Schematic Reference

More and more electronic devices are portable and run off batteries. It is no surprise, then, that so many flat batteries find their way into the bin - and often far too early. When a set of batteries can no longer run some device - for example, a flashgun - the cells are not necessarily completely discharged. If you put an apparently unserviceable AA-size cell into a radio-controlled clock with an LCD display it will run for months if not years. Of course not every partially discharged cell can be put in a clock. The circuit presented here lets you squeeze the last Watt-second out of your batteries, providing a bright ‘night light’ - for free!

The circuit features a TBA820M, a cheap audio power amplifier capable of operating from a very low supply voltage. Here it is connected as an astable multivibrator running at a frequency of around 13 kHz. Together with the two diodes and electrolytic capacitor this forms a DC-DC converter which can almost double the voltage from between four and eight series-connected AA-, C- or D-size cells, or from a PP3-style battery. The DC-DC converter is followed by a constant current source which drives the LED. This protects the expensive white LED: the voltages obtained from old batteries can vary considerably.

With the use of the DC-DC converter and 20 mA constant current source a much greater range of usable input voltages is achieved, particularly helpful at the lower end of the range when old batteries are used. With the constant current source on its own the white LED would not be adequately bright when run from low voltages. An additional feature is the ‘automatic eye’. The LDR detects when the normal room lighting is switched on or when the room is lit by sunlight: its resistance decreases. This reduces the UBE of the transistor below 0.7 V, the BC337 turns off and deactivates the LED.

This prolongs further the life of the old batteries. A further LDR across capacitor C reduces the quiescent current of the circuit to just 4mA (at 4V). Light from the white LED must of course not fall on the LDR, or the current saving function will not work.

The LP2951 regulator is manufactured by National Semiconductors. The choice of values is from an application note "Battery Charging", written by Chester Simpson. Diode D1 can be any diode from the 1N00x series, whichever is conveniently available. It functions as a blocking diode, to prevent a back flow of current from the battery into the LP2951 when the input voltage is disconnected. Charging current is about 100+mA, which is the internally-limited maximum current of the LP2951. For those wondering, this is compatible with just about any single-cell li-ion battery since li-ion can generally accept a charging current of up to about 1c (i.e. charging current in mA equivalent to their capacity in mAh, so a 1100mAh li-ion cell can be charged at up to 1100mA and so on).

A lower charging current just brings about a correspondingly longer charge time. IMHO 100mA is quite low, low enough that the circuit can be used for an overnight charger for many typical single-cell li-ion batteries. The resistors are deliberately kept at large orders of magnitude (tens/hundred Kohm and Mohm range) to keep the off-state current as low as possible, at about 2?A. Resistor tolerances should be kept at 1% for output voltage accuracy. The 50k pot allows for an output voltage range between 4.08V to 4.26V - thus allowing calibration as well as a choice between a charging voltage of 4.1V or 4.2V depending on the cell to be charged. The capacitors are for stability, especially C2 which prevents the output from ringing/oscillating.

Parts List:

IC1 = LP2951, voltage regulator
D1 = 1N4002, General purpose diode
R1 = 2M, 1%, metal-film
R2 = 806K, 1%, metal-film
P1 = 50K, potentiometer
C1 = 0.1uF, polyester
C2 = 2.2uF/16V, electrolytic
C3 = 330pF, ceramic

The above pictured schematic diagram is just a standard constant current model with a added current limiter, consisting of Q1, R1, and R4. The moment too much current is flowing biases Q1 and drops the output voltage. The output voltage is: 1.2 x (P1+R2+R3)/R3 volt. Current limiting kicks in when the current is about 0.6/R1 amp. For a 6-volt battery which requires fast-charging, the charge voltage is 3 x 2.45 = 7.35 V. (3 cells at 2.45v per cell). So the total value for R2 + P1 is then about 585 ohm. For a 12 V battery the value for R2 + P1 is then about 1290 ohm. For this power supply to work efficiently, the input voltage has to be a minimum of 3V higher than the output voltage. P1 is a standard trimmer potentiometer of sufficient watt for your application. The LM317 must be cooled on a sufficient (large) coolrib. Q1 (BC140) can be replaced with a NTE128 or the older ECG128 (same company). Except as a charger, this circuit can also be used as a regular power supply.

