Over the past couple of months several current and former Electrathon have helped me get started on my first vehicle. One way I thought that I could repay the generosity is to share my experiences, both good and bad, as I try to build an Lithium-ion battery pack.

My initial premise is that electrathon is very cool. It is 'simple' enough that a team of high school students can build and test a competitive vehicle in a year for under a $1000 dollars. Yet, it is complicated enough the final product can propel a student around a short track for 60 minutes at 30 miles per hours. The project could be even more fun if the weight could be reduced by 50 lbs.

However, I understand there are many challenges with racing Li-io batteries:

1.Safety. At the end of the day, this is a student project. If the teachers and race stewards don't feel safe the batteries are not ready to be used at their events.

2. Fairness. There has been an ongoing challenge how to race and score various vehicles with different battery characteristics.

3. Price. With any new technology it can quickly become an arms race to buy a victory. That seems contrary to the principle vision behind electron.

To be honest, I am going to ignore the idea of fairness for now. Once a few team have started running Li-io, the stewards can visit how to score them. In this thread, we will follow the rules but not get involved in setting them. Li-io is fairly safe, as long as one follows a few simple rules, you are not going to have problems. Hopefully we can use this thread to prove that:) Finally, with regard to price. The budget for this build is $100 of consumable per vehicle and $500 of tools per team.

Next week, we can look at design parameters:)

My primary challenge when researching batteries was the inconsistency of available documentation. The most interesting was confusion caused by switching between metric and english systems of measurement. Electrathonamerica references vehicle speed in MPH and battery weight in lbs while volts and watts are SI units. It is a minor things but, it trips up allot of people. The second dichotomy is the use case for various batteries. What works for a manufacturer of electric vehicles is not necessarily the same as what works for RC hobbyist and vice versa. Electrathon vehicles are somewhere in the middle.

Lets start by defining some terms and units.

Volt -- We need to be familiar with the idea of a volt as the electric potential between two points of an electrical system. Using the water flow analogy, voltage is the difference in water pressure between two points.

Traditional electrathon vehicles use 24 volts because it is easy to create 24 volts by wiring two 12 volt batteries in series. Voltage can be a bit confusing because it is is not always clear if someone is talking about nominal or actual voltage. Nominally, lead-acid batteries are considered 12 volts. Actually a charge lead-acid battery is about 13.2 volts. while a discharge lead-acid battery is about 10.5 volts.

Amp -- From a practical perspective an amp can be thought of as the rate at which charged particles flowing through a point. Using the water flow analogy, amperage is the rate of flow through the system.

Most electrathon vehicles are set up average about 40 amps. While under heavy acceleration, amperage can go up to 100 or more amp. While braking or sitting in the pits amperage can go down to near zero.

Watt -- A watt is a derived unit of power. watts = Volts * Amps.

Most electrathon vehicles use power at a rate of 24 volts * 40 amps = 960 Watts.  

KWh  -- A kilowatt Hour or kWh is a derived unit of energy. Which calculated by multiplying power in watts by time in hours.

This is important because a rough estimate of the two lead acid batteries commonly in use can hold about 1 kWh of power.

Gram -- I would like to stress the use of grams as our unit for measuring mass for this excercise. Even though electrathon rules talk about battery weight in pounds, we will use grams in this thread to reduce possible confusion caused by converting pound to kilograms on the fly.

Current regulations state that vehicles can use:

Lead-Acid 73 lbs. or 33.1 kg

Nickel-Iron 58 lb. or 26.3 kg

Nickel-Metal-Hydride 41 lb. or .

Nickle-Zinc 44 lb.

Lithium-Iron-Phosphate 29lb.

Silver-Zinc 23 lb.

Lithium-Ion 15 lb.

Lithium-Polymer 15 lb. or 6.8 kg

While appearing somewhat arbitrary, my understanding is that these rules were design to create a level playing field such that each type of battery would hold about one kWh of power.

The brings us to our primary design specification. 1. Build battery pack which provides one kWh of energy which weighs less than the allowed weight for a particularly chemistry.

