Overheating - Cleaning Fans and Troubleshooting Hot CPU Failure

Many of us own laptops where the fan seems to run an annoying amount of the time, but it's better than a laptop fan that doesn 't run at all. When the fan in your notebook fails, the usual symptom is an overheating CPU and automatic shutdown. Sometimes the laptop runs for 10 minutes, sometimes for a half hour, it depends on the task and the power. Troubleshooting a hot CPU can be as simple as making sure you aren't blocking the airflow outside the laptop. Often times, you can access the fans and heatsink without a huge job, on this Toshiba A65, all it takes is some careful prying and two screws. As always, remove the battery before working on any laptop. The first step is removing the combination blank and hinge cover that allows access to the two screws securing the keyboard. It is snapped into place with little tabs on the long edge and a big tab on each end. The picture to the right shows how the blank and the hinge cover are all one piece of plastic, so you have to open the screen to fully remove the blank. I should mention for desperate people with no mechanical skills, one non-invasive way to extend the length of time a notebook with poor cooling will run between shutdowns is to run it on battery and choose the most aggressive (ie, longest life) power saving mode. BTW, I'm working on this notebook because I was asked to bypass the power connector as a favor. However, the owner's got five kids in the house and I burned my finger on the metal part of the 100BaseT port after I got it charging again, so I'm afraid to return it for safety reasons.

The problem I was originally asked to fix should have tipped me off to the overheating issues this notebook would suffer. The pin had broken out of the center of the power connector on the motherboard, which meant the battery couldn't be charged. My fix (really a kludge) was to solder a new wire directly to the connector and bring it out the I/O port blank on the back of the notebook. The coax power cable was HUGE. Turned out to be rated for 2.0 amps. Two screws secure the keyboard in the laptop, and it's not even necessary to disconnect the ribbon connector from the motherboard to access the fans. This Toshiba laptop uses two fans, a large exhaust fan that pulls cooling air through the body of the notebook, and a smaller CPU fan on top of the large heatsink. The large fan to the right draws cooling air in the bottom of the laptop, the smaller fan in the center is directly cooling the CPU. Troubleshooting overheating is pretty much limited to making sure the laptop fans operate and aren't completely burried in lint. This is a pretty robust cooling system and likely to keep most laptops from shutting down from thermal overload protection, but they stuck a P4 2.8 Ghz CPU in this puppy! When I went out on the web and read some reviews of how this laptop performed when new, I saw comments like "Keyboard gets too hot to type" and "Frequent shutdowns from overheating." Sony made a Vaio with the same CPU, HP made an Pavilion and Dell an Inspiron. I've seen similar overheating and battery life complaints for them as well. You can always try an external cooler, the best rated is the Logitech for $29.95.

As you can see to the left, there was a build-up of dust on the CPU fan which I blew out with compressed air, but not enough to cause any thermal shutdown issues. The reason I burned my finger on this boat anchor is it just runs too hot as a design flaw. OK, I didn't get a blister, but I'm afraid to put this laptop with a kludged power connector in a family home. I really don't know how they could have forced more air through this thing than with the big intake fan they are already using, so troubleshooting includes making sure you have unobstructed airflow to the BOTTOM of the laptop for the fan to function. If you run the thing on a bed, or on your lap with the fan grille blocked by your leg, it's going to cook for sure!

Troubleshooting and Repair of AC Adapters

This collection of information deals with the troubleshooting, repair, and use (normal or unconventional) of AC (wall) adapters, transformers, equipment power supplies (nonswitching type), and batteries used in portable electronic devices and power tools.

AC adapters (may also be called: wall adapters, power packs, or wall warts) are those boxes hanging on the end of the power cords of many modern consumer electronic devices. Their output may be a single AC or DC voltage, or several, with or without regulation. Most of those outputting AC are simple transformers and with the addition of a diode or bridge and filter capacitor for DC. (Often, people refer to all types of AC adapters including those outputting DC as 'transformers' but this is not really correct.) Others (those that are compact and weigh almost nothing) may be sophisticated switchmode power supplies. In most cases, the output will be totally isolated from the power line for safety. However, some that are designed for applications like battery chargers may not be line isolated and should not therefore be used for any other purposes. Transformers in this context refer to the actual magnetic components which may be found in AC adapters or equipment power supplies. Equipment power supplies are those portions of the equipment that provide various (usually DC) voltages for its operation. Note that in this document, nonswitchmode types are discussed. Batteries are found in all sorts of portable equipment nowadays. The discussions in this document related directly to problems and repair. This is not intended as a comprehensive battery "FAQ".

SAFETY
For the common transformer based AC adapter, there is no danger anywhere inside the device once unplugged. Any internal overcurrent fuses or thermal fuses represent essential safety features of an AC adapter. These must not be removed except during testing. Where a fuse is found to be blown, use only an exact replacement. I really don't recommend running a repaired cobbled together AC adapter unattended in any case since even the sealed case provides some additional amount of fire protection. Inexpensive replacements are generally available. For power supplies inside equipment, the same basic precautions apply but access and repair are generally much more easily accomplished. The only real danger from an unplugged heavy iron transformer would be accidentally dropping it on your foot. :(

General AC Adapter Information

AC Adapter Basics
It seems that the world now revolves around AC Adapters or 'Wall Warts' as they tend to be called. There are several basic types. Despite the fact that the plugs to the equipment may be identical THESE CAN GENERALLY NOT BE INTERCHANGED. The type (AC or DC), voltage, current capacity, and polarity are all critical to proper operation of the equipment. Use of an improper adapter or even just reverse polarity can permanently damage or destroy the device. Most equipment is protected against stupidity to a greater or lessor degree but don't count on it. The most common problems are due to failure of the output cable due to flexing at either the adapter or output plug end. See section: AC Adapter Testing. 1. AC Transformer. All wall warts are often called transformers. However, only if the output is stated to be 'AC' does the device consist of only a (stepdown) transformer. These adapters typically put out anywhere from 3 to 20 VAC or more at 50 mA to 3 A or more. The most common range from 6 to 15 VAC at less than 1 A. The regulation is typically very poor so that an adapter rated at 12 VAC may put out 15 VAC with no load and drop to less than 12 VAC at rated load. However, some may actually output up to two times the rated voltage or more with a light load. To gain agency approval, the transformer needs to be protected internally so that there is no fire hazard even if the output is shorted. There may be a fuse or thermal fuse internally located (and inaccessible).

If the output tested inside the adapter (assuming that you can get it open without total destruction - it is secured with screws and is not glued or you are skilled with a hacksaw - measures 0 or very low with no load but plugged into a live outlet, either the transformer has failed or the internal fuse had blown. In either case, it is probably easier to just buy a new adapter but sometimes these can be repaired. Occasionally, it will be as simple as a bad connection inside the adapter. Check the fine wires connected to the AC plug as well as the output connections. There may be a thermal fuse buried under the outer layers of the transformer which may have blown. These can be replaced but locating one may prove quite a challenge. 2. DC Power Pack. In addition to a step down transformer, these include at the very least a rectifier and filter capacitor. There may be additional regulation but most often there is none. Thus, while the output is DC, the powered equipment will almost always include an electronic regulation. As above, you may find bad connections or a blown fuse or thermal fuse inside the adapter but the most common problems are with the cable. 3. Switching Power Supply. These are complete low power AC-DC converters using a high frequency inverter. Most common applications are laptop computers and camcorders. The output(s) will be fairly well regulated and these will often accept universal power - 90-250 V AC or DC. Again, cable problems predominate but failures of the switching power supply components are also possible. If the output is dead and you have eliminated the cable as a possible problem or the output is cycling on and off at approximately a 1 second rate, then some part of the switching power supply may be bad. In the first case, it could be a blown fuse, bad startup resistor, shorted/open semiconductors, bad controller, or other components. If the output is cycling, it could be a shorted diode or capacitor, or a bad controller.

About AC Adapter Ratings

The following mainly applies to AC adapters using transformers. Those based on switchmode power supplies adapters have tended to be well designed with decent regulation and realistic ratings. Of course, they are generally also much more expensive! There is no standard for rating AC adapters. When a particular adapter is listed as, say, 12 V, 1 A max, there's a good chance the output will average 12 V when outputting 1 A but what it does at lower currents is not known. In fact, lightly loaded, the output voltage may be more than double its nameplate rating! This could be disastrous where a piece of equipment is plugged into it that doesn't expect such a high voltage. The rating also doesn't say anything about the ripple (for DC models) - it could be almost anything. The lifetime of an AC adapter (particularly those outputting DC) when run at or near its nameplate rating may be very short. Why? Because they often use low temperature (cheap!) components that can't take the heat. For AC output models, the transformer itself

may fail (or at least the thermal fuse). For DC models, the electrolytic capacitor(s) may go bad very quickly. The likely result will be that the output voltage will disappear entirely (AC models) or drop in value with greatly increased ripple (DC models). Where the adapter is used with its intended equipment, one can presume the manufacturer did the proper testing to assure compatibility and adequate life (though this isn't always the case!). However, where it is used in some other application, the life of the adapter and the equipment may be much shorter than expected, possibly failing almost immediately.