Parts List:

R1 = 0.56 Ohm, 5W, WW
R2 = 470 Ohm C2 = 220nF
R3 = 120 Ohm
R4 = 100 Ohm
C1 = 1000uF/63V
Q1 = BC140
Q2 = LM317, Adj. Volt Reg.
C3 = 220nF (On large coolrib!)
P1 = 220 Ohm

Except for use as a normal Battery Charger, this circuit is perfect to 'constant-charge' a 12-Volt Lead-Acid Battery, like the one in your flight box, and keep it in optimum charged condition. This circuit is not recommended for GEL-TYPE batteries since it draws to much current. The above circuit is a precision voltage source, and contains a temperature sensor with a negative temperature coλficient. Meaning, whenever the surrounding or battery temperature increases the voltage will automatically decrease. Temperature coλficient for this circuit is -8mV per °Celcius. A normal transistor (Q1) is used as a temperature sensor. This Battery Charger is centered around the LM350 integrated, 3-amp, adjustable stabilizer IC. Output voltage can be adjusted with P1 between 13.5 and 14.5 volt.
T2 was added to prevent battery discharge via R1 if no power present. P1 can adjust the output voltage between 13.5 and 14.5 volts. R4's value can be adjusted to accommodate a bit larger or smaller window. D1 is a large power-diode, 100V PRV @ 3 amp. Bigger is best but I don't recommend going smaller. The LM350's 'adjust' pin will try to keep the voltage drop between its pin and the output pin at a constant value of 1.25V. So there is a constant current flow through R1. Q1 act here as a temperature sensor with the help of components P1/R3/R4 who more or less control the base of Q1. Since the emitter/base connection of Q1, just like any other semiconductor, contains a temperature coλficient of -2mV/°C, the output voltage will also show a negative temperature coλficient.

That one is only a factor of 4 larger, because of the variation of the emitter/basis of Q1 multiplied by the division factor of P1/R3/R4. Which results in approximately -8mV/°C. To prevent that sensor Q1 is warmed up by its own current draw, I recommend adding a cooling rib of sorts. (If you wish to compensate for the battery-temperature itself, then Q1 should be mounted as close on the battery as possible) The red led (D2) indicates the presence of input power.Depending on what type of transistor you use for Q1, the pads on the circuit board may not fit exactly (in case of the BD140).

Here is a deluxe version of the simple charge rate limiter, using the same idea but with the ability to charge two packs simultaneously from a single wall charger. For circuit description and parts list, see the simple charger page. Since wall chargers provide about 55mA, you should not use this dual circuit to charge batteries at rates greater than 27mA (for a total of 54mA).

Some workbenches can’t help ending up looking like a rats nest of cables and equipment, so its always an advantage if a piece of mains equipment can be removed from somewhere to free up an extra mains socket. Here we are using the ubiquitous PC as a battery charger. An unused serial interface port can supply enough current to charge (or trickle charge) low-capacity Nickel Cadmium (NiCd) batteries. You could for example, use the batteries in a radio and charge them during use.

The three serial port connections TxD, DTR, and RTS, when not in use, are at –10 V and can supply a current of around 10 to 20mA (they are short-circuit protected). The circuit shown supplies a charging current of approximately 30mA. If it is necessary to alter the polarity of the charging circuit then it is a simple job to reverse the diodes and using software, switch the port signals +10 V. Those interested could also write a software routine to automatically recharge the batteries.

The ASUS Eee is a fantastic ultra-portable notebook with almost everything required for geeks (and nothing that isn’t). Plus it features fantastic build quality and is very well priced. If you live in New Zealand you can get them from DSE; at the time of writing they are the exclusive supplier. I worked out it’s the same cost as importing one once you include all the duties and tax, plus you get the advantage of a proper NZ-style mains charger. Anyway, being so small I thought it would be nice to be able to carry this around in the car. Unfortunately I couldn’t find a car charger available anywhere at the time so I decided to tackle the problem myself. As a bonus this provides an opportunity for an external high-capacity battery.

Commercial Equivalent:
I thought at this stage it would be worth noting that a commercial car charger is now available for less than it cost me to build this from Expansys and is available in most countries (select your location on their site). It outputs 9.5v from 10-18v in at up to 2.5A. I’d actually recommend it over the design here is it seems to perform better at lower voltages (that one works down to 10V). However I have kept this page up as a reference for those who enjoy tinkering.

Design:
The charger included with the Eee is rated at 9.5v, 2.315A. There isn’t a fixed voltage regulator available for this exact voltage, so the circuit needed to be designed around an adjustable regulator. I decided to design the charger around the LM2576 “Simple Switcher” IC from National Semiconductor. There are tons of ICs like this available, many of which are a bit more efficient, however I selected this one because it is readily available and relatively cheap. It also has a lower drop-out voltage (~2V) than many other chips I looked at which is important when powering the device from a car or 12v SLA battery.

