Classic Computer Magazine Archive COMPUTE! ISSUE 146 / NOVEMBER 1992 / PAGE 62

Salt and battery. (notebook computer batteries) (Column)
by Mark Minasi

I love notebook computers!

That's my song. Hey, I'm on the road all the time. Without my notebook, I couldn't write books, articles, course handouts... heck, I'd be out of business.

For example, I recently did an Alaska-Ireland-England-Italy, series of lectures. With the big Alaska flight coming up, I charged up two laptop batteries. I got about 45 minutes' work out of the first battery--less than I'd expected but not the end of the world. I saved my work, changed the battery, and powered up, expecting another 45 minutes or more. It was just when I was typing in a really great riff that the second battery died, after 15 minutes ... and no warning.

Which reminds me of my other song or, perhaps better, dirge:

I hate laptop batteries!

I've owned six PCs that weighed less than eight pounds each and ran on batteries. And every one of them has left a pain in the pit of my stomach whenever I used it with on-board juice rather than AC power.

When the battery is new, all is well with the world, and you get almost two hours of work out of a 386SX notebook. But after six months, it's like playing Russian roulette every time you trust your work to batteries. Your car has a gas gauge; why can't a laptop tell you how much zap's left in its battery?

To find out, I called up Phil DeSantis. Phil is the marketing and sales manager for the Battery Systems Division of HM Electronics in San Diego. Phil's been in the battery business for 20 years, and I was quite surprised by some of the things that he told me.

It's pretty much common knowledge that the ni-cad batteries that are in laptops have a memory--a phenomenon whereby they remember how much they were charged up and eventually lose all charge capacity beyond the last charge.

For example, if you always recharge your battery when it still has a half charge on it, it'll eventually retain only the capacity to hold a half charge. That means that you have to be really religious about completely discharging your laptop before charging it again--no topping off.

All right, now, how many of you knew that or had heard something like it before? Hmmm ... lots of hands raised. I'd heard that, too.

"Baloney," says Phil. "The 'memory' problem was solved years ago. You can charge and recharge your laptop all you like."

Hang on, you might be thinking; you can deny the memory phenomenon all you like, but it still happens. My old laptop batteries just don't take as much charge as they did when they were new.

Phil explains that the problem with your laptop's battery isn't memory; it's dendrites. These are salt-like crystalline formations that grow inside a battery when you overcharge it. When you try to charge a full battery, it gets hot, and the dendrites form.

Similarly, when you charge a battery that's empty or partially empty but hot, the charging process is much less efficient, and you get dendrite growth to boot. Dendrites act like an internal resistance in your battery, reducing its efficiency and its service time when charged.

So the person who plugs a laptop into the wall every night and gives it an eight-hour charge, when it only needs an hour or two, is doing the battery two hours of good and six hours of harm--and slowly growing a veritable dendrite farm.

Worse yet, the heat can cause other problems, such as internal chemical changes that cause short circuits right inside the battery. That kind of bad news means shorter battery life and greater heat.

Once your battery has dendrites, is it no good? No, it's recoverable. The battery can be conditioned. The way that's done is simple. Just discharge it until each cell produces no more than one volt--don't discharge any further, for reasons we'll take up in a few paragraphs--and then recharge. Do this discharge/recharge cycle three times, and most of the dendrites will be gone.

How Chargers Work

Ordinarily, a charger pumps power into a battery at about the same rate that the battery puts it out, though some go down to 1/2 that rate. That means a battery that can drain itself in one hour in the laptop should take no more than an hour or two to recharge itself.

Once it's charged, a certain amount of the charge dissipates, so a charger should then shift down to a tiny 1/20-1/50 of the previous charge rate.

The problem with this is that detecting when a battery is fully charged is a complicated process. Once the battery is full, a charger should detect a slight reverse voltage.

That's not completely reliable, however, as a slight power surge from the wall current could look like a voltage reversal. If the charger misses that voltage shift, it could notice that the battery is increasing in temperature.