Protect Yourself from "Unknown AC Adapter Syndrome"

Apparently, manufacturers of equipment powered by AC adapters have discovered that they can improve their bottom line by not printing the AC adapter ratings on the device itself, and possibly not even in the user manual. I don't know whether this is actually done for liability reasons (so you aren't tempted to actually use an AC adapter other than their own exorbitantly priced replacement) or just to same 3 microcents on printing ink but the net result is that the owner has no idea what adapter in that drawer that collects adapters is the correct one. They could at least specify a particular model adapter if they don't think the average consumer has an intelligence greater than a carrot. For example, I own 2 U.S. Robotics modems. One uses a 9 VAC adapter; the other uses a 20 VAC adapter. The power jacks are identical and totally unmarked. Guess what happens if I guess wrong? With too little voltage, the modem may appear to work but be unreliable. With too much voltage, the smoke will very likely be released instantly. :( To save yourself a lot of hassle and possible damaged equipment, put a label on each AC adapter powered device you own with the voltage, current, AC or DC (with polarity), and model number of the adapter (make one up if nothing is obvious and put it on the device and adapter). Then, if you misplace the adapter, you'll know what to look for and if it is nowhere to be found, will have enough information to purchase a replacement.

Why do AC Adapters Usually Use Heavy Transformers?

The main reasons are safety and cost. Line isolation is essential for safety with respect to electrical shock - no part accessible to the user must be connected to either side of the power line. A regular transformer provides this automatically. While combinations of passive components can reduce the risk of shock, nothing quite matches the virtually fail-safe nature of a simple transformer between the power line and the low voltage circuitry. To achieve similar isolation without a line transformer generally requires a switchmode power supply which actually contains a small high frequency transformer to provide the isolation. Until recently, such systems were much more expensive than a simple iron transformer but that is changing and many modern devices do now use a wall adapter based on this approach. These can be recognized by their light weight, DC (probably regulated) output, and the required warnings NOT to cut them off and replace them with an ordinary plug! I wonder how many people have ignored the warnings when their equipment stopped working and replaced that fat "plug"? What a scenario for disaster! 7

Compact AC Adapters
These use switchmode power supply technology and can therefore be quite small and light weight. In addition to the applications noted below, they are turning up on a variety of other high tech gadgets from shavers to Personal Digital Assistants. WARNING: DON'T attempt to disassemble or repair one of these unless you are familiar with the safety and troubleshooting information for larger switchmode power supplies they can be quite deadly. For some reason I've been fascinated by tiny wall wart AC adaptors that use switch mode power supplies, since they're light and can supply more current than similar linear power cubes. One type that keeps catching my eye is used a lot for "AC travel charger" accessories for cellular phones. These things connect via a cable to the bottom of a cell phone, much like the cigarette-lighter "charger/saver" accessories, only these are driven by house current. The typical wart is a small rectangular box, about the size of two 9V transistor batteries side by side, manufactured in China or Taiwan. The wall side is distinguished by the fact that the AC prongs line up with the long axis of the box, rather than the other way around as with most wall cubes. This makes it possible to put them side by side on an AC power strip. The opposite face contains a tri-mode LED which may display red, green, or orange under conditions I've yet to figure out. Recently I noticed one of these thingies in K-Mart as part of a modular power system for cell phones. There are several models of cigarette lighter cords, however the actual 12VDC car plug in _interchangeable_ and connected to the cable using a 4-pin modular telephone handset jack. Each model comes with a cable constructed to mate with the phone it's sold for. Next to these on the pegboard is a variant of the wall wart being discussed, also having a 4-pin handset socket, and sold as an accessory to the DC cords. Instead of using the cigarette lighter plug, you connect the cable to the wall wart and create a new device which uses house current. So I picked up the wall wart and started to play. It's marked as being capable of 5-15 VDC at 750 mA. Playing with the 4 output pins; one is ground, two are tied together and supply 14.35 VDC open circuit, and can deliver about 1.5 amps. The other reads about 13 volts between it and the ground. Unpowered there is a small leakage between the ground and the "13 volt" pin. Looking inside, there are two 8-pin DIPs on the PC board; both having identifiers sanded off. One is near the transformer end and the other is near the DC output end. All of the DC side output traces lead, directly or indirectly, to the second IC. My guess is that the "13 volt" pin is really used to program the output voltage between ground and the other two pins that are tied together. The cable sold for any specific phone

has some passive components inside that will cause the second IC to produce the required output voltage. Am I warm? A note on a similar if not the same system. I hacked a bit into the Igo Juice power supplies with hopes of charging various devices via solar panels. I bought a bunch of close-out tips for the system and broke them open to see how the pins relate to output. There were resistors between pins in the tips and I would assume as you did that this "told" the power supply what output to give on what pins. Cheaper than hacking more was to just buy tips with output voltages I needed and adapt the plugs if necessary to sizes that fit my devices. I used the discontinued (close-out and cheap at the time) Igo dual power adapter which can handle a range of DC voltages and output stable voltages to charge cell phones, cameras, music players, etc.

Substituting AC Adapters
This relates to replacing a missing or broken adapter for which the specifications are known.

They must both be the same type with respect to AC or DC. (In rare instances, either will work but this is the exception and there is no way to know without examining the circuit unless this capability is explicitly stated on the device or in the user manual.) If DC, the polarity must be the same. Even if the connectors are identical, it's a coin toss as to whether the center is positive or negative. However, the cable can be cut and spliced if incorrect polarity is the only problem.

The replacement much have at least the same current rating. Higher current is fine as long as it isn't ridiculously more (like 5 times). The connectors should be identical (including the size of the center hole). If the old adapter is dead, and it's not a cable problem near the connector, cutting and splicing in the old connector on the new adapter is acceptable as long as it's done very carefully and double checked for wiring errors and short circuits. The output voltage with no load must not be much greater than the listed value unless you know for sure that the equipment can take it, or that the other adapter behaved similarly. If the original adapter was regulated, then the new one should be as well. If the original adapter was a switchmode type (light as a feather), then it was almost certainly regulated. The replacement doesn't need to be switchmode unless weight is an issue, but it must be regulated.

Use a multimeter to check wall adapters found in your junk drawer and take one along to garage sales, thrift stores, and the like - even Radio Shack! Some simple tests of no load voltage and polarity can quickly identify a suitable adapter as long as their current rating is known (from the label).

 AC Adapter Troubleshooting and Repair

AC Adapter Testing
AC adapters that are not the switching type (1) and (2), above, can easily be tested with a VOM or DMM. The voltage you measure (AC or DC) will probably be 10-25% higher than the label specification. If you get no reading, wiggle, squeeze, squish, and otherwise abuse the cord both at the wall wart end and at the device end. You may be able to get it to make momentary contact and confirm that the adapter itself is functioning. The most common problem is one or both conductors breaking internally at one of the ends due to continuous bending and stretching. Make sure the outlet is live - check with a lamp. Make sure any voltage selector switch is set to the correct position. Move it back and forth a couple of times to make sure the contacts are clean. If the voltage readings check out for now, then wiggle the cord as above in any case to make sure the internal wiring is intact - it may be intermittent. Although it is possible for the adapter to fail in peculiar ways, a satisfactory voltage test should indicate that the adapter is functioning correctly. It's also possible that the power jack on the device itself is damaged from use or abuse. If possible, confirm proper operation with a COMPATIBLE adapter. With battery operated devices, there is usually a set of contacts that should close when the adapter is removed to connect the internal battery to the circuitry. If these don't operate properly, the device may not work off batteries (they may appear to not be charged), the AC adapter, or both. Check the jack for obvious signs of damage (cracked, loose, etc.). A squirt of contact cleaner into the jack may clear up intermittent contact problems not due to actual damage.

Pocket Wall Adapter Tester/Polarity Checker

This handy low cost device can be built into an old ball point pen case or something similar to provide a convenient indication of wall adapter type, operation, and polarity:
Probe(+) o-----/\/\-----+----|>|----+---o Probe(-) 1K, 1/2 W | Green LED | +----|<|----+ Red LED

The green LED will light up if the polarity of an adapter with a DC output agrees with the probe markings.

10

The red LED will light up if the polarity of an adapter with a DC output is opposite of the probe markings. Both LEDs will light up if your adapter puts out AC rather than DC. The LED brightness can provide a rough indication of the output voltage.

Getting Inside an AC Adapter

Manufacturers come up with all sorts of creative ways of making access a challenge:

Some are secured with screws - possibly with strange heads. If this is the case, disassembly is possible without damage, at least in principle. However, you may need to find or improvise for the special tool. Some are fully potted and impossible to open without dynamite. Forget it, move on with your life. :) These will feel solid and there will be no 'give' when pressing the sides.

For those that are glued:

A hacksaw or thin file can be used to carefully cut along the glue line just deep enough so that the two halves can be popped apart. Make sure you don't rip into internal components! Gently whacking a large knife with a soft mallet may be a bit more persuasive. :) A vise can be used to squeeze on diagonally opposing corners which will hopefully pop the case open along the glue line (or somewhere!).

After the repair, the two halves (or pieces!) can be glued back together using something like Duco Cement or windshield sealer.

AC Adapter Repair
Although the cost of a new adapter is usually modest, repair is often so easy that it makes sense in any case. The most common problem (and the only one we will deal with here) is the case of a broken wire internal to the cable at either the wall wart or device end due to excessive flexing of the cable. Usually, the point of the break is just at the end of the rubber cable guard. If you flex the cable, you will probably see that it bends more easily here than elsewhere due to the broken inner conductor. If you are reasonably dextrous, you can cut the cable at this point, strip the wires back far enough to get to the good copper, and solder the ends together. Insulate completely with several layers of electrical tape. Make sure you do not interchange the two wires for DC output adapters! (They are usually marked somehow either with a stripe on the insulator, a thread inside with one of the conductors, or copper and silver colored conductors. Before you cut, make a note of the proper hookup just to be sure. Verify polarity after the repair with a voltmeter.