This circuit could have used a standard three pin regulator IC such as the LM317, however most types require an external transistor when handling so much current and not to mention the fact that they are very inefficient; they draw the same amount of current from the input as the load and the difference in power is dissipated as heat. The main problem with using the LM2576 is the fact it needs quite a large inductor due to its somewhat low switching frequency. The inductor I used is made by Pulse Engineering, part number PE92108KNL. I’d prefer a smaller one, however I couldn’t find one capable of supplying the required current that I could purchase in single units. Besides the PE92108KNL is apparently designed specifically to work with the LM257x series.

The circuit also includes a low voltage cut-out based on a 9.1v Zener diode and BC337 transistor that will shut down the regulator if the input voltage is below 11.5V. This prevents unstable operation of the regulator at lower input voltages, and also helps prevent accidental flattening of the supply battery. Substituting this transistor for similar type may affect the cut-out voltage; the Vbe of the transistor should be 1.2v.All of the components used should be pretty readily available in most areas. I got everything from Farnell. Jaycar also sells everything except the inductor. Make sure you specify high temperature, low ESR capacitors as these help result in more stable operation and better efficiency of the charger.

Unfortunately the end result is a charger that is slightly bulkier than I would really like. I attempted to fit this inside an old mobile phone charger case so the whole thing could hang out of the cigarette lighter, however I ran into trouble making the circuit stable enough and dissipating all the heat. Due to the high current involved compared to a mobile phone charger the components are much bulkier so it’s pretty tricky to get all to fit! If I do get it finished I’ll add an update.

Parts List:

• 2x 10k resistor (R1 & R4)
• 2x 22k resistor (R2 & R3)
• 1x 1.5k resistor (R5)
• 1x 120μF 25v electrolytic capacitor (C1)
• 1x 2200μF 16v electrolytic capacitor (C2)
• 1x 1N5822 Schottky diode (or equivalent)
• 1x 9.1v 0.5W Zener diode
• 1x BC337 NPN transistor
• 1x LM2576T-ADJ IC
• 1x 100uH, 3A inductor (e.g. Pulse PE92108KNL)
• 25°C/W or better minature heatsink (e.g. Thermalloy 6073)
• Cigarette lighter plug with 3A fuse and 2.1mm DC plug (e.g. DSE P1692)
• 2.1mm DC chassis mount socket
• 1.7mm x 4.75mm (ID x OD) DC plug and cable
• Small plastic enclosure

Building It:
Make yourself a PCB using the template below (600dpi). I simply laser print (or photocopy) the design onto OHP transparency sheet and then transfer the toner onto a blank PCB using a standard clothes iron. Any missing spots can be touched up with a permanent marker before etching. This is quick, usually results in pretty tidy boards and hardly costs a thing. There is a tutorial on a variation of this method at http://max8888.orcon.net.nz/pcbs.htm.
Install the components on the PCB and triple check the layout before soldering. It is much easier to start with the low profile components such as resistors and diodes, then install the larger components after-wards. Don’t forget the wire link; this is shows as a red line on the layout guide above. Remember to smear a small amount of heatsink compound on the regulator tab before mounting the heatsink.

For a case I used a small plastic enclosure from DSE, part H2840, as it was all the local store had in stock that was remotely suitable. The PCB is designed to fit into this particular case, however any small box should be suitable. If you have a dead laptop charger lying about it might be worth ripping the guts out of that and salvaging the case. If your enclosure is different you may need to modify the design to suit, so I have provided the schematic and PCB design files for download. They were created using Eagle. The Eee uses a standard 1.7mm DC power connector with a positive tip.
Testing:
Connect the circuit to a 12v supply. If you use a car or lead acid battery ensure you have a 3A fuse fitted in line with the circuit before connecting it, just in case. Use your multimeter to check that the circuit outputs about 9.45v with no load. Connect a 12V, 21W lamp (e.g. old brake lamp from a car) or similar load across the output and check that the voltage doesn’t vary much. You should now be able to connect your Eee. The circuit design should be good for up to 2.5A, so there is plenty of margin for the Eee to fully function and charge its own battery off this supply.