Chargers today aren't very accurate in detecting when a battery is full, and when they do detect it, they don't drop to 1/20 or 1/50 of the normal current--they only drop to 1/10. The result is a warming battery, with all the attendant problems.

Handling Your Battery

I had problems with the first battery that my notebook came with, problems that looked like memory problems. So I figured that the right thing to do was to completely discharge the battery.

Unfortunately, my laptop refused to run at all when the battery was low, so I never had a chance to completely discharge the battery. I had a brainstorm, however. As the battery pack is just four D-cell ni-cads, I just got a flashlight that took four D-cells and ran the flashlight until the light went out.

Rather pleased with myself, I did this through a few charge/discharge cycles, and then the batteries refused to charge at all.

"You did the exact wrong thing," Phil told me. (Now he tells me!) Nicads should never be discharged below one volt per cell. If you do that, they move into an unstable state and can reverse polarity.

Then when you try to recharge them, it's as if you put the batteries into the laptop backward before you started up the charge cycle, an activity that can be dangerous. In my case, my laptop heated up to over 130 degrees before giving up the ghost. Now that I have new batteries, it runs cool again.

And when you get a new battery, give it a good charge--don't worry about overcharging it the first time. The reason is something called stand loss. Here's how it works.

A battery discharges all by itself about 1 percent per day. Now suppose you have a laptop battery composed of four cells. This battery gets shipped over from Japan, sits in the warehouse, and sees a month or two of inactivity before you use it.

The problem is that each of the four cells discharges at a different rate, so the four of them may be in very different places in their discharge cycles. Remember the possible polarity reversal if the voltage drops below a volt? If you've got a cell that's at about one volt and turn on the laptop, you'll discharge all the cells, but the one that's on the fence may just flip.

This could leave you with a battery composed of three cells that agree on polarity and one that fights the others, with the resulting hot battery and short charge life.

One way to keep that from happening is to let your battery charge for 24 hours the first day that you get it. There will be some overcharge, but the effects of that are far outweighed by the benefits of ensuring that all the cells in your battery are in step.

Once you start using your laptop regularly, how should you handle the charging? First of all, try to get an idea of the ratio of charging time to running time.

Do an experiment to find out the longest running time for the battery. Then charge the battery for that amount of time, and see how long a run you get from that charge.

With a little experimentation, you can determine the running/charging time ratio that your system uses. Suppose you find that to get an hour's use from the battery, you must charge for 1 1/2 hours. Then you should only recharge for the amount of time equal to how long you last used the computer while on batteries, multiplied by 1 1/2.

Battery Gas Gauge

But what about my initial question? Why doesn't my laptop give me a warning sometime before it dies?

"Easy," says Phil. "Ni-cads have a funny discharge pattern. They give you just about full voltage, with very little decrease, until they quit." There's no straightforward way to know when your battery is about to die.

The computers that give you a percentage-depleted number are merely guessing, based on how long the battery's been running and when you last told the computer that you'd charged the battery. They assumed the battery discharges at a pretty steady rate--a fairly heroic assumption.

What all this boils down to is that you can't trust the full- and low-power indicators on many laptops.

HM Electronics, however, sells a neat little doodad that solves many battery problems. This battery analyzer/charger unit does three things.

First, it charges your battery safely, as it knows when to stop charging. Once the battery has reached peak voltage, the charger continues to provide current, but only at a small fraction of the normal charging rate, so as not to heat up the battery and form dendrites.

Second, it discharges the battery, which (1) discharges the battery safely, not allowing it to drop below one volt, and (2) monitors the discharge process, reporting exactly how much charge it held. You can use this to measure your battery's capacity, as you can charge and then discharge the battery to determine exactly how much juice it can hold.

Finally, it performs a conditioning function whereby it does three charge/discharge cycles, dedendriting your battery.

Currently, HM has units for only the Grid and Dataworld portables. I use one for my Dataworld portable, and I must admit that I love it. It has taken the mystery out of my battery capacity. It's as if I drove a car for years without a gas gauge, and all of a sudden someone installed one--I don't know how I lived without it.

The unit I bought cost $200, but Phil says that there should be a generic unit out for about $160 by the time you read this.