11

The same procedure can be followed if the break is at the device plug end but you may be able to buy a replacement plug which has solder or screw terminals rather than attempting to salvage the old one. Once the repair is complete, test for correct voltage and polarity before connecting the powered equipment. This repair may not be pretty, but it will work fine, is safe, and will last a long time if done carefully. If the adapter can be opened - it is assembled with screws rather than being glued together - then you can run the good part of the cable inside and solder directly to the internal terminals. Again, verify the polarity before you plug in your expensive equipment. WARNING: If this is a switching power supply type of adapter, there are dangerous voltages present inside in addition to the actual line connections. Do not touch any parts of the internal circuitry when plugged in and make sure the large filter capacitor is discharged (test with a voltmeter) before touching or doing any work on the circuit board. If it is a normal adapter, then the only danger when open are direct connections to the AC plug. Stay clear when it is plugged in.

AC Adapter Substitution and Equipment Damage

Those voltage and current ratings are there for a reason. You may get away with a lower voltage or current adapter without permanent damage but using a higher voltage adapter is playing Russian Roulette. Even using an adapter from a different device - even with similar ratings, may be risky because there is no real standard. A 12 V adapter from one manufacturer may put out 12 V at all times whereas one from another manufacturer may put out 20 V or more when unloaded. A variety of types of protection are often incorporated into adapter powered equipment. Sometimes these actually will save the day. Unfortunately, designers cannot anticipate all the creative techniques people use to prove they really do not have a clue of what they are doing. The worst seems to be where an attempt is made to operate portable devices off of an automotive electrical system. Fireworks are often the result, see below and the section on: "Automotive power". If you tried an incorrect adapter and the device now does not work there are several possibilities (assuming the adapter survived and this is not the problem): 1. An internal fuse or IC protector blew. This would be the easiest to repair. 2. A protection diode sacrificed itself. This is usually reverse biased across the input and is supposed to short out the adapter if the polarity is reversed. However, it

12

may have failed shorted particularly if you used a high current adapter (or automotive power). 3. Some really expensive hard to obtain parts blew up. Unfortunately, this outcome is all too common. Some devices are designed in such a way that they will survive almost anything. A series diode would protect against reverse polarity. Alternatively, a large parallel diode with upstream current limiting resistor or PTC thermistor, and fuses, fusable resistors, or IC protectors would cut off current before the parallel diode or circuit board traces have time to vaporize. A crowbar circuit (zener to trigger an SCR) could be used to protect against reasonable overvoltage. I inherited a Sony Discman from a guy who thought he would save a few bucks and make an adapter cord to use it in his car. Not only was the 12-15 volts from the car battery too high but he got it backwards! Blew the DC-DC converter transistor in two despite the built in reverse voltage protection and fried the microcontroller. Needless to say, the player was a loss but the cigarette lighter fuse was happy as a clam! Moral: those voltage, current, and polarity ratings marked on portable equipment are there for a reason. Voltage rating should not be exceeded, though using a slightly lower voltage adapter will probably cause no harm though performance may suffer. The current rating of the adapter should be at least equal to the printed rating. The polarity, of course, must be correct. If connected backwards with a current limited adapter, there may be no immediate damage depending on the design of the protective circuits. But don't take chances - double check that the polarities match - with a voltmeter if necessary - before you plug it in! Note that even some identically marked adapters put out widely different open circuit voltages. If the unloaded voltage reading is more than 25-30% higher than the marked value, I would be cautious about using the adapter without confirmation that it is acceptable for your equipment. Needless to say, if you experience any strange or unexpected behavior with a new adapter, if any part gets unusually warm, or if there is any unusual odor, unplug it immediately and attempt to identify the cause of the problem. Or, a more dramatic result of the same principles: A guy brought a Johnson Messenger CB to my shop a few decades back. He had been told it would run on 12 VDC *and* 115 VAC - so he tried it! I never saw so many little leads sticking up from any PCB since - that once were capacitors and top hat transistors. There was enough fluff from the caps to have the chassis rated at least R-10 :->).

Power Reversal - Better Pray

"That's right, I reversed power and ground on a Sony XR-6000 AM/FM cassette car stereo. (12V negative ground). The little fellow made a stinky smell, so I assume that at least one component is cooked."

13

If it had not been turned on before you discovered your error, the damage may have been limited to the display and some filter caps. Then again... The problem is that an auto battery has a very high current capacity and any fuses respond too slowly to be of much value in a situation such as this. Any capacitors and solid state components on the 12 V bus at the time power was applied are likely fried well done. "Is there any hope of my repairing it? (This assumes I show more ability than I did when installing it.) Which part(s) are likely damaged?" Well, based on that last statement ;-)

Bad: car batteries can provide amps and amps of current (much worse than reverse connecting a wall adapter for example.) Good: The stinking might be due to a component getting too hot and vaporizing the solder paste/preserver/dust on it, but not actually giving up the ghost. I would find and check any fuses, or components directly in-line with or parallel to the power lines (the latter might include the IC's unfortunately...)

Good: There might have been a protecting diode somewhere (but why did it stink then (^_^) Neutral: Did you disassemble it to see if there were any blackened areas/components? Smell from a close distance; I can often locate a burnt component that way even after a long time.

If not, join the happy crowd, and gut the good old stereo for parts!

Determining Voltage and Polarity of AC Adapter Powered Devices

This is often required when the original adapter is lost or misplaced or isn't labeled so you are not sure if it is the correct one for your device. It's amazing how many things like modems and phone answering machines don't list the voltage and polarity on the case it's not like the extra printing would cost anything! While I would stop short of calling this a conspiracy, there does appear to be an industry-wide practice of leaving out key information to encourage replacement of the equipment rather than the much less costly and much less profitable repair or replacement of only the wall adapter. Information on voltage, current, and AC or DC polarity, is often missing on the equipment itself. And, absolutely totally incompatible wall adapters having similar plugs can be attached with the possible result being instant destruction of the device. This even applies to equipment from the same manufacturer! At least wall sockets are standardized - wall adapters are not. If you are simply replacing a broken adapter with a universal type, check the label on the old one - they almost always provide this information. There are three issues: AC versus DC, the voltage, and polarity. Unfortunately, fully determining these requirements

14

experimentally can be non-trivial. While many devices have built in protection for reverse polarity (which would probably also include putting AC into a device requiring DC), others do not and may be damaged or may at least blow an internal fuse. Few devices protect against extreme overvoltage. If you have a multimeter, there are also some tests you can perform without opening the device but they are not foolproof. Here are some general guidelines. The more of these you can confirm, the greater the confidence of avoiding disaster. 1. The best way would be to find the information without serious testing. It may be readily available: o Examine the device for labels, either embossed near the power jack or on the rear or underneath such as:
o o o o o DC 5V ---____ AC 12 V ~ _

If there is a voltage listed but no indication of AC/DC, 6 V or less is likely to be DC (and may require decent regulation; higher voltages could be either AC or DC (probably filtered but unregulated though not always). o A symmetric (non-polarized) jack means it is supposed to operate on AC. o If the device has a metal case or you can get to the metal shields on connectors, check for continuity to the power jack. This probably is the negative input (though no guarantee - some manufacturers do really strange things!). o Check your user manual! o Contact the manufacturer or their Web site. 2. The next best way would be to open it up and trace enough of the power circuitry to identify components which have obvious voltage ratings and polarities like electrolytic capacitors. There may even be labeling on the circuit board. o There will almost always be at least one electrolytic cap very near the power input. o If there is nothing between it and the power jack, then polarity will be that of the cap and you will have an upper bound on voltage (but the actual safe operating voltage will probably be considerably less). o If there is a diode in series with the cap, then the voltage and polarity will be as above (except for the 0.7 or so V diode drop) and the device is probably designed to operate on DC (and possibly AC but there may not be enough filtering). o If there is a bridge rectifier or multiple rectifier diodes between the input and any DC loads, it is probably designed to operate on AC. o If the device also has a battery compartment and the battery powers the device the same way as the adapter (possibly with one connection going through a diode or an interlock on the power jack), then the AC adapter
o

15

polarity and voltage will be the same (+/- 0.7 V or so) as the battery. However, some devices use totally different means of powering themselves with battery and AC operation! 3. If you have a multimeter for which you know the polarity of its output on the ohms ranges (VOMs may be reversed from the probes; DMMs are often the same - this can be determined by testing a diode or with another meter), then test on the low ohms range first in one direction, than the other. This is like applying a very low safe voltage to the device: o Open in one direction and a charging cap (resistance starts low and increases relatively slowly) indicates a series diode (protection or a rectifier). The probe polarity where the cap is charging is the correct one. (Note: once the internal caps charge up, reversing the leads again may result in an apparent open reading.) o A diode drop in one direction and charging cap in the other indicates a parallel protection diode. Again, the slowly charging direction is correct. o Symmetric behavior may indicate it is supposed to use AC. However, this could just mean that a filter cap is directly across the input and DC is required. Anything else will probably require you do (1) or (2). And, except for manufacturer supplied information, even these are no guarantee of anything! Once AC versus DC and polarity (if relevant) are determined, start low on voltage to see at what point the device behaves normally. Depending on design, this may be quite low compared to the recommended input voltage or very near it - no way to really know. Devices with motors and solenoids may appear to operate at relatively low voltage but fail to do the proper mechanical things reliably if at all. RF devices capable of transmitting may behave similarly when asked to transmit. Devices with more constant power requirements may operate happily at these reduced voltages. However, depending on the type of power supplies they use, running at a low voltage may also be stressful (e.g., where DC-DC converters are involved). NOTE: Some devices with microcontrollers and/or logic will require a fast power turn-on so it may be necessary to switch off and then on for each input voltage you try for proper reset. Again, determining the requirements from the manufacturer is best!