SLA Battery Carry-bag:
Jaycar have a really cool carry bag with a shoulder strap designed to perfectly fit a 12v 7AH sealed lead acid battery. The bag features a fused cigarette lighter socket and is the perfect compliment to this charger. It works well with the Eee and provides hours of extra use. The shoulder strap means it’s not too bothersome to carry about and the charger circuit itself zips up neatly inside the bag. The under-voltage cut-off means the battery will never run completely flat, and the Eee will simply cut over to its internal battery once the SLA runs out. I got my SLA battery from Rexel as they are much cheaper (approx NZ$18 including GST last time I bought one) and they don’t sit as long on the shelf as many other suppliers.

Disclaimer:
This circuit is intended for people who have had experience in constructing electronic projects before. The circuit design and build process are provided simply as a reference for other people to use and I take no responsibility for how they are used. If you proceed with building and/or using this design you do so entirely at your own risk. You are free to use the content on this page as you wish, however I do ask that you include a link or reference back to this page if you distribute or publish any of the content to others.

The battery discharger published in this website may be improved by adding a Schottky diode (D3). This ensures that a NiCd cell is discharged not to 0.6–0.7 V, but to just under 1 V as recommended by the manufacturers. An additional effect is then that light-emitting diode D2 flashes when the battery connected to the terminals is flat. The circuit in the diagram is based on an astable multivibrator operating at a frequency of about 25 kHz. When transistor T2 conducts, a current flows through inductor L1, whereupon energy is stored in the resulting electromagnetic field. When T2 is cut off, the field collapses, whereupon a counter-emf is produced at a level that exceeds the forward voltage (about 1.6 V) of D2.

A current then flows through the diode so that this lights. Diode D1 prevents the current flowing through R4 and C2. This process is halted only when the battery voltage no longer provides a sufficient base potential for the transistors. In the original circuit, this happened at about 0.65 V. The addition of the forward bias of D3 (about 0.3 V), the final discharge voltage of the battery is raised to 0.9–1.0 V. Additional resistors R5 and R6 ensure that sufficient current flows through D3. When the battery is discharged to the recommended level, it must be removed from the discharger since, in contrast to the original circuit, a small current continues to flow through D3, R2-R3, and R5-R6 until the battery is totally discharged.

The flashing of D2 when the battery is nearing recommended discharge is caused by the increasing internal resistance of the battery lowering the terminal voltage to below the threshold level. If no current flows, the internal resistance is of no consequence since the terminal voltage rises to the threshold voltage by taking some energy from the battery. When the discharge is complete to the recommended level, the LED goes out. It should therefore be noted that the battery is discharged sufficiently when the LED begins to flash.

The Over-the-Top type of operational amplifier is ideal for use as a current sense for battery charger applications. The design described here can be used with chargers for rechargeable batteries (Lead/acid or NiCd etc). The 5 V operating supply for the circuit is derived from the battery on charge. The circuit uses a sense resistor R8 to determine the value of current flowing in or out of the battery. An LED output shows whether the battery is charging or discharging and an analogue output displays the battery charge or discharge current. The circuit can also be altered to shown different ranges of charging current to cater for higher capacity cells. IC1a and IC1b together with T1 and T2 form two current sources, which produce a voltage across R5. The voltage across R5 is proportional to the current through resistors R8 and R1 (for IC1a) or R8-R3 (for IC1b).

The current source formed by IC1a and T1 is active when the batteries are discharging and IC1b and T2 is active when the batteries are being charged. In each case the inactive opamp will have 0V at its output and the corresponding transistor will be switched off. IC1d amplifies the voltage across R5, which is proportional to the sense current. The component values given in the diagram produce an amplification factor or 10. A sense current of 0.1 A will produce an output voltage of +1 V. The supply voltage to the circuit is +5 V so this will be the maximum value that the output can achieve. This corresponds to a maximum charge/discharge current of 0.5 A To display currents from 0 to 5.0 A, resistor R7 can be omitted to give IC1d a voltage gain of 1.

Higher currents can be displayed by using a lower value of sense resistor R8. A DVM or analogue meter can be used at Vout to give a display of the charge/discharge current. The constant current sources can only function correctly when the supply to the voltage regulator circuit (UBatt. e.g. 6V or 12V) is greater than the operating voltage of the opamps (+5 V). The supply voltage to the LT1639 can be in the range of +3 V and +44V and voltages up to 40V over the supply voltage are acceptable at the inputs to the opamp. IC1c controls the charging/discharging LED output. The inputs to this opamp are connected to the outputs of the current source opamps and its output goes high when the battery is being charged and low when it is discharging.

