More on batteries. (batteries for computers) (Column)
by Mark Minasi

The more I find out about this battery stuff, the more there is to know. This month, a bit more on the nicads and chargers, a discussion of their would-be successors--the nickel metal hydrides, and a quick peek at the El Dorado of batteries . . . lithium.

Last month, I talked about some of the basic problem with nicad batteries and their chargers. For those who are just joining us, here's a quick review. Nicads are the most popular form of batteries that store electricity for portable computers, hand-held radios, videocassette recorders, and the like. Nicads can't really store that much juice, so anything that really needs a lot of power won't work well with nicads. That's why you'll never see nicads under the hood of an electrically powered car.

Look back at the list of devices that use nicads--VCRs, notebooks, and walkie-talkies--and you see devices that are basically solid-state. It always amazes people when I tell them this, but computers basically use no power at all. For example, my 386SX notebook contains 16MB of RAM, a floppy drive, a backlit LCD screen, and a 120MB hard disk, yet it only draws 15 watts of power. Fifteen watts! That's about one-fourth of the amount used by the common 60-watt bulb that you probably have in your desk lamp. My earlier 8088-based laptop with no hard disk and a backlit screen drew an even more pusillanimous 8 watts, but that computer was a power spendthrift when compared to the 386SX notebook. Even though the notebook uses more power, it's doing a lot more.

Notebooks have to be more miserly in their use of power, largely because it's hard to make batteries better, and batteries are heavy. The battery in my Dataworld notebook is just a tube containing four D-cell batteries, and that probably accounts for 20 percent of the weight of the notebook all by itself. By the way, in the process of researching this article, I found that the battery classification system that's based on letters was developed in 1926 by the American National Standards Institute--ANSI to its friends. It includes not only the common AA, AAA, C, and D batteries, but an A (kind of long and narrow like the AA and AAA), as well as a B (which pretty much doesn't exist any more), an E, and an F. You can find an F by opening a lantern battery; there's a bunch of them in there. Nowadays, there are also AAAA (really tiny), as well as G, J, N, and 6. Some of these can actually be found in your local Radio Shack; I know because I need the N batteries for my hand-held laser pointer that I use in class.

Merely saying that a battery provides, say, 1.5 volts doesn't tell the whole story. Battery capacities are rated in terms of milliamp hours, which tell how many milliamps the battery can provide for a period of one hour. The batteries in my laptop, for example, provide about 6000 milliamp hours.

Milliamp hours measure capacity, but, again, we're interested in capacity that's lightweight. That's where a new unit of measure comes in. Energy density is a measure of how many milliamp hours each pound of a battery provides. Obviously, if my battery could hold 6000 milliamp hours in one ounce, my laptop would be a lot lighter. That's where nickel metal hydride comes in . . . or so they say.

According to the makers of nickel-metal-hydride batteries, these batteries have an energy density that's 25 to 50 percent better than that of nicads. But the more you know about them, the less you like them.

First, most makers of nicads underreport the capacities of their batteries by 20 percent or so. Nickel metal hydride, on the other hand, doesn't get under-reported. (I don't know why. Perhaps because it's a new technology and needs some hyping?) As a result, nicads don't look as good as they would if put on an equal basis with nickel metal hydride. While that dulls hydride's luster a bit, don't give up hope for battery improvement in general.

An even newer technology, the lithium battery, promises an energy density two or three times better than that of nicads. But it's still on the drawing boards. And with rechargeable batteries, it's always a guessing game as to how much longer they'll work until they die. To see why, take a look at figure 1. What you see is a graph of the voltage output of a battery plotted against its remaining capacity. As the graph marches from left to right, more of the battery capacity is gone, but so also goes the output voltage. As most batteries discharge, their voltage drops steadily. That makes it easy for a sensing circuit to predict when the voltage will drop below some critical point. It's also why the built-in battery tester on the Energizer batteries works--it's measuring output voltage and using that as a proxy for the remaining milliamp hours. But look at the same curve for a rechargeable battery in figure 2.

Rechargeable batteries provide just about the maximum voltage level until right before they die. That's why your laptop runs just fine up until the moment it dies. It's also why you need a computer-controlled charger and battery analyzer, as I explained last month. If you use a laptop, I can't stress strongly enough that you must have some kind of charger and analyzer. I used to get about an hour's use from my laptop when I used the charger that came with the laptop. But after buying the HME System 90 charger (call 800-233-6868 or 619-458-1500 for information), I've gotten almost three hours of work out of a single battery. It's simply wonderful that I can carry three batteries with me and get a solid eight hours' worth of computer time to write, draw, and, of course, play Risk for Windows. With the kind of deadlines that I've had this month, I simply couldn't have written my latest book, The Windows Problem Solver, on time without it.

The second thing to consider about nickel metal hydrides is that, as with nicads, there's the old problem of recharging these things. I told you last month that nicads must be treated properly when they're being recharged or they'll grow crystal dendrites that reduce their charging capacity. That's related to the common notion that nicads can develop a memory problem whereby they lose their capacity to charge. Nickel metal hydrides don't have that problem, which sounds good.

Nicads grow dendrites if they're overcharged or charged when hot. Removing the dendrites involves a process called conditioning the battery. Again, an analyzer and. charger can help; mine took a severely abused battery (OK, I was the abuser, but I didn't know any better at the time) that could deliver only about 4100 milliamp hours and raised its capacity to 5900 milliamp hours!

The bad news with nickel metal hydride is that it can also be damaged by overcharging, but when it's damaged, it's damaged for good--no conditioning is possible. Worse, a damaged battery may exhibit a discharge cycle like the one in figure 3. About halfway through, the battery drops its voltage output dramatically, perhaps below the voltage level needed for the laptop. Result--you've instantly halved the useful capacity of the battery. Again, no fix.

Another popular feature of many rechargers is a fast-charge feature whereby a battery charges fully in only an hour or two. That's possible with nicads, but not nickel metal hydrides. You need a fairly complex charge cycle to safely charge them quickly, and even then they don't charge as quickly as nicads.

While on the subject of recharging, there's another problem with nickel metal hydrides. Battery chargers use charging circuits that detect when the battery is charged so the charger can throttle back to a trickle mode, rather than continuing to force power into the already-full battery. One way of doing this is negative voltage detection; the nicad kind of splashes back power when it's full. Some charger circuits use this method, but it won't work on nickel metal hydrides--they don't show a negative voltage when full.

The third problem is outgassing. In plain English, that means the emission of gases by the battery when it's charging or discharging. If charged when hot, nickel metal hydrides outgas hydrogen gas--you know, the stuff that blew up the Hindenburg? I mean, I like a hot notebook as well as the next guy, but there are limits . . . I should mention here that outgassing is one of the big reasons why lithium batteries are still on the drawing boards; they produce some fairly toxic gases.

Finally, nickel metal hydrides just don't last as long. Nicads can be charged and discharged many times more than nickel metal hydrides can. If you routinely charge a nickel metal hydride to 80 percent of its capacity, you'll only get 50 percent of the service life you would've gotten from a corresponding nicad.

So what's the bottom line? It seems to me that the money spent on laptops that use nickel metal hydrides isn't well spent. If you want good capacity and less trouble, get a nicad laptop and a computerized analyzer and charger.

But what about making the laptops use less power? That's next month.