AC Adapter Modifications or Enhancements

Using AC Adapters in Series for Multiple Voltages
Where a bipolar DC power supply is needed, it is possible to create this with a pair of DC output adapters in series. Each adapter must have voltage and current ratings adequate for 16

your application. They can be used with or without external regulators . Since they are fully isolated from the AC line and each other, they can be tied together with any desired polarity and common point. The only cautions are that if one of them is unpowered for any reason (it falls out of the AC outlet!) or the current rating of one of the adapters is exceeded, then current may be forced through the other one in the wrong direction possibly damaging its electrolytic capacitors or other components. To prevent this possibility, place a rectifier like a 1N4002 (this is 1 A, use a larger one if your adapters are really huge) in REVERSE across each output. This will bypass current safely around the internal circuitry. The idea of using multiple adapters can be extended to even more outputs but this is left as an exercise for the student.

Using AC Adapters in Series for Higher or Lower Voltage

Wall adapters are totally isolated from everything (except possibly for a very high value resistor to one side of the AC line which for this purpose can be ignored) so using one set of wires as a common for the series connection won't blow anything. However, obtaining an AC adapter with the proper ratings for long term use would be a good idea. There are two cases: 1. DC output. It's probably not recommended and only the sum case (higher voltage output) is generally possible but will work (at least for testing purposes) as long as neither current rating is exceeded and both units are powered. (Else, you could end up with the wrong polarity on one of the unit's electrolytic capacitors.) 2. AC output. There should be no problem as long as the current rating of neither adapter is exceeded. Unless they are identical units, you will probably have to experiment with the phasing to get the sum or difference of the voltages WITHOUT the equipment attached! WARNING: If one of the adapters is not plugged in, high voltage (possibly even more than the normal line voltage) may appear on its exposed prongs due to the AC from the other adapters present on its output (being stepped up going the wrong way through the transformer). The voltage and available current may be enough to be dangerous in some cases. CAUTION: For the difference case, if one of the units isn't powered, you may get a HIGHER voltage than expected at the output of the series combination which may let the smoke out of your equipment. :(

17

Replacing Batteries With an AC Adapter

While most appliances that run off of internal batteries also include a socket for an wall adapter, this is not always the case. Just because there is no hole to plug one in doesn't necessarily mean that you cannot use one. The type we are considering in this discussion are plug-in wall adapter that output a DC voltage (not AC transformers). This would be stated on the nameplate. The first major consideration is voltage. This needs to be matched to the needs of the equipment. However, what you provide may also need to be well regulated for several reasons as the manufacturer may have saved on the cost of the circuitry by assuming the use of batteries:

The maximum voltage supplied by a battery is well defined. For example, 4 AA cells provide just over 6 V when new. The design of the device may assume that this voltage is never exceeded and include no internal regulator. Overheating or failure may result immediately or down the road with a wall adapter which supplies more voltage than its nameplate rating (as most do especially when lightly loaded). Most wall adapters do not include much filtering. With audio equipment, this may mean that there will be unacceptable levels of hum if used direct. There are exceptions. However, there is no way of telling without actually testing the adapter under load. The load on the power source (batteries or adapter) may vary quite a bit depending on what the device is doing. Fresh batteries can provide quite a bit of current without their voltage drooping that much. This is not always the case with wall adapters and the performance of the equipment may suffer.

Thus, the typical universal adapter found at Radio Shack and others may not work satisfactorily. No-load voltage can be much higher than the voltage at full load - which in itself may be greater than the marked voltage. Adding an external regulator to a somewhat higher voltage wall adapter is best. The other major consideration is current. The rating of the was adapter must be at least equal to the *maximum* current - mA or A - drawn by the device in any mode which lasts more than a fraction of a second. The best way to determine this is to measure it using fresh batteries and checking all modes. Add a safety factor of 10 to 25 percent to your maximum reading and use this when selecting an adapter. For shock and fire safety, any wall adapter you use should be isolated and have UL approval.

Isolation means that there is a transformer in the adapter to protect you and your equipment from direct connection to the power line. Most of the inexpensive types consist of nothing more than a transformer (and for DC types), rectifier and filter capacitor. However, if what you have weighs almost nothing and is in a tiny

18

case, it may be meant for a specific purpose like a battery chargers or rechargeable device where human contact is not possible and may not include line isolation. But, if there is a low voltage plug with exposed contacts and/or the powered equipment has exposed shields or other parts, the compact light-weight types are actually miniaturized switchmode power supplies which are functionally equivalent to the heavier, bulkier adapters and do provide line isolation. UL (Underwriters Lab) approval means that the adapter has been tested to destruction and it is unlikely that a fire would result from any reasonable internal fault like a short circuit or external fault like a prolonged overload condition.

To wire it in, you can obtain a socket like those used on appliances with external adapter inputs - from something that is lying in your junk-box or a distributor like MCM Electronics. Use one with an automatic disconnect (3 terminals) if possible. Then, you can retain the optional use of the battery. Cut the wire to the battery for the side that will be the outer ring of the adapter plug and wire it in series with the disconnect (make sure the disconnected terminal goes to the battery and the other terminal goes to the equipment). The common (center) terminal goes to other side of the battery, adapter, and equipment as shown in the example below. In this wiring diagram, it is assumed that the ring is + and the center is -. Your adapter could be wired either way. Don't get it backwards!
+--+ X V | (Inserting plug breaks connection at X) Battery (+) o------- | Adapter (+) o---------+------------------o Equipment (Ring, +) \______ o===+ Battery/ | Adapter (-) o-----------------------+----o Equipment (Center, -)

WARNING: if you do not use an automatic disconnect socket, remove the battery holder or otherwise disable it - accidentally using the wall adapter with the batteries installed could result in leakage or even an explosion! A possibly simpler alternative is to fashion a 'module' the size and shape of the battery or battery pack with screw contacts at the same locations and connect your external power supply to it. For example, a couple of pieces of wooden dowel rod about 2-1/4" long taped together with wood screws in the appropriate ends would substitute for a pair of side-by-side AA batteries. Then, you don't need to modify the Walkman or whatever at all (or at most just file a slot for the wire to exit the battery door).

Converting an AC Output Wall Adapter to DC

Where a modest source of DC is required for an appliance or other device, it may be possible to add a rectifier and filter capacitor (and possibly a regulator as well) to a wall adapter with an AC output. While many wall adapter output DC, some - modems and some phone answering machines, for example - are just transformers and output low voltage AC. 19

To convert such an adapter to DC requires the use of:

Bridge rectifier - turns AC into pulsating DC. Filter capacitor - smooths the output reducing its ripple. Regulator - produces a nearly constant output voltage.

Depending on your needs, you may find a suitable wall adapter in your junk box (maybe from that 2400 baud modem that was all the rage a couple of years ago!). The basic circuit is shown below:

Bridge Rectifier

Filter Capacitor

AC o-----+----|>|-------+---------+-----o DC (+) ~| |+ | In from +----|<|----+ | +_|_ Out to powered device AC wall | | C ___ or voltage regulator Adapter +----|>|----|--+ - | | | | AC o-----+----|<|----+------------+-----o DC (-) ~ -

Considerations:

An AC input of Vin VRMS will result in a peak output of approximately 1.4 Vin 1.4 V. The first factor of 1.4 results from the fact that the peak value of a sinusoid (the power line waveform) is 1.414 (sqrt(2)) times the RMS value. The second factor of 1.4 is due to the two diodes that are in series as part of the bridge rectifier. The fact that they are both about 1.4 is a total coincidence. Therefore, you will need to find an AC wall adapter that produces an output voltage which will result in something close to what you need. However, this may be a bit more difficult than it sounds since the nameplate rating of many wall adapters is not an accurate indication of what they actually produce especially when lightly loaded. Measuring the output is best.

Select the filter capacitor to be at least 10,000 uF per 1000 mA of output current with a voltage rating of at least 2 x Vin. This rule of thumb will result in a ripple of less than 1 V p-p which will be acceptable for many devices or where a voltage regulator is used (but may be inadequate for some audio devices resulting in some 120 Hz hum. Use a larger or additional capacitor or a regulator in such a case. Suitable components can be purchased at any electronics distributor as well as Radio Shack. The bridge rectifier comes as a single unit or you can put one together from 1N400x diodes (the x can be anything from 1 to 7 for these low voltage applications). Observe the polarity for the filter capacitor!

The following examples illustrate some of the possibilities. 20

Example 1: A typical modem power pack is rated at 12 VAC but actually produces around 14 VAC at modest load (say half the nameplate current rating). This will result in about 17 to 18 VDC at the output of the rectifier and filter capacitor. Example 2: A cordless VAC battery charger adapter might produce 6 VAC. This would result in 6 to 7 VDC at the output of the rectifier and filter capacitor.

Limiting your load to the VA ratings of the transformer should keep it from overheating. Whether you will get a decently smooth output will depend on how much filtering you have AND on the peak current available from the transformer to recharge the filter capacitors on each half-cycle. A high quality transformer (e.g. something from a manufacturer like Stancor or Thorderson that is designed with much more copper) will be much much better in this respect. A wall adapter is likely to have limited peak current and significant droop. Adding an IC regulator to either of these would permit an output of up to about 2.5 V less than the filtered DC voltage.