The design of the charger is similar to that of many commercially available chargers. The charger consists of a mains adaptor, two resistors and a light-emitting diode (LED). In practical use, this kind of charger is perfectly all right. Resistor R1 serves two functions: it establishes the correct charging current and it drops sufficient voltage to light the diode. This means that the LED lights only when a charging current flows into the battery. The charging current is about 1/4 of the battery capacity, which allows a slight overcharging, and yet the charging cycle is not too long (4–5 hours).

The value of the resistors may be calculated as follows, for which the nominal e.m.f. and the capacity of the battery must be known. Adjust the output of the mains adaptor to 1.17 times the nominal battery voltage plus 3.3 V, which is the potential across R1. Note that the adaptor must be capable of supplying a current of not less than half the battery capacity. The value of R1 in ohms is equal to 3.3 divided by 1/4 of the battery capacity. The value of the resistors for various battery voltages is given in the Table. The battery capacity is taken as 1 Ah.

The rating of R1 should be 5 W. If the battery to be charged has a different capacity, the theoretical value of R1 in the table must be divided by the battery capacity. Its actual value is the nearest one in the E12 series. For instance, if a 6 V battery with a nominal capacity of 600 mAh is to be charged, the value of R1 must be 20/0.6 = 33R.

Although you may well be the proud owner of the very latest NiCd battery charger, you may still come across the odd 'incompatible' battery, for example, one having a rare voltage or requiring a much higher charging current than can be supplied by your off-the-shelf charger. In these cases, many of you will resort to an adjustable mains adaptor (say, a 500-mA type) because that is probably the cheapest way of providing the direct voltage required to charge the battery. Not fast and not very efficient, this 'rustic' charging system works, although subject to the following restrictions:

1. You should have some idea of the charging current. In case you use an adaptor which is adjustable but of the unregulated, low output current type, you can adjust the current by adjusting the output voltage.
2. You have to know if the current actually flows through the battery. A current-detecting indicator is therefore much to be preferred over a voltage indicator.
3. To prevent you from forgetting all about the charging cycle, the indicator should be visible from wherever you pass by frequently. Using the circuit shown here, the LED lights when the baseemitter potential of the transistor exceeds about 0.2 V. Using a resistor of 1 ? as suggested this happens at a current of about 200 mA, or about 40 mA if R1 is changed to 4.7?. The voltage drop caused by this indicator can never exceed the base-emitter voltage (UBE) of the transistor, or about 0.7V. Even if the current through R1 continues to increase beyond the level at which UBE = 0.7 V, the base of the transistor will 'absorb' the excess current. The TO-220 style BU406 transistor suggested here is capable of accepting base currents up to 4A. Using this charging indicator you have overcome the restrictions 2 and 3 mentioned above.

The design of solar panel systems with a (lead-acid) buffer battery is normally such that the battery is charged even when there is not much sunshine. This means, however, that when there is plenty of sunshine, a regulator is needed to prevent the battery from being overcharged. Such controls usually arrange for the superfluous energy to be dissipated in a shunt resistance or simply for the solar panels to be short-circuited. It is, of course, an unsatisfactory situation when the energy derived from a very expensive system can, after all , not be used to the full. The circuit presented diverts the energy from the solar panel when the battery is fully charged to another user, for instance, a 12V ice box with Peltier elements, a pump for drawing water from a rain butt, or a 12V ventilator.

It is, of course, also possible to arrange for a second battery to be charged by the super-fluous energy. In this case, however, care must be taken to ensure that when the second battery is also fully charged , there is also a control to divert the superfluous energy. The shunt resistance needed to dissipate the superfluous energy must be capable of absorbing the total power of the panel, that is, in case of a 100W panel, its rating must be also 100 W. This means a current of some 6–8 A when the operating voltage is 12 V. When the voltage drops below the maximum charging voltage of 14.4V growing to reduced sunshine, the shunt resistance is disconnected by an n-channel power field effect transistor (FET), T1.

The disconnect point is not affected by large temperature fluctuations because of a reference voltage provided by IC1. The necessary comparator is IC2, which owing to R9 has a small hysteresis voltage of 0.5V. Capacitor C5 ensures a relatively slow switching process, although the FET is already reacting slowly owing to C4. The gradual switching prevents spurious radiation caused by steep edges of the switched voltage and also limits the starting current of a motor (of a possible ventilator). Finally, it prevents switching losses in the FET that might reach 25W, which would m a ke a heat sink unavoidable. Setting up of the circuit is fairly simple. Start by turning P1 so that its wiper is connected to R5.