In the last post we looked at some common terms and units for calculating the capacity of a battery. The biggest takeaway is that in order to be competitive we need a battery with an energy capacity of about one kWh.

We need to acknowledge that the winning team are getting more than one kwh out of their batteries. However, we are using one kWh as a target to keep things simple. If we focus too early on making the most effective battery possible it is very easy to get into the common trap of premature optimization.

The initial step will be to verify that it is possible to meet our mass constraint for different chemistries based on their energy densities.




Original image from batteryuniversity.com

We can start by looking at lead-acid batteries. In theory a 33.1kg battery should be able to hold

33.1kg * 40 Wh/kg = 1324 Wh

This is a bit higher than our target of 1 kWh. But as we will see, actual capacity is often lower than theoretical capacity. Later, we will look at some the the inefficiencies which reduce batteries to below their theoretical capacity. For now, it is enough to say that in theory is it possible to create a lead-acid battery which meets our initial design specification.

As a second example we can look at Cobalt Li-ion. In theory a 6.8kg battery should be able to hold

6.8kg * 165 Wh/kg = 1122 Wh

This shows that we can meet our initial design specification with Cobalt Li-ion.

For further reading, there is allot of information on energy density on the internet:)

So far, we have looked at some pretty hand wavy theory. There are many variable which can alter a batteries capacity. Once we use theory to get us in the ballpark, we must test to verify what our battery an actually discharge during a race or training session.

This post we are going to look at something a bit more concrete, how individual cells are configured in the battery packs. The 12 volt lead acid batteries commonly used in electrathon vehicles are made up of 6 cells arranged in series.

Orginal image from http://www.boundless.com

Putting two 12 volt lead acid batteries in series gives us a 24 volt pack.'

Just a word of caution about actual and nominal voltages:

Nominal voltage for a lead acid cell is 2 volts.

Voltage for a fully charged lead-acid cell is 2.1 volts.

Voltage for a fully discharged lead-acid cell is about 1.95 volts.

Lead-acid is well understood. Manufactures of other batteries can rate their batteries differently so be sure to test.

Back to series. If we are going to assemble a battery pack, we are going to have to arrange the right number of cells in series to get the proper output voltage. For example:

Nominal voltage for a li-ion cell is 3.7 volts.

Voltage for a fully charged li-ion cell is 4.2 volts.

Voltage for a fully discharged li-ion cell is about 2.4 volts.

If we were looking for system that could output a nominal 24 volts we could chose six cells in series.

6 cells * 3.7 volts = 22.2 volts

A couple of words of caution. Most accidents with batteries are the result of overcharging, undercharging, or rapidly discharging individual cells. When testing be careful to not overcharge, undercharge, or discharge your batteries faster than they are designed to handle.

Most high density battery packs will require some type of BMS (Battery Management System) to ensure the battery is operating within desired parameters.

The next post is going to continue on the theme of safety by looking at the discharge rates of various battery chemistry and how to prevent discharging at an excessive rate.

In the last post we looked at how a battery pack is made up of one or more cells arranged in series to create the desired output voltage. We will need to take steps to ensure that overvoltage or undervoltage situations are never met for the entire battery.

Our second category of safety precautions is not charging or discharging a cell too rapidly. All batteries have a C-Rating, formally known as its Coulomb Capacity. For practical purposes we can think of the C-rate as a factor which indicates how rapidly a battery can be discharged relative to its maximum capacity.

A 2C battery can be discharged in .5 hours. A 1C battery can be fully discharged in 1 hour. A .5C battery can be discharged in 2 hours.

ThreeAs part of these experiments, we are going to take precautions to monitor and shut down the battery in the case of excessive current. Each cell will have a fuses. the pack with have thermal couplers to monitor temperature.

C-rate can have a significant on battery efficiency. Each of the three effects of exceeding a batterys C-rate can reduce the batterys output. This effect is pretty straight forward in a long track competition where the vehicle gets up to speed and then runs at maximum efficiency for one hour. One hour at maximum discharge is by definition 1C.

Things get a bit more complicated for a short track competition. In every curve the vehicle goes through a cycle of coast/brake into the turn, accelerate through the turn until optimal speed is achieved, run at optimal speed until the next turn. As a result, the C-rate is varying between 0C while coasting, greater than 1C while accelerating , about 1C while traveling at optimum speed.