Adding an IC Regulator to a Wall Adapter or Battery

For many applications, it is desirable to have a well regulated source of DC power. This may be the case when running equipment from batteries as well as from a wall adapter that outputs a DC voltage or the enhanced adapter described in the section. The following is a very basic introduction to the construction of a circuit with appropriate modifications will work for outputs in the range of about 1.25 to 35 V and currents up 1 A. This can also be used as the basis for a small general purpose power supply for use with electronics experiments. For an arbitrary voltage between about 1.2 and 35 V what you want is an IC called an 'adjustable voltage regulator'. LM317 is one example - Radio Shack should have it along with a schematic. The LM317 looks like a power transistor but is a complete regulator on a chip. Where the output needs to be a common value like +5 V or -12 V, ICs called 'fixed voltage regulators' are available which are preprogrammed for these. Typical ICs have designations of 78xx (positive output) and 79xx (negative output). For example:
Positive Voltage Regulator ----------------------7805 +5 V 7809 +9 V 7812 +12 V 7815 +15 V Negative Voltage Regulator ----------------------7905 -5 V 7909 -9 V 7912 -12 V 7915 -15 V

21

and so forth. Where these will suffice, the circuit below can be simplified by eliminating the resistors and tying the third terminal to ground. Note: pinouts differ between positive and negative types - check the datasheet! Here is a sample circuit using the LM317:

I +-------+ O Vin (+) o-----+---| LM317 |---+--------------+-----o Vout (+) | +-------+ | | | | A / | | | \ R1 = 240 | | | / | ___ _|_ C1 | | +_|_ C2 |_0_| LM317 ___ .01 +-------+ ___ 1 uF | | Adjust | uF | - | |___| Output | \ | ||| Input | / R2 | 123 | \ | | | | Vin(-) o------+-------+----------------------+-----o Vout (-)

1 2 3 -

Note: Not all voltage regulator ICs use this pinout. If you are not using an LM317, double check its pinout - as well as all the other specifications. For the LM317: 1. R2 = (192 x Vout) - 240, where R2 in ohms, Vout is in volts and must be at between 1.2 V and 35 V. 2. Vin should be at least 2.5V greater than Vout. Select a wall adapter with a voltage at least 2.5 V greater than your regulated output at full load. However, note that a typical adapter's voltage may vary quite a bit depending on manufacturer and load. You will have to select one that isn't too much greater than what you really want since this will add unnecessary wasted power in the device and additional heat dissipation. 3. Maximum output current is 1 A. Your adapter must be capable of supplying the maximum current safely and without its voltage drooping below the requirement in (2) above. 4. Additional filter capacitance (across C1) on the adapter's output may help (or be required) to reduce its ripple and thus the swing of its input. This may allow you to use an adapter with a lower output voltage and reduce the power dissipation in the regulator as well.

22

Using 10,000 uF per *amp* of output current will result in less than 1 V p-p ripple on the input to the regulator. As long as the input is always greater than your desired output voltage plus 2.5 V, the regulator will totally remove this ripple resulting in a constant DC output independent of line voltage and load current fluctuations. (For you purists, the regulator isn't quite perfect but is good enough for most applications.) Make sure you select a capacitor with a voltage rating at least 25% greater than the adapter's *unloaded* peak output voltage and observe the polarity! Note: wall adapters designed as battery chargers may not have any filter capacitors so this will definitely be needed with this type. Quick check: If the voltage on the adapter's output drops to zero as soon as it is pulled from the wall even with no load - it does not have a filter capacitor. 5. The tab of the LM317 is connected to the center pin - keep this in mind because the chip will have to be on a heat sink if it will be dissipating more than a watt or so. P = (Vout - Vin) * Iout. 6. There are other considerations - check the datasheet for the LM317 particularly if you are running near the limits of 35 V and/or 1 A.

Troubleshooting and Repair of Batteries/Battery Packs

Battery Technology
The desire for portable power seems to be increasing exponentially with the proliferation of notebook and palmtop computers, electronic organizers, PDAs, cellular phones and faxes, pagers, pocket cameras, camcorders and audio cassette recorders, boomboxes - the list is endless. Two of the hottest areas in engineering these days are in developing higher capacity battery technologies (electrochemical systems) for rechargeable equipment and in the implementation of smart power management (optimal charging and high efficiency power conversion) for portable devices. Lithium and Nickel Metal Hydride are among the more recent additions to the inventory of popular battery technologies. A variety of ICs are now available to implement rapid charging techniques while preserving battery life. Low cost DC-DC converter designs are capable of generating whatever voltages are required by the equipment at over 90% efficiency However, most of the devices you are likely to encounter still use pretty basic battery technologies - most commonly throwaway Alkaline and Lithium followed by rechargeable Nickel Cadmium or Lead-Acid. The charging circuits are often very simple and don't really do the best job but it is adequate for many applications.

23

For more detailed information on all aspects of battery technology, see the articles at: Many major battery manufacturers have extensive technical information on their Web sites, though not all of it may be unbiased. There is more on batteries than you ever dreamed of ever needing. The sections below represent just a brief introduction.

Battery Basics
A battery is, strictly speaking, made up of a number of individual cells (most often wired in series to provide multiples of the basic cell voltage for the battery technology - 1.2, 1.5, 2.0, or 3.0 V are most common). However, the term is popularly used even for single cells. Four types of batteries are typically used in consumer electronic equipment: 1. Alkaline - consisting of one or more primary cells with a nominal terminal voltage of 1.5 V. Examples are AAA, AA, C, D, N, 9V ('transistor'), lantern batteries (6V or more), etc. There are many other available sizes including miniature button cells for specialty applications like clocks, watches, calculators, and cameras. In general recharging of alkaline batteries is not practical due to their chemistry and construction. Exceptions which work (if not entirely consistently as of this writing) are the rechargeable Alkalines (e.g., 'Renewals'). Advantages of alkalines are high capacity and long shelf life. These now dominate the primary battery marketplace largely replacing the original carbonzinc and heavy duty types. Note that under most conditions, it not necessary to store alkaline batteries in the 'fridge to obtain maximum shelf life. 2. Lithium - these primary cells have a much higher capacity than alkalines. The terminal voltage is around 3 volts per cell. These are often used in cameras where their high cost is offset by the convenience of long life and compact size. Lithium batteries in common sizes like 9V are beginning to appear. In general, I would not recommend the use of lithiums for use in applications where a device can be accidentally left on - particularly with kids' toys. Your batteries will be drained overnight whether a cheap carbon zinc or a costly lithium. However, for smoke alarms, the lithium 9V battery (assuming they hold up to their longevity claims) is ideal as a 5-10 year service life without attention can be expected. 3. Nickel Cadmium (NiCd) - these are the most common type of rechargeable battery technology use in small electronic devices. They are available in all the poplar sizes. However, their terminal voltage is only 1.2 V per cell compared to 1.5 V per cell for alkalines (unloaded). This is not the whole story, however, as NiCds terminal voltage holds up better under load and as they are discharged. Manufacturers claim 500-1000 charge-discharge cycles but expect to achieve these optimistic ratings only under certain types of applications. In particular it is usually recommended that NiCds should not be discharged below about 1 V per cell and should not be left in a discharged state for too long. Overcharging is also an enemy of NiCds and will reduce their ultimate life. An electric shaver is an

24

example of a device that will approach this cycle life as it is used until the battery starts to poop out and then immediately put on charge. If a device is used and then neglected (like a seldom used printing calculator), don't be surprised to find that the NiCd battery will not charge or will not hold a charge next time the calculator is used. 4. Nickel-Metal-Hydride (NiMH) - These are gradually replacing NiCds in portable devices. They have the same terminal voltage (1.2 V per cell) and are charged in a similar manner to NiCds. Advantages include a supposed total lack of any voltage depression or memory effect. Disadvantages include a higher self discharge rate and less tolerance of overcharging. NiMH cells CANNOT be charged using the same fast charger or built in charger of typical devices that were designed for NiCds. 5. Lead Acid - similar to the type used in your automobile but generally specially designed in a sealed package which cannot leak acid under most conditions. These come in a wide variety of capacities but not in standard sizes like AA or D. They are used in some camcorders, flashlights, CD players, security systems, emergency lighting, and many other applications. Nominal terminal voltage is 2.0 V per cell. These batteries definitely do not like to be left in a discharged condition (even more so than NiCds) and will quickly become unusable if left that way for any length of time.