When the battery reaches the voltage at which it will be switched off, that is, 13.8 – 14.4V, adjust P1 slowly until the output of comparator I C2 changes from low to high, which causes the load across T1 to be switched in. Potentiometer P1 is best a 10-turn model. When the control is switched on for the first time, it takes about 2 seconds for the electrolytic capacitors to be charged. During this time, the output of the comparator is high, so that the load across T1 is briefly switched in. In case T1 has to switch in low-resistance loads, the BUZ11 may be replaced by an IRF44, which can handle twice as much power (150 W) and has an on-resistance of only 24 mR. Because of the very high currents if the battery were short-circuited, it is advisable to insert a suitable fuse in the line to the regulator. The circuit draws a current of only 2 mA in the quiescent state and not more than 10mA when T1 is on.

If the lead acid battery is not using for long time, self discharge takes place at the rate of 4 % per week at 27 degree. For example, a 125 Ah tubular battery self discharge at the rate of 5 Amps current per week if it is not in charge / discharge cycles. This add on circuit can be used to detect the self discharge of the 12 volt lead acid battery below the safe level of 11.6 volts. It is ideal to monitor the charge level in the battery, it the battery is not using for long period. The circuit uses only a few components and its working is simple. A PNP transistor T1 act as a switch to light the LED, if the battery voltage drops below the safe level. The base bias of T1 is controlled by a Zener diode ZD. Its rating is 10 volt 1 W.

The Zener diodes usually requires 1.6 volts excess than its rated value to enter into the “Avalanche state”. So, as long as the battery voltage is above 11.6 (10+1.6), Zener conducts and keeps the base of T1 high. Since T1 is a PNP transistor, it will not conduct till its base becomes negative. So LED remains dark. When the voltage in the battery reduces below 11 volts, Zener turns off and the base of T1 becomes negative. T1 then conducts and LED lights. So the battery can be charged again to keep it in top condition. Preset VR can be used to set the exact point at which LED turns on.

Although not a new device, the LM317 is still a high-performance regulator. Its output voltage is essentially immune to fluctuations in load, supply voltage and temperature and this makes it ideal as the central element in a float charger for NiMH cells. Float charging has the advantage of keeping the cells fully charged and ready to use without the potential damage of long-term trickle charging or the cost of low-discharge cells. This works because NiMH cells do not have the memory problems associated with Ni-cads. The circuit is based on a conventional LM317 regulator. Resistors R2 & R3 and trimpot VR1 set the maximum output voltage to between 1.3V and 1.4V per cell. VR1 should be adjusted for a value of 1.35V per cell at the regulator output. Resistor R2 has been fixed at 240O. The formula for the voltage output is: Vout = 1.25*(1 + (R3 + VR1)/R2).
Diode D1 protects the circuit against reverse polarity of the power supply and protects the LM317 should the power be disconnected while it is still connected to a charged battery pack. Resistor RCL and transistor Q1 limit the maximum current in the event of a short circuit or the connection of a severely discharged battery pack. LED2 provides an indication of voltage input to the charger. LED1 and the 680O resistor provide the same function for the charger output and also provide a minimum load for the regulator when the battery pack is nearing full charge. This is necessary to keep the regulator output from drifting up and damaging the batteries. The circuit uses an external DC plugpack and is suitable for four NiMH cells rated at 2.5Ah.

Table 1 gives alternative values for 1-10 batteries in series at peak charge currents of between 200mA to 600mA. If you are using the specified plug-pack and the TO-220 packaged LM317T, you will need a heatsink rated at 12°C/W or better for any design other than the 200mA single cell charger. A TO-3 packaged device with the correct plug-pack will be OK without a heat-sink for any of the 200mA configurations and up to four cells charging at 400mA.

This circuit is even simpler and employs a 6V feed from one of the stator connections on the vehicle’s alternator. This is connected to a 6V automotive relay (RLY1) which controls a Continuous Duty Solenoid (RLY2). This solenoid electrically connects or isolates the batteries. When the engine is started and the alternator stator voltage rises, the 6V relay turns on. This turns on the Continuous Duty Solenoid to connect the two batteries in parallel. As long as the engine is running, the vehicle’s alternator will maintain charge in both batteries.

When the engine is shut down, the alternator stator voltage drops and the Continuous Duty Solenoid switches off, thus isolating the second battery from the vehicle’s electrical system. Provided that camping accessories are only connected to the second battery, the main battery should never discharge. Because the concept is entirely dependent upon the alternator’s stator output voltage, you cannot forget to turn the system on or off as it happens automatically.