As with overall capacity, theory and reading the manufacturer's specification can get you in the ballpark. It takes testing to determine how your battery is going to behave in a particular situation.

This post introduced the idea of C-rate and how rate of discharge affects a batterys safety and efficiency. The next post will look at another key safety concept, cell balancing. things tend to happen if a batterys C-rate is exceeded:

1. Voltage sags

2. Undesirable chemical reactions occur which damage the battery.

3. Energy is lost as heat.

We are most concerned about the generation of excess heat. Some battery chemistries can catch fire or explode if pushed to an extreme. Undesirable chemical reactions can reduce the useful life of a battery.

In the last two posts, we looked at how to put cells in series to create a desired output voltage and C-rate. In this post we are going to extend those ideas to cell balancing.

I took quite a while to get my head around cell balancing. In the DIY and hobbyist site some people seem very passionate in their positions without provide facts to back up those positions. This can make it hard to get through a discussion without getting confused :(

My first question was why do some types of cells need balancing and others dont. The best answer I found was by user36129 on stackexchange. I highly recommend stack exchange as a reference. It has attracted a very large community of extremely knowledgeable people.

To make a long story a little shorter. Due to manufacturing imperfections all cells have slightly different capacities and C-rates. Whenever batteries are charged and discharged in series, they can become unbalanced. Some type of batteries handle over and under voltage better than others.

Orginal image from wikipedia

Lead-acid cells are rather forgiving. They tend to give off gas when overcharged and die gracefully when run below their minimum voltage. My wow these things are amazing instance occurred one day while helping a friend put tow lights on his new truck. He accidently dropped a wrench across the batteries terminals. For a few seconds the points of contact between the wrench and the terminals arced brightly until the wrench welded itself to the terminals. Then the wrench started to glow red until it melted near the middle. Much the energy in the battery discharged across the wrench in less than 30 seconds. without catastrophic results. That is impressive.

Lithium is much less forgiving. Operating out of specification can cause lithium batteries to release their stored energy explosively. Because of the greater danger associate with lithium we need to take step to ensure that individual cells are not over or under charged.

On a practical level we will need to balance the individual cells in our pack often enough that individuals cells dont go above or below their operating voltage. How often we need to balance depend on factors such as: the quality of our materials, the quality of our construction, and how close we operate our cells to their limits

There are many methods of balancing cells. The basic ideas is that the BMS must be able to detect the voltage of each cell and then charge or bleed the cell as necessary to bring the pack back into balance.

Original images from instructables

The final post for this section will be about narrowing down our choice of batteries. We are at an fascinating point for battery development. We are surrounded by consumer electronics which compete franticly on size, weight, and battery life. This drives down cost.

Darpa is heavily involved in drone research. This tends to push basic research forward on related technologies forward.

There are several electric vehicle that have been in use for over 10 years. This results in a large body of data on how those cars and their batteries perform in the long term. It also establishes a strong safety record.

For this project I have used five categories of battery to keep things clear in my mind; SLA, EV, Exotic, hobby, Commodity.

SLA -- Sealed Lead-Acid batteries currently used by most electrathon racers. Optima Red and Yellow top are the most common and well understood.

EV -- Electric Vehicle. A second class of batteries which show potential are modules from battery packs currently in use by production electrivehicls. The biggest advantage of these modules is that they come in premade assembles ready to bolt onto a vehicle. The downside is the low C-Rate common in these batteries. Most EVs are designed to go for a least 3-4 hours between recharges. C-Rates between .25 and .5 are acceptable.

Exotic -- Exotic. This is where a lot of the cool research is happening. I would love to get into this someday. Prices are prohibitive. Prices between $1000 and $10,000 for a 1kWh pack is common. This is above the budget for me and most teams. To be honest, it doesn't seem much fun to buy ones way to victory lane. I would rather limp my way back to pit lane with a smoking battery that I built myself:)

Hobby -- Hobby batteries are where things start getting reasonable for an electrathon team. By hobby batteries, I mean HobbyKing and friends :) There is a lot of interesting work happening in the DIY crowd with these batteries. They are still a bit pricy, but they could be within the budget of many teams if they can be proven to last for a couple of years. The biggest drawback of this type of battery is their Quality Control. In order to build a pack one must do a significant amount of testing on individual cells.