Battery Chargers
The (energy storage) capacity, C, of a battery is measured in ampere hours denoted a A-h (or mA-h for smaller types). The charging rate is normally expressed as a fraction of C e.g., .5 C or C/2. In most cases, trickle charging at a slow rate - C/100 to C/20 - is easier on batteries. Where this is convenient, you will likely see better performance and longer life. Such an approach should be less expensive in the long run even if it means having extra cells or packs on hand to pop in when the others are being charged. Fast charging is hard on batteries - it generates heat and gasses and the chemical reactions may be less uniform. Each type of battery requires a different type of charging technique. 1. NiCd batteries are charged with a controlled (usually constant) current. Fast charge may be performed at as high as a .5-1C rate for the types of batteries in portable tools and laptop computers. (C here is the amp-hour capacity of the battery. A .5C charge rate for a 2 amp hour battery pack would use a current equal to 1 A, for example.) Trickle charge at a 1/20-1/10C rate. Sophisticated charges will use a variety of techniques to sense end-of-charge. Inexpensive chargers (and the type in many cheap consumer electronics devices) simply trickle charge at a constant current. Rapid chargers for portable tools, laptop computers, and camcorders, do at least sense the temperature rise which is one indication of having reached full charge but this is far from totally reliable and some damage is probably unavoidable as

25

some cells reach full charge before others due to slight unavoidable differences in capacity. Better charging techniques depend on sensing the slight voltage drop that occurs when full charge is reached but even this can be deceptive. The best power management techniques use a combination of sensing and precise control of charge to each cell, knowledge about the battery's characteristics, and state of charge. While slow charging is better for NiCds, long term trickle charging is generally not recommended. Problems with simple NiCd battery chargers are usually pretty easy to find - bad transformer, rectifiers, capacitors, possibly a regulator. Where temperature sensing is used, the sensor in the battery pack may be defective and there may be problems in the control circuits as well. However, more sophisticated power management systems controlled by microprocessors or custom ICs and may be impossible to troubleshoot for anything beyond obviously bad parts or bad connections. 2. Lead acid batteries are charged with a current limited but voltage cutoff technique. Although the terminal voltage of a lead-acid battery is 2.00 V per cell nominal, it may actually reach more than 2.5 V per cell while charging. For an automotive battery, 15 V is still within the normal range of voltages to be found on the battery terminals when the engine (and alternator) are running. A simple charger for a lead-acid battery is simply a stepped down rectified AC source with some resistance to provide current limiting. The current will naturally taper off as the battery voltage approaches the peaks of the charging waveform. This is how inexpensive automotive battery chargers are constructed. For small sealed lead-acid batteries, an IC regulator may be used to provide current limited constant voltage charging. A 1 A (max) charger for a 12 V battery may use an LM317, 3 resistors, and two capacitors, running off of a 15 V or greater input supply. Trickle chargers for lead-acid batteries are usually constant voltage and current tapers off as the battery reaches full charge. Therefore, leaving the battery under constant charge is acceptable and will maintain it at the desired state of full charge. Problems with lead-acid battery chargers are usually pretty easy to diagnose due to the simplicity of most designs.

Comments on Building Charger for Small Lead-Acid Batteries

The following applies to the sort of lead-acid batteries found in some camcorders and other portable equipment:

26

The simple way is to build a power supply that outputs 13.8 volts regulated, with a current limit of 0.5 A. 13.8 V can be left connected to the battery forever without damage - this is called a float charge. The 0.5 A current limit protects the battery from drawing too much current and overheating if it's been deeply discharged. This sort of charger will get the battery back up to 80% charge within a few hours, so it's fine for most uses. However, when designing it, make sure the charger doesn't self-destruct if the input voltage goes away (due to AC power failure) while still connected to the battery. With a standard series regulator, when the input power fails the whole battery voltage gets applied to the base- emitter junction of the output transistor in reverse. Many transistors are only specified to withstand about 6 V reverse base-emitter voltage, so with this design your charger will be toast at the first power failure. If you want higher-performance charging, there are special charge controller chips that provide 3 or more charge phases. They are:

Constant current charge at maximum safe current (see battery spec sheet) until the voltage rises to about 14.5 V. Constant voltage charge at 14.5 V until the current drops to a fraction of the initial current limit. Float charge at 13.65 V after that.

By using the 14.5 V instead of 13.8 V for the initial charge voltage, this type of charger gets the battery back up to 90% charged in considerably less time. But if you only care about charging overnight, you don't need the extra complexity. On the other hand, NiCd batteries can safely be charged in less than an hour with suitable electronics. Lead-acid simply can't be recharged that fast.

Substituting NiCds for Alkalines

First note that rechargeable batteries are NOT suitable for safety critical applications like smoke detectors unless they are used only as emergency power fail backup (the smoke detector is also plugged into the AC line) and are on continuous trickle charge). NiCds self discharge (with no load) at a rate which will cause them to go dead in a month or two. For many toys and games, portable phones, tape players and CD players, and boomboxes, TVs, palmtop computers, and other battery gobbling gadgets, it may be possible to substitute rechargeable batteries for disposable primary batteries. However, NiCds have a lower terminal voltage - 1.2V vs. 1.5V - and some devices will just not be happy. In particular, tape players may not work well due to this reduced voltage not being able to power the motor at a constant correct speed. Manufacturers may specifically warn against their use. Flashlights will not be as bright unless the light bulb is also replaced with a lower voltage type. Other equipment may perform poorly or fail to operate entirely on NiCds. When in doubt, check your instruction manual. And, there is a slight, but nonzero chance that some equipment may actually be damaged. This might occur if its

27

design assumed something about the internal resistance of the batteris; the resistance is much lower for NiCds than Alkalines.

Can a Large Electrolytic Capacitor be Substituted for a NiCd?

The quick answer is: probably not. The charger very likely assumes that the NiCds will limit voltage. The circuits found in many common appliances just use a voltage source significantly higher than the terminal voltage of the battery pack through a current limiting resistor. If you replace the NiCd with a capacitor and the voltage will end up much higher than expected with unknown consequences. For more sophisticated chargers, the results might be even more unpredictable. Furthermore, even a SuperCap cannot begin to compare to a small NiCd for capacity. A 5.5 V 1 F (that's Farad) capacitor holds about 15 W-s of energy which is roughly equivalent to a 5 V battery of 3 A-s capacity - less than 1 mA-h. A very tiny NiCd pack is 100 mA-h or two orders of magnitude larger.

Determining the Actual Capacity of a NiCd Battery Pack

When a battery pack is not performing up to expectations or is not marked in terms of capacity, here are some comments on experimentally determining the A-h rating. When laying eggs, start with a chicken. Actually, you have to estimate the capacity so that charge and discharge rates can be approximated. However, this is usually easy to do with a factor of 2 either way just be size:
Size of cells Capacity range, A-h --------------------------------------------AAA .2 - .4 AA .4 - 1 C 1 - 2 D 1 - 5 Cordless phone .1 - .3 Camcorder 1 - 3+ Laptop computer 1 - 5+

First, you must charge the battery fully. For a battery that does not appear to have full capacity, this may be the only problem. Your charger may be cutting off prematurely due to a fault in the charger and not the battery. This could be due to dirty or corroded contacts on the charger or battery, bad connections, faulty temperature sensor or other end-of-charge control circuitry. Monitoring the current during charge to determine if the battery is getting roughly the correct A-h to charge it fully would be a desirable first step. Figure about 1.2 to 1.5 times the A-h of the battery capacity to bring it to full charge. Then discharge at approximately a C/20 - C/10 rate until the cell voltages drops to about 1 V (don't discharge until flat or damage may occur). Capacity is calculated as average current x elapsed time since the current for a NiCd will be fairly constant until very near the end.

28

NiCd Batteries and the Infamous 'Memory Effect'

Whether the NiCd 'memory effect' is fact or fiction seems to depend on one's point of view and anecdotal evidence. What most people think is due to the memory effect is more accurately described as voltage depression - reduced voltage (and therefore, reduced power and capacity) during use. The following are the most common causes of application problems wrongly attributed to 'memory': 1. Cutoff voltage too high - basically, since NiCds have such a flat voltage vs. discharge characteristic, using voltage sensing to determine when the battery is nearly empty can be tricky; an improper setting coupled with a slight voltage depression can cause many products to call a battery "dead" even when nearly the full capacity remains usable (albeit at a slightly reduced voltage). 2. High temperature conditions - NiCds suffer under high-temp conditions; such environments reduce both the charge that will be accepted by the cells when charging, and the voltage across the battery when charged (and the latter, of course, ties back into the above problem). 3. Voltage depression due to long-term overcharge - Self-explanatory. NiCds can drop 0.1-0.15 V/cell if exposed to a long-term (i.e., a period of months) overcharge. Such an overcharge is not unheard-of in consumer gear, especially if the user gets in the habit of leaving the unit in a charger of simplistic design (but which was intended to provide enough current for a relatively rapid charge). As a precaution, I do NOT leave any of my NiCd gear on a charger longer than the recommended time UNLESS the charger is specifically designed for long-term "trickle charging", and explicitly identified as such by the manufacturer. 4. There are a number of other possible causes listed in a "miscellaneous" category; these include o Operation below 0 degrees C. o High discharge rates (above 5C) if not specifically designed for such use. o Inadequate charging time or a defective charger. o One or more defective or worn-out cells. They do not last forever. To close with a quote from the GE note: "To recap, we can say that true 'memory' is exceedingly rare. When we see poor battery performance attributed to 'memory', it is almost always certain to be a correctable application problem. Of the problems noted above, Voltage Depression is the one most often mistaken for 'memory'..... This information should dispel many of the myths that exaggerate the idea of a 'memory' phenomenon."

Memory Effect in NiMH Batteries?

The party line is that Nickel-Metal-Hydride batteries do not have any memory effect. Perhaps, perhaps not. 29

Care and Feeding of NiCds

Here are six guidelines to follow which will hopefully avoid voltage depression or the memory effect or whatever: (Portions of the following guidelines are from the NiCd FAQ written by: Ken A. Nishimura (KO6AF)) 1. DON'T deliberately discharge the batteries to avoid memory. You risk reverse charging one or more cell which is a sure way of killing them. 2. DO let the cells discharge to 1.0V/cell on occasion through normal use. 3. DON'T leave the cells on trickle charge for long times, unless voltage depression can be tolerated. 4. DO protect the cells from high temperature both in charging and storage. 5. DON'T overcharge the cells. Use a good charging technique. With most inexpensive equipment, the charging circuits are not intelligent and will not terminate properly - only charge for as long as recommended in the user manual. 6. DO choose cells wisely. Sponge/foam plates will not tolerate high charge/discharge currents as well as sintered plate. Of course, it is rare that this choice exists. All of which tends to support my basic operating theory about the charging of nickelcadmium batteries: 1. Man is born in sin and must somehow arrange for the salvation of his immortal soul. 2. All nickel-cadmium batteries must be recharged. 3. There is no proper method of performing either task (1) or task (2) to the satisfaction of anyone.