A simple NiCd charger can be built using ‘junk box’ components and an inexpensive LM317 or 78xx voltage regulator. Using a current limiter composed of R3 and a transistor, it can charge as many cells as desired until a ‘fully charged’ voltage determined by the voltage regulator is reached, and it indicates whether it is charging or has reached the fully charged state. If the storage capacitor (C1) is omitted, pulsed charging takes place. In this mode, a higher charging current can be used, with all of the control characteristics remaining the same. The operation of the circuit is quite simple. If the cells are not fully charged, a charging current flows freely from the voltage regulator, although it is limited by resistor R3 and transistor T1.

The limit is set by the formula Imax ≈ (0.6 V) ÷ R3 For Imax = 200 mA, this yields R3 = 3 Ω. The LED is on if current limiting is active, which also means that the cells are not yet fully charged. The potential on the reference lead of the voltage regulator is raised by approximately 2.9 V due to the voltage across the LED. This leads to a requirement for a certain minimum number of cells. For an LM317, the voltage between the reference lead and the output is 1.25 V, which means at least three cells must be charged (3 × 1.45 V > 2.9 V + 1.25 V). For a 78xx with a voltage drop of around 3 V (plus 2.9 V), the minimum number is four cells.


When the cells are almost fully charged, the current gradually drops, so the current limiter becomes inactive and the LED goes out. In this state, the voltage on the reference lead of the regulator depends only on voltage divider R1/R2. For a 7805 regulator, the value of R2 is selected such that the current through it is 6 mA. Together with the current through the regulator (around 4 mA), this yields a current of around 10 mA through R1. If the voltage across R1 is 4 V (9 V – 5 V), this yields a value of 390 Ω. The end-of-charge voltage can thus be set to approximately 8.9 V. As the current through the regulator depends on the device manufacturer and the load, the value of R1 must be adjusted as necessary. The value of the storage capacitor must be matched to the selected charging current. As already mentioned, it can also be omitted for pulse charging.

It is impossible to imagine present day society without any batteries. Count the number of gadgets in your house that are powered from batteries, you will be stunned by the number of batteries you will find. The majority of these devices use penlight batteries and if you’re a little environmentally conscious you will be using rechargeable batteries. A few years ago these batteries were invariably NiCd types. However, these batteries suffer from a relatively high rate of self-discharge and from the so-called memory-effect. It is now more common to use NiMH batteries.

The advantage is that these batteries do not suffer from the memory-effect and generally also have a much higher capacity, so that they last longer before they have to be recharged again. From the above you can conclude that every household these days needs, or could use, a battery charger. A good charger needs to keep an eye on several things to ensure that the batteries are charged properly. For one, the charger has to make sure that the voltage per cell is not too high. It also needs to check the charging curve to determine when the battery is fully charged. If the charging process is taking too long, this is an indication that something is wrong and the charger must stop charging.


Sometimes it is also useful to monitor the temperature of the cells to ensure that they do not get too hot. The circuit presented here is intended for charging NiMH batteries. The MAX712 IC used here contains all the necessary functionality to make sure that this happens in a controlled manner. Figure 1 shows the schematic of the charger. The heart of the circuit is easily recognized: everything is arranged around IC1, a MAX712 from Maxim. This IC is available in a standard DIP package, which is convenient for the hobbyist because it can be directly fitted on standard though-hole prototyping board.

IC1 uses T1 to regulate the current in the battery. R1 is used by IC1 to measure the current. While charging, IC1 attempts to maintain a constant voltage, equal to 250 mV, across R1. By adjusting the value of R1 the charging current can be set. The value of R1 can be calculated using the formula below: R1 = 250 mV / I charge For a charging current of 1 A, the value of R1 has to be 250 mV / 1 A = 0.25 Ω. The power dissipated by R1 equals U × I = 0.25 × 1= 250 mW. A 0.5-watt resistor will therefore suffice for R1. Transistor T1 may need a small heatsink depending on the charging current and supply voltage.


IC1 needs a small amount of user input regarding the maximum charging time and the number of cells in the battery to be charged. IC1 has four inputs, PGM0 to PGM3, for this purpose. These are not ordinary digital inputs (which recognise only 2 states) but special inputs that recognise 4 different states, namely V+, Vref, BATT– or not connected. To make this a little bit more user friendly, we’ve brought out the necessary connections to 2 connectors (K3 and K4). A number of dongles have been made (Figure 2) that can be plugged into these connectors and set the number of cells and the maximum charging time. When determining the maximum charging time we have to take into account the charging current and the capacity of the cells that are connected.

