IAN LANG ELECTRONICS
Robots are electrical and electronic devices. They use motors. They need a control system. The more features on a robot, the more power hungry they are. If you read the power article in this section you'll know that a mobile robot has to have an on-board power supply. You'll also know that the best and sometimes only choice is a battery or batteries.
Rudebot, you will remember, uses six AA types. At the moment he's working with zinc-carbon types, which have a voltage output of 1.5V. As there's six of them in series that's 9V.
Zinc-carbon has a serious drawback. It's what is known as a primary cell type. Primary cells can't be recharged. If I try shoving through current the wrong way what I'll get is a very hot battery and eventually a gooey pile of melted toxins. This means that very soon I will have a large pile of spent zinc-carbon AA batteries which I'll take to Maplin or Asda to let disposal become their problem. Bwu-ha-ha-ha!!!
My problem is of course the cost. So what I can do and will do eventually when Rudebot's up and running is run him on some secondary cells. These are rechargeable and the cost will be far less to charge them than it will be to keep buying new batteries. You'll find that most robots can use rechargables of one sort or another for that very reason. I may need to use eight of them though (9.6V) if Rudebot does not perform.
In small bots like this we are stuck with the AAA, AA, C and D sizes. There are such things as full traction batteries which have a very high milliAmp/hour rating and a high energy density (i.e. how much energy you can get out of something for the size it is). The problem with full traction batteries is that a. they are expensive and b. They are huge and heavy. Wheelbarrow could handle something the size and weight of a car battery (which isn't a full traction battery, it's a starter-lighting-ignition battery or SLI) but Rudebot would struggle to move under that weight and he'd be at least four times bigger than he is. He's already 10 inches long and anything any bigger would be scantly portable and would have a large turning circle.
If you are considering full traction batteries (aka traction or electric vehicle battery) consider the weight of the battery as well as the energy density because in most applications the weight of the battery accounts for at least half that of the vehicle. You'll also want to consider the motor load. I know of at least one person who has blown a fuse in a mobility scooter trying to get it up a steep, grass-covered incline. The motor's been trying to draw oodles of current which the manufacturers don't think the battery will be happy with and so they stuck a fuse in there to stop it happening. You will look daft if your bot gets stranded because it's been trying to drag too much weight uphill, especially if half the weight is your battery.
Li-ion (lithium ion) have a good energy density of ~200 Wh/kg but are short lived and more seriously catch fire if they get punctured. Some are made to be fire-resistant but this lowers signifcantly the energy density.
Ni-MH (nickel metal hydride) have a somewhat lower energy density of ~60 Wh/kg on average (some are a lot lower and approach only 30, others are higher and top 80) but they do have extremely long lives. The trouble is they are not efficient at charging, wasting a lot of the charging power and can self-discharge so they are no good for standby robots that may not have to move for months at a time. Neither type performs well in the cold and Li-ion may not deliver a charge at all.
SLA (sealed lead-acid).are the best bet for robots on a large scale. I want to draw your attention to the LEAD part of it. Lead is very dense, and being very dense quite small bits of it are quite heavy. So are the batteries. You can get them in a range of voltages with a nominal 6V and 12V being common and generally speaking the bigger the battery in physical dimensions the bigger the mAh rating. Don't use a vented lead-acid battery- I draw your attention to the ACID part.
Whatever battery you intend to use the important part is the milliAmp/hour rating (mAh) as this tells you how long the battery is going to keep putting out the nominal voltage under the conditions of current draw. In theory. A 2500 mAh battery will, if it's used on a motor that draws a constant 2500 mA, which is 2.5A, run it normally for an hour and then it will flag and start to droop. You really shouldn't run a battery over the rating. Common sense says that if it will run a 2.5A motor for an hour, it'll run a 5A motor for 30 minutes. Common sense needs to shut up because what will actually happen is that the battery will get hot and it won't do 30 minutes, probably fifteen or twenty if you're lucky. It's the same the other way round too. If you run under the rating, as most will, the battery will last less long then you'd expect from the common sense point of view. It's difficult without looking at the manufacturer's methodology and test data to predict what the run time will be at a constant rate (in Rudebot's case about 0.5A) let alone a starty-stoppy affair but as a rule of thumb I work it out as rating/current draw and then halve the result. Any longer is a bonus.
You also have to remember that if you use rechargeable AAA, AA, C or D size you will not get 1.5V like a zinc carbon. You will in fact get 1.2V as this is what the chemistry is capable of. For what look like 9V ones you get 7.2V and not much mAh rating.
The more voltage you give to an electric motor the more speed. The more current, the more torque, and multiplying the two gives you the power at the output shaft. It's safe to say that the more grunt you get through a motor the more force the turning wheels will have. The most important things to remember are:
Do not run a motor over the design voltage as this is a recipe for dramatically shortening its lifespan.
Ensure your supply can provide enough current to the motor(s).
Choose a battery that gives you the best energy density for the size and weight.
Choose a battery for the voltage required for the highest voltage your system requires. You can always convert down for other parts of the system.
Try and make your battery no more than 50% and ideally less than 10% of the weight of your robot.
Of course if it is something like a robot arm then it does not need a mobile platform. If it does not need a mobile platform, use a mains adaptor to run it. In the long run it will be much cheaper and the supply will be constant, unless you forget to pay your electricity bill. If you are worried about power cuts you could run it from an uninterruptable power supply. These are designed to give a power output for a couple of hours with no power coming in and do so by means of a rechargeable battery. The range tends to be fairly limited as far as voltages go but it is possible to build a regulator circuit to get it down to where you want.
In summary, electric motors are power-hungry and gobble batteries like Eric Cartman gobbles KFC (and if you've never seen South Park you won't get that so never mind it) and the most cost effective way to power your bot if it has to be mobile is to use secondary cells (i.e. rechargables). The trouble with the smaller ones is a lower voltage, and you may have to design this factor in at the beginning. I haven't with Rudebot and I've got a feeling that I'm going to be tweaking the Arduino code towards the end to get a better performance out of him, either that or redesigning the battery box to hold eight AA batteries which might be a better idea as they'll last a bit longer. Power is probably the trickiest part of the development you'll face in bot-building and it's not unknown for people to go back to the drawing board entirely even when they've got a working device because they aren't satisfied with the power performance. That's the point of all the above- an attempt to steer you past some of the pitfalls. Good luck!
I'm not fat- I'm big-boned!
Ian Lang, August 2013