Nickel Cadmium Versus Nickel-Metal-Hydride in a Nutshell

NiCds are inexpensive, reliable, and easy to charge, but may suffer from voltage depression (what people call the memory effect) from repeated shallow discharge cycles. NiMHs have slightly higher capacity and no memory effect but have higher initial cost and are more sensitive to overcharging. Must be used with compatible charger.

First Aid for NiCd Battery Packs

CAUTION: Opening these battery packs will of course void any warranty but you knew that. Also, make notes of exactly how the cells and anything else inside is arranged. Improper reassembly can result in damage to equipment and/or risk of overheating should cells short inside the pack due to lack of or misplaced insulation. Under no circumstances should all thermal switches be removed - not only are they a safety device to prevent excessive temperatures but may also be part of the charging circuit. So, if they are removed, your next charge may be your last! I'd highly recommend that all of them be

30

replaced (from another pack as a last resort) and installed in exactly the same positions they were originally. Many "name brand" camcorder and other similar battery packs contain two or even 3 thermal switches (those rectangular, un-identifiable, wired between the cells). They contain a bimetal strip operating a set of contacts which open at a preset temperature. Often only one of these will fail, resulting in a $40 NiCad that won't charge. Since these little suckers are pricey if ya kind find them, a safe and cheap fix, is to test the thermal switches for continuity (they should be closed at room temp) and remove the defective one. If needed move the other, or at least one, to the mid-point of the cells series. If a battery pack has 8 separate cells, (i.e.: a 9.6 V VHS-C camera pack) the thermal switch should be wired between the 4th and 5th, and as far away from the charging contacts as possible. The extra switches were added as a safety factor but since the average one is designed to open at 87C, there is no fire hazard so long as the pack is re-sealed after working on it. A quick fix for a NiCad pack left on the dashboard. Since good ol' solar power can heat a battery pack to the point where the thermal protection can open (and even warp a case) you can be stuck at the soccer game with what seems like dead batteries. The trick is to drop the temp below 87C. Wrap the battery in plastic so the contacts won't get wet, and stick it in the cooler with the kids lunch and your six-pack. A few minutes and the thermistor should close. letting the batter work normally. Also, if the cord is long enough, never recharge a NiCad inside the car. Place the battery and charger under the car, in the shade, so it doesn't heat quickly and will get a full charge.

Identifying Technology of Unmarked Battery Packs

Since the nominal (rated) voltages for the common battery technologies differ, it is often possible to identify which type is inside a pack by the total output voltage:

NiCd packs will be a multiple of 1.2 V. Lead-acid packs will be a multiple of 2.0 V. Alkaline packs will be a multiple of 1.5.

Note that these are open circuit voltages and may be very slightly higher when fully charged or new. Therefore, it is generally easy to tell what kind of technology is inside a pack even if the type is not marked as long as the voltage is marked. Of course, there are some - like 6 V that will be ambiguous.

Powering LEDs with Batteries

LEDs look like diodes with a high forward voltage drop. Above the that voltage, the incremental resistance is very low and without current limiting, the current would be critically dependent on the exact voltage of the power source. Most of the time, they are spec'd at a particular maximum current and need some means to limit the current to that value based on the input voltage. Some devices may depend on the internal resistance of 31

the batteries to provide the current limiting - this is a poor approach and depends greatly on the type and capacity of the batteries being used. Most common is just a resistor but this provides no regulation and poor efficiency. Better designs (used in LED flashlights) will use a DC to pulse inverter with regulation achieving constant light output regardless of battery state-of-charge and high efficiency. LEDs can usually withstand short high current pulses and this allows the circuit to be designed with low losses. The specifications for LEDs you see in electronics distributor's catalogs may look the same as those for incandescent lamps but they are not. Incandescent lamps provide their own current limiting; LEDs do not. It's possible to luck out and happen to have a given LED work without current limiting with a particular set of batteries but it hardly an acceptable design approach. Slight variations in battery parameters will result in gross changes in light intensity and possible shortening of life or outright destruction of the LED.

Battery Problem Troubleshooting and Repair

Problems with Battery Operated Equipment
For primary batteries like Alkalines, first try a fresh set. For NiCds, test across the battery pack after charging overnight (or as recommended by the manufacturer of the equipment). The voltage should be 1.2 x n V where n is the number of cells in the pack. If it is much lower - off by a multiple of 1.2 V, one or more cells is shorted and will need to be replaced or you can attempt zapping it to restore the shorted cells. If the voltage drops when the device is turned on or the batteries are installed - and the batteries are known to be good - then an overload may be pulling the voltage down. Assuming the battery is putting out the proper voltage, then a number of causes are possible: 1. Corroded contacts or bad connections in the battery holder. 2. Bad connections or broken wires inside the device. 3. Faulty regulator in the internal power supply circuits. Test semiconductors and IC regulators. 4. Faulty DC-DC inverter components. Test semiconductors and other components. 5. Defective on/off switch (!!) or logic problem in power control. 6. Other problems in the internal circuitry.

32

NiCd Battery Pack Will Not Hold a Charge

This applies if the pack appears to charge normally and the terminal voltage immediately after charging is at least 1.2 x n where n is the number of cells in the pack but after a couple of days, the terminal voltage has dropped drastically. For example, a 12 V pack reads only 6 V 48 hours after charging without being used. What is most likely happening is that several of the NiCd cells have high leakage current and drain themselves quite rapidly. If they are bad enough, then a substantial fraction of the charging current itself is being wasted so that even right after charging, their capacity is less than expected. However, in many cases, the pack will deliver close to rated capacity if used immediately after charging. If the pack is old and unused or abused (especially, it seems, if it is a fast recharge type of pack), this is quite possible. The cause is the growth of fine metallic whiskers called dendrites that partially shorts the cell(s). If severe enough, a dead short is created and no charge at all is possible. Sometimes this can be repaired temporarily at least by 'zapping' using a large charged capacitor to blow out the whiskers or dendrites that are causing the leakage (on a cell-bycell basis) but my success on these types of larger or high charge rate packs such as used in laptop computers or camcorders has been less than spectacular. See the section: Zapping NiCds to Clear Shorted Cells.

What is This Thing in my NiCd Battery Pack?

In addition to the NiCd cells, you will often find one or more small parts that are generally unrecognizable. Normally, you won't see these until you have a problem and, ignoring all warnings, open the pack. If it is a little rectangular silver or plastic box in series with one of the positive or negative terminals of the pack, it is probably a thermostat and is there to shut down the charging or discharging if the temperature of the pack rises too high. (The manufacturer name "Klixon" would be a dead giveaway to identity. Izuzu also makes these things.) If it tests open at room temperature, it is bad. With care, you can safely substitute a low value resistor or auto tail light bulb and see if the original problem goes away or at least the behavior changes. However, if there is a dead short somewhere, that device may have sacrificed its life to protect your equipment or charger and going beyond this (like shorting it out entirely) should be done with extreme care. These may be either mechanical (bimetal strip/contacts) or solid state (Polyfuse(tm) - increases resistance with overcurrent). If it looks like a small diode or resistor, it could be a temperature sensing thermistor which is used by the charger to determine that the cells are heating which in its simple minded way means the cells are being overcharged and it is should quit charging them. You can try using a resistor in place of the thermistor to see if the charger will now cooperate. Try a variety of values while monitoring the current or charge indicators.

33

However, the problem may actually be in the charger controller and not the thermistor. The best approach is to try another pack. It could be any of a number of other possible components but they all serve a protective and/or charge related function. Of course, the part may be bad due to a fault in the charger not shutting down or not properly limiting the current as well.