Commodity -- I came across commodity batteries by accident. Many consumer products use batteries that are made of common cells. One of the most common is a cell called the 18650. They are everywhere. Most laptops have 6-9 of these cells.

For this build, I settled on 18650 cells as good enough at a reasonable price. Other advantages is a lot people have tried this before, there are plenty of people to ask for help. As a mixed blessing the Tesla-S uses battery packs made of 18650 cell. On the plus side, a major manufacture chose to use these cells. On the negative side, supply is pretty tight so prices are higher than normal.

Next post, we will design a battery pack out of 18650 cell that will work in an electrathon racers.

Sorry, the last couple of posts have been coming fast and furiously. I had a bunch of notes that seemed like they could be broke down into daily topics of discussion between a coach and bright high school team member who wanted to learn enough background to look into experimental batteries.

Rather than jump to battery design, this post might be a good time to look at easily available references.

Wikipedia -- While, understandably, teachers might prefer that students dont reference wikipedia in papers. It is an awesome first reference in the technical fields. A few facts might be out of data or an author might have a biased point of view. But overall, it is a good introduction to many topics.

Battery University -- This is a good site for learning about batteries and their use.

Endless-sphere -- I highly recommend several thread on this forum. Particular some of the ones on home made battery packs. But PLEASE keep your bull**** detectors on at all times. Online forums are the wild west. Many participants have more enthusiasm than experience or empathy :(

Electric Samba -- If you like videos, there is a guy who is documenting his project to make an electric bus. The episodes on the electric system and interesting and fun to watch.

The first thing we are going to need is an 18650 spec sheet. While the reading can be pretty dry, it is invaluable.  When a engineer from somewhere like Apple is deciding what battery to use in a new line of laptop their first stop is usually a pile spec sheets from vendors. After checking the spec sheet, they will order a few test units to verify that they meet specs.

If the test units dont meet the listed specifications, the engineer moves on to a different vendor and is pissed they wasted their time with the first vendor. Not only did the first vendor lose the sale, the engineer is going to be unlikely to even bother considering that vendor until they earn his trust again.

Looking at the spec sheet linked to above we see:

Nominal capacity-----------------------------------------------------------------2200mAh

Weight-----------------------------------------------------------------------------------48 g (ref)

Maximum charge voltage-----------------------------------------------------4.20 0.05V

Normal Voltage------------------------------------------------------------------------------3.7V

Minimum discharge end voltage--------------------------------------------------------2.75V

First let's see how many we can fit in our pack given the weight limit of 6.8.kg

6800g / 48g/cell = 141.7 cells

Let's see how many amps that can store.

141 * 2200mAh = 310 ah

Finally we need to convert that to kWh

310kWh * 3.7v = 1.15 kWh

That is in the right ballpark. Just to be safe let's assume we are going to lose 10% of our available weight to things wiring

1.15 kWh * .90 = 1.035 kWh

Good enough that we can proceed with a prototype.

Time for another aside.... The market for 18650 is pretty turbulent right now. As far as I can tell there are a handful of factories in China make all or most of the 18650s. Battery vendors buy from those factories put on their own brand. If you look at a site like http://www.batteryjunction.com/18650.html  you can see the only distingusing charastic of an 18650 cell is a shrink wrap sleeve with some printing. Whenever you buy cells, buy a few and test them thoroughly before making a larger purchase.

There is an interesting review of some of the common brands at http://lygte-info.dk/review/batteries2012/Common18650Summary%20UK.html .

Rather than spending the money on 150 new 18650 cells for our prototype, this prototype is going reuse cells from laptop batteries. They are dirt cheap, you can find them from recyclers or ebay for as low as $0.25 per cell. They will work well enough so serve as a proof of concept that we are on the right track. Or they will allow us to learn that we are heading down the wrong path before we have spent too much money.

