on most sites you’ll read something like “C-rating is the maximum current you can draw from the battery, but it’s really a meaningless number that you shouldn’t trust”. So, how are we supposed to know what a battery is capable of and how it compares to other batteries?
So, I decided to gather empirical data to try and learn more. I bought several batteries, armed myself with a voltmeter, ammeter, and several fixed resistance loads. I methodically discharged batteries from 100% charge down to empty (well, as close to empty as I could without damaging the battery) and recorded load/resting voltage and current draw along the way.
In this series, I’ll go through the data I collected and present findings. There’s a lot that lines up with the accepted theory, but there’s also a good deal that surprised me – either relationships that weren’t obvious or things that flat out contradict pieces of conventional wisdom. I’ll also draw direct applications to electric RC vehicles/motors where appropriate.
There’s lots of great LiPo 101 articles out there (how to charge, discharge, what a multi-cell pack is, so on). I’d suggest you take a moment and go read one or two of these until you feel you have a good working knowledge of LiPo basics. Here are two I recommend:
- A Guide to Understanding LiPo Batteries (rogershobbycenter.com)
- Understanding RC LiPo Batteries (rchelicopterfun.com)
Before proceeding, I also must highlight that LiPos are dangerous. A relatively volatile chemistry is used to achieve rapid discharge. This chemical makeup is prone to fires and explosions if mishandled or abused. Please don’t brush this off. The risk is real, and you must know what you’re doing before trying to use these batteries. Again, check out the links below for some information on proper handling and usage. If you still need conivncing, look for some videos about LiPo fires on YouTube.
Also, here’s a quick glossary of the more important terms I’ll be using in this study:
- Load: Anything that draws a current out of the battery.
- Load Voltage: The voltage across the battery terminals with a connected load.
- Resting Voltage: The voltage across the battery terminals without a connected load.
- Internal Resistance (or “IR”): Resistance can be thought of as an electrical ‘force’ that opposes the flow of current. Internal resistance refers to current-opposing force that occurs within the battery itself. IR is arguably the most important (and most complicated) characteristic of a LiPo cell. As such, I’ll be dedicating an entire post to study and discuss IR.
- State of Charge (SOC): Remaining capacity in a battery. A battery at 100% SOC is full, and 0% SOC is empty.
Tools & Setup
I chose two different LiPo batteries for this study: A 300mAh Turnigy single cell rated at 20-40C, and another 300mAh Turnigy single cell rated at 45-90C. This choice was extremely intentional: The batteries have the same capacity, have the same manufacturer, are from the same product family (nano-tech), and have the same connector type (connectors affect resistance). By keeping all of these factors constant, I hoped to be able to extract some useful relationship between the printed C rating and the actual battery performance. I chose a single-cell battery just to keep the experiment simpler (and safer) and not have to worry about parallel connections between multiple cells. I also bought several of each of these batteries to test for consistency. I’ll be referring to them as by C-rating and a number I assigned to each pack (20C #1, 20C #2, 45C #1, etc).
I also wanted to test the batteries at a variety of fixed loads. In order to get a reasonably high current draw, I used a variety of low-resistance coils of copper wire. The approximate resistances used, the expected current draws (at 3.7V – this is the voltage the industry has settled on to discuss the nominal voltage of a LiPo cell), and the corresponding C rating of that current draw are as follows:
|Load Resistance||Expected Current Draw at Nominal Voltage (3.7 V)||C Rating of Expected Current Draw (assuming 300mAh capacity)|
|0.5 Ohm||7.4 Amps||24.67 C|
|0.9 Ohm||4.11 Amps||13.7 C|
|1.3 Ohm||2.85 Amps||9.48 C|
|1.9 Ohm||1.95 Amps||6.49 C|
|3.7 Ohm||1 Amp||3.33 C|
I used a pretty basic multimeter to measure current, and a 5v Arduino’s ADC to measure voltage. I chose the Arduino partially because I didn’t have a second multimeter, but also because I could apply an offset in software to calibrate the ADC against the multimeter. These tools definitely aren’t the state of the art when it comes to precision, but they’re good enough to collect some useful information. The precision matters the most when it comes to measuring internal resistance, so I’ll spend some time on precision when I discuss IR.
The data collection took some trial and error to get right, but in the end it’s pretty straightforward:
- Start with a fully charged LiPo.
- Apply a fixed load through an ammeter, with a voltmeter attached to the battery leads.
- Wait a fixed period of time (with circuit connected).
- I ended up using 20 seconds. I’ll explain why in a future post.
- Record load voltage and current under load.
- Disconnect circuit immediately.
- Wait for another fixed period of time (with circuit disconnected).
- I ended up using 180 seconds. I’ll explain why in a future post.
- Record the resting voltage.
- Repeat steps 2 through 7 until the battery is depleted.
I repeated this process on several different batteries with several different load resistances. It ended up taking 1-2 hours per battery per load, so I didn’t test every possible combination. Instead, I focused on getting a representative set of battery/load combinations.
In the part 2 of the series, I’ll dive in to the results. I’ll look at the basic characteristics of a LiPo cell and how they might differ from battery to battery. I’ll also discuss the implication of all of this is on your RC models.
If you’ve got any questions or comments, please leave them below!