Zapping NiCds to Clear Shorted Cells

Nickel-Cadmium batteries that have shorted cells can sometimes be rejuvenated - at least temporarily - by a procedure affectionately called 'zapping'. The cause of these bad NiCd cells is the formation of conductive filaments called whiskers or dendrites that pierce the separator and short the positive and negative electrodes of the cell. The result is either a cell that will not take a charge at all or which self discharges in a very short time. A high current pulse can sometimes vaporize the filament and clear the short. The result may be reliable particularly if the battery is under constant charge (float service) and/or is never discharged fully. Since there are still holes in the separator, repeated shorts are quite likely especially if the battery is discharged fully which seems to promote filament formation, I have used zapping with long term reliability (with the restrictions identified above) on NiCds for shavers, Dustbusters, portable phones, and calculators. WARNING: There is some danger in the following procedures as heat is generated. The cell may explode! Take appropriate precautions and don't overdo it. If the first few attempts do not work, dump the battery pack. Attempt sapping at your own risk!!! You will need a DC power supply and a large capacitor - one of those 70,000 uF 40 V types used for filtering in multimegawatt geek type automotive audio systems, for example. A smaller capacitor can be tried as well. Alternatively, a you can use a 50 to 100 A 5 volt power supply that doesn't mind (or is protected against) being overloaded or shorted. Some people recommend the use of a car battery for NiCd zapping. DO NOT be tempted - there is nearly unlimited current available and you could end with a disaster including the possible destruction of that battery, your NiCd, you, and anything else that is in the vicinity. OK, you have read the warnings:

34

Remove the battery pack from the equipment. Gain access to the shorted cell(s) by removing the outer covering or case of the battery pack and test the individual cells with a multimeter. Since you likely tried charging the pack, the good cells will be around 1.2 V and the shorted cells will be exactly 0 V. You must perform the zapping directly across each shorted cell for best results. Connect a pair of heavy duty clip leads - #12 wire would be fine - directly across the first shorted cell. Clip your multimeter across the cell as well to monitor the operation. Put it on a high enough scale such that the full voltage of your power supply or capacitor won't cause any damage to the multimeter. Wear your eye protection!!! 1. Using the large capacitor: o Charge the capacitor from a current limited 12-24 V DC power supply. o Momentarily touch the leads connected across the shorted cell to the charged capacitor, + to +, - to -. CAUTION: Polarity is critical - do it backwards and you will make the problem worse, probably terminal. There will be sparks. The voltage on the cell may spike to a high value up to the charged voltage level on the capacitor. The capacitor will discharge almost instantly. 2. Using the high current power supply: o Turn on the supply. o Momentarily touch the leads connected across the shorted cell to the power supply output, + to +, - to -. CAUTION: Polarity is critical - do it backwards and you will make the problem worse, probably terminal. There will be sparks. DO NOT maintain contact for more than a couple of seconds. The NiCd may get warm! While the power supply is connected, the voltage on the cell may rise to anywhere up to the supply voltage. Now check the voltage on the (hopefully previously) shorted cell. If the dendrites have blown, the voltage on the cell should have jumped to anywhere from a few hundred millivolts to the normal 1 V of a charged NiCd cell. If there is no change or if the voltage almost immediately decays back to zero, you can try zapping couple more times but beyond this is probably not productive. If the voltage has increased and is relatively stable, immediately continue charging the repaired cell at the maximum SAFE rate specified for the battery pack. Note: if the other cells of the battery pack are fully charged as is likely if you had attempted to charge the pack, don't put the entire pack on high current charge as this will damage the other cells through overcharging. One easy way is to use your power supply with a current limiting resistor connected just to the cell you just zapped. A 1/4 C rate should be safe and effective but avoid

35

overcharging. Then trickle charge at the 1/10 C rate for several hours. (C here is the amphour capacity of the cell. Therefore, a 1/10 C rate for a 600 mA NiCd is 50 mA.) This works better on small cells like AAs than on C or D cells since the zapping current requirement is lower. Also, it seems to be more difficult to reliably restore the quick charge type battery packs in portable tools and laptop computers that have developed shorted cells (though there are some success stories). My experience has been that if you then maintain the battery pack in float service (on a trickle charger) and/or make sure it never discharges completely, there is a good chance it will last. However, allow the bad cells to discharge to near 0 volts and those mischievous dendrites will make their may through the separator again and short out the cell(s).

Comments on NiCd Care and Feeding

Measuring NiCd capacity - I use a very simple/effective system. Put a 2.5 ohm resistor across the contacts of a cheap travel analog clock, which will time the rundown. It is quite consistent for good cells. A good typical AA NiCd will run one hour. NiCd zapping - I use a 1 ohm power resistor in series with a car battery, though a series headlight will also work. I charge for about 30 secs or until warm, which will clear the whisker and put in enough charge to see if the cell is salvageable.

Battery Juice and Corroded Contacts

Unless you have just arrived from the other side of the galaxy (where such problems do not exist), you know that so-called 'leak-proof' batteries sometimes leak. This is a lot less common with modern technologies than with the carbon-zinc cells of the good old days, but still can happen. It is always good advice to remove batteries from equipment when it is not being used for an extended period of time. Dead batteries also seem to be more prone to leakage than fresh ones (in some cases because the casing material is depleted in the chemical reaction which generates electricity and thus gets thinner or develops actual holes). In most cases, the actual stuff that leaks from a battery is not 'battery acid' but rather some other chemical. For example, alkaline batteries are so called because their electrolyte is an alkaline material - just the opposite in reactivity from an acid. Usually it is not particularly reactive (but isn't something you would want to eat). The exception is the lead-acid type where the liquid inside is sulfuric acid of varying degrees of strength depending on charge. This is nasty and should be neutralized with an alkaline material like baking soda before being cleaned up. Fortunately, these sealed leadacid battery packs rarely leak (though I did find one with a scary looking bulging case, probably due to overcharging - got rid of that is a hurry).

36

Scrape dried up battery juice from the battery compartment and contacts with a plastic or wooden stick and/or wipe any liquid up first with a dry paper towel. Then use a damp paper towel to pick up as much residue as possible. Dispose of the dirty towels promptly. If the contacts are corroded, use fine sandpaper or a small file to remove the corrosion and brighten the metal. Do not an emery board or emery paper or steel wool as any of these will leave conductive particles behind which will be difficult to remove. If the contacts are eaten through entirely, you will have to improvise alternate contacts or obtain replacements. Sometimes the corrosion extends to the solder and circuit board traces as well and some additional repairs may be needed - possible requiring disassembly to gain access to the wiring.

Exploding Batteries - These Things Really Happen

When I was about 10 years old I was sitting in my dad's driveway in a '65 Plymouth Fury III station wagon while he disconnected the trickle charger from the '67 Fiat in the garage. I heard a pop and saw my dad throw his hands over his face, run to the back door and start kicking it to get someone to open it. Fortunately he wasn't injured. But it was an eye opener. It was probably 30 or below, there was no flame present, and the double garage door was open (this happened in Connecticut). Also in a Fiat 850 sport coupe the battery is in the trunk (front) so there really isn't anything up there that would cause a spark (engine & gas tank in back). So it must have been a spark off of the charger when he pulled it off the terminal (he hadn't unplugged the charger).

Soldering Tabs Onto NiCd Batteries

When replacing NiCd batteries in packs or portable tools, it is often necessary to attach wires to the individual cells. It may be possible to obtain NiCds with solder tabs attached (Radio Shack has these) but if yours do not, here are two ways that work. They both require a (Weller) high wattage soldering gun. I use a high power Weller (140 W) soldering gun. Use fine sandpaper to thoroughly clean and roughen up the surface of the battery cell at both ends. Tin the wires ahead of time as well. Arrange the wire and cell so that they are in their final position - use a vise or clamp or buddy to do this. Heat up the soldering gun but do not touch it to the battery until it is hot - perhaps 10 seconds. Then, heat the contact area on the battery end while applying solder. It should melt and flow quite quickly. As soon as the solder adheres to the battery, remove the heat without moving anything for a few seconds. Inspect and test the joint. A high power soldering iron can also be used. Here is a novel approach that appears to work: There is really no great amount of danger spot welding tabs! They usually are made of pure nickel material. I put two sharp pointed copper wires in a soldering gun, place both on the tab in contact with the battery case and pull the trigger for a short burst. The battery remains cool.

37

Of course! A soldering gun is a source of about 1.5 V at 100 A RMS. Should make a fine spot-welder. You should write that up for QST ("Hints and Kinks") or better yet, send it in a letter to the editor of "Electronics Now" (the magazine I write for).

Battery Related Information

Automotive Power
While it is tempting to want to use your car's battery as a power source for small portable appliances, audio equipment, and laptop computers, beware: the power available from your car's electrical system is not pretty. The voltage can vary from 9 (0 for a dead battery) to 15 V under normal conditions and much higher spikes or excursions are possible. Unless the equipment is designed specifically for such power, you are taking a serious risk that it will be damaged or blown away. Furthermore, there is essentially unlimited current available from the battery (cigarette lighter) - 20 A or more. This will instantly turn your expensive CD player to toast should you get the connections wrong. No amount of internal protection can protect equipment from fools. My recommendation for laptop computers is to use a commercially available DC-AC inverter with the laptop's normal AC power pack. This is not the most efficient but is the safest and should maintain the laptop's warranty should something go wrong. For CD players and other audio equipment, only use approved automotive adapters.

How Do Those On-Battery or On-the-Package Battery Testers Work?

There is a graded width resistance element that gets connected when you pinch those two points. It heats up - substantially, BTW. Some sort of liquid crystal or other heat sensitive material changes from dark to clear or yellow at a fairly well defined temperature. Incidentally, since the current is significant, repeated 'testing' will drain the batteries - as with any proper under-load battery test! This isn't an issue for occasional testing but if the kids figure how to do this.... Personally, I would rather use a $3 battery checker instead of paying for throw-away frills!

Battery Eliminator for Laptop or Appliance with Dead NiCds

Even where you have the AC adapter, it is quite likely that simply removing the (shorted) battery pack will not allow you to use it. This is because it probably uses the battery as a smoothing capacitor. You cannot simply replace the battery with a large electrolytic capacitor because the battery also limits the voltage to a value determined by the number of cells in the pack. Without it, the voltage would be much too high, possibly resulting in

38

damage. You could use N power diodes in series (i.e., N=Vb/.7) to drop the approximate voltage of the battery pack AND a large capacitor but you would be wasting a lot of power in the form of heat. One alternative is to substitute a regulated power supply with an output equal to the the battery voltage and current capacity found by dividing the VA rating of the normal wall adapter by the battery's nominal terminal voltage (this will be worst case - actual requirements may be less). Connect this directly in place of the original battery pack. Unless there is some other sort of interlock, the equipment should be perfectly happy and think it is operating from battery power!

39