If you look around old laptop batteries are pretty easy to find. It is time consuming, but they are pretty easy to take apart. I was able to take apart 40 batteries I bought on ebay in about 3 hours while watching a movie. That yielded me 240 18650 cells. Some of them will be junk. Hopefully 150 of them will be good enough to use in the pack.

If you want to build along grab yourself some 18650 cells. Tomorrow, we will start looking at testing methodologies to sort out the good cells from the bad cells.

One of the things which makes electrathon, and all racing, fascinating is how so many subassemblies work together to make the whole vehicle. A good engineer must continually zoom his or her focus in and out. They switch between details and the whole. In that vein, let's take a step back from battery details and look at what we are trying to achieve.

The ultimate goal is to use the battery to propel a vehicle around a track as many times as possible in exactly one hour. In long track racing, speeds are relatively consistent. Energy drawn from the battery is relatively consistent over the entire hour. In short track racing the vehicle speed varies significantly as the vehicle accelerates, cruises, and decelerates between turns. Energy drawn varies from zero when decelerating to whatever significantly higher than average during acceleration.

This is significant because for our purposes, total capacity doesn't really matter. What matters how much energy we can draw in an hour at the expected rates. One way to observe this is to test our battery against a dummy load.

There are several way of constructing this load. I chose to build a resistor bank consisting of 20 1.5 Ohm power resistors. Each resistor is capable of dissipating 100W.

I came up with these number by combining the formulas I = V / R and P = I * V to get P = V * V / R . This gave me 12V * 12V / 1.5Ohms = 96 Watts . We can vary the load by steps of 96 watts for each resistor we add in parallel.

I have read about people varying the discharge rate by adding or removing resistors. Since we already have a PWM controller, I am planning on running the control between the battery and the resistor bank. Adjusting the throttle should enable us to precisely vary the current draw and exactly mimic the conditions on the track. (As an aside -- I need to see if it is necessary to include some inductance in the circuit when connecting a PWM controller to a resistive load)

I've ordered the resistors. Hopefully they will arrive in a couple of weeks.

I have spent the last several evenings hunched over my electronics bench watching batteries charge. I think the rest of the family has decided that I have finally lost my mind:)

At first it was a bit boring. After a while things got interesting as I started to drag out my mostly forgotten chemistry knowledge and speculate about what was happening inside the cell.

Along the way I came across three outstanding sources.

The TAB Battery Book - An In-Depth Guide to Construction, Design, and Use by Michael Root. I found this to be appropriately paced for someone who had not opened a chemistry book in 20 years. Probably a good fit for a 'hands on' high school student who has been playing with batteries for a while and wants to figure how they work.

Once you understand the basic chemistry Jack Rickard @ EVCCON2013 : Lithium Ion Battery Basics provides a good low level overview of lithium ion battery. His knowledge is extensive once you get past his humor.

Finally Why do Li-ion Batteries die and how to improve the situation provides many specific details about why cells fail.

BMS - Battery Management System. There is a good introductory article at Wikipedia. Just as with battery capacity, the BMS is something we need to look at with our specific vehicle needs in mind. The primary purpose of a BMS is Protection.

Overvoltage. This is a big issues since over voltage can cause damage to the battery or start a fire. For now we will be charging the battery with an external charger which is specifically designed for our needs.

Undervoltage. Undervoltage in many type of batteries can greatly reduce its life. For now, we are going to prevent undervoltage by setting the minimum voltage cutout in the controller.

Overcurrent. As we undervoltage, we will be depending on the motor controller to prevent over current situations. We will also use fuses.

Temperature. For now, we are going to use a handheld Infrared Thermometer. It is a bit more flexible.

Balancing. We will go into this in more detail later. For now we will use balancing charger to correctly rebalance at every charge.

The final aspect of a BMS is to provide information on the state of the battery such as charge remaining. For now, we will do all those things manually in order to get a good understanding of how they are calculated. While it is a bit more work, it will help us learn to recognize and diagnose problems. Later if we decide to add a dedicated BMS we will understand what it is doing.

Speaking of mistakes. I have always thought of batteries as the equivalent of high pressure air tanks.

You use an air compressor to fill the tanks.
The tank contains a certain amount of energy which you can use to run an air tool.
How long the tool runs depends on the volume of air delivered (cubic feet) at a certain pressure. (pounds per square inch)

A simple battery appears to be quite similar. You have pressure (voltage) and volume of stuff. (Amperage) If you want your battery to store more energy, you either need to increase the volume or the pressure.

This model works well enough if you are just dealing with simple batteries, however if fails to explain enough to work safely with modern batteries in series. The error is that air pressure stores energy by squeezing molecules together. Everyone should remember PV = nRT.

Electricity (for our purposes) works by moving charged particles around a circuit. Charging a battery works by moving those charged particles from somewhere they want to be, to somewhere they dont want to be. Discharging allows the particles to return to where they want to be.

A key to understanding batteries is to understand that all we are doing is moving charged particles arround

I am going make an abstraction at this point and just talk about the flow of charged particles. When looked at in detail we learn that these charged particles can have either a positive or negative charge and can be either electrons of various types of ions. However, it is easy to get caught up in the details of how things are named and lose sight of how things work.

We need to start by imagining a piece of wire. If we put a charged particle in one end of the wire, a charged particle will come out the other end.

Something similar happens if we split that wire in two segment which we re-attach with a battery. If we put a charged particle in one end:

The charged particle flows to the end of the first segment.
The charged particle flows from one terminal of the battery to the other.
The charged particle flows from the beginning of second segment to the the end.

The same thing happens if we put multiple batteries in series along the wire. If a charge goes in one end, it pass through every battery before coming out the other side. It does not matter it an individual cell is full or empty, whenever a charged particle flow down the wire it goes through each cell.

The next abstraction I am going to make is to consider a battery that can hold 10 units of charge. The voltage of the cell varies as defined in the table below.

0    0.0V
1    2.0V
2    3.0V
3    3.2V
4    3.4V
5    3.6V
6    3.8V
7    3.9V
8    4.0V
9    4.1V
10  4.2V

If we start with each cell having no charge and push 10 units of charge into the system, each cell will raise to 4.2v. We can then safely discharge 8 units to bring each cell back to 3v. We will call this the standard charge cycle. In an ideal world, we can do this forever.

However, in the real world of manufacturing tolerances each cell is not exactly the same. For example a cell can have a higher than normal internal resistance. The cell might waste 0.1 unit of charge during each charge cycle as heat.

As a result, if we follow the standard charge cycle, after 10 cycles that cell will be charging charging and discharging between 4.0V and 2V which is outside of the safe operating range.

As a second example, as a cell gets older, crud builds up on the electrodes and its capacity can go down. As a result that cell might hit 4.2V when only 9 units of charge have been put into the cell.

The problems start to happen when we try to use these slightly different cells in series. Lets say we have 8 perfect cells, one high internal resistance, and one lower capacity cell in series. At the end of the 10 standard discharge cycle the overall battery voltage will look like

8 * 3v + 1 * 2v + 1 * 3v = 29v

Even though the overall pack voltage is pretty close to normal, one cell is down to 1 volts.

We can imagine all sorts of situations where individual cell are out of the safe operating range while the pack still looks pretty good. We need to deal with this situation intelligently because, a surprisingly large percent of accidents (and cell damage) with high capacity cells such as lithium ion happen when a cell is over or under charged.

Next we will look at battery balancing.

The Nissan Leaf modules arrived last week. I am impressed. They appear to be well designed and very suitable for electrathon use with two exceptions.
1. They are slightly overweight.
2. The cells are arranged as 2s2p.

This arrangement results in a nominal 7.6 volts per module or 15.2 volts for a pair of modules. At this voltage we start getting significant losses due to resistance unless we use a large amount of copper for our wiring and motors.

So, we must take each module apart and rewire them in series. There is an interesting series of teardown videos at



The biggest takeaway from the video is that the cases are totally destroy during dis-assembly.

I have disassembled mine to the point where I need to use a Dremel to cut the copper bus bars between cells However, I want to think a few days before chopping up the bus bar with a cutting disk :)