The charging time can be calculated with the formula: Tcharge = Ccell / I charge × 1.2 where Ccell is the capacity in Ah (e.g., 1200 mAh = 1.2 Ah). After the nominal charging time has been calculated, we can use the first dongle that has a value that is equal or greater than the calculated charging time. For example, if we calculated a maximum charging time of 38 minutes, we have to select the dongle for 45 minutes. When IC1 is replaced by a MAX713, the charger becomes suitable for charging NiCd batteries (but not suitable for NiMH batteries any more!). The only difference between these two ICs is the value of the detection point at which the cell(s) are considered to be completely charged. The ICs are otherwise identical with regard to pin-out, method of adjustment, etc. To make it easy to swap between the ICs, we recommend an IC-socket for IC1.

This circuit charges two NiCad cells with a constant current and features dual charging rates, voltage cutoff and an audible alarm. The circuit is powered by a 12VAC centre-tapped mains transformer, together with two rectifier diodes (D1 & D2) and a 1000mF filter capacitor. A 7806 3-terminal regulator is used to generate a 6V rail for the remainder of the circuit. Transistor Q1 and LED1 constitute a basic constant-current source. The forward voltage of the red LED (about 1.5V) minus Q1’s base-emitter voltage (about 0.6V) appears across the 6.8W or 15W emitter resistors, depending on the position of S1. With a 15W resistance in the emitter circuit, the charging current is about 60mA, whereas with 6.8W it is about 130mA.

This is sufficient to charge 600mAH "AA" cells in 14 hours and five hours, respectively. An LM393 voltage comparator (IC1) is used for the voltage cutoff function. Its inverting input is set to 2.9V (nominal) via trimpot VR1, while the non-inverting input senses battery voltage. This means that while the cells are being charged, the output transistor (in the LM393) is switched on, also switching on Q1 and enabling the current source. Once the cells are charged to approximately 80% or more of capacity, their terminal voltages will exceed 1.45V, so the voltage at the non-inverting input (pin 3) of IC1 will exceed the reference voltage on the inverting input (pin 2).


This causes IC1’s output to switch off, in turn switching Q1 off and disabling the current source. To prevent rapid switching action around the voltage cutoff point, a 100nF capacitor provides feedback between the output and inverting input of the comparator. Four NAND gates are used to build two simple oscillators of different frequencies. When cascaded together, the result is a pulsed tone from the piezo transducer to indicate charge completion.

Editors note:
Absolute terminal voltage is not always a reliable indicator of Nicad battery charge state. Importantly, batteries should never be charged for longer than the manufacturer’s specified period.

A number of readers have asked how to tell when the Lead-Acid Battery Zapper has done its job and battery desulphation is complete. In the author’s experience, batteries that are going to respond to this treatment will generally show quite a high peak voltage across the terminals at the beginning of the treatment. If this steadily decreases and practically disappears, then the treatment is near to complete. This may take anything from a week to many months, depending on the size and condition of the battery. In the absence of an oscilloscope to monitor the voltage peaks, a simple peak detector can be constructed from a fast diode and 100nF capacitor. Any high-impedance multimeter (eg, most digital types) can then be used to measure the average DC voltage across the capacitor.

This circuit may be used to replace the single current limiting resistor often found in dirt cheap battery chargers. The alternative shown here will eventually pay off because you no longer have to throw away your NiCds after three months or so of maltreatment in the original charger. The circuit diagram shows an LM317 in constant-current configuration but without the usual fixed or variable resistor at the ADJ pin to determine the amount of output current. Also, there is no switch with an array of different resistors to select the charge currents for three cell or battery types we wish to charge: AAA, AA and PP3 (6F22). When, for example, an empty AAA cell is connected, the voltage developed across R1 causes T1 to be biased via voltage dropper D1.


This results in about 50 µA flowing from the LM317’s ADJ pin into the cell, activating the circuit into constant-current mode. D4 is included to prevent the battery being discharged when the charger is switched off or without a supply voltage. The charging current I is determined by R1/R3/R3 as in R(n) = (1.25 + Vsat) / I where Vsat is 0.1 V. The current should be one tenth of the nominal battery capacity — for example, 170 mA for a 1700-mAh NiCd AA cell. It should be noted that ‘PP3’ rechargeable batteries usually contain seven NiCd cells so their nominal voltage is 8.4 V and not 9V as is often thought.

If relatively high currents are needed, the power dissipation in R1/R2/R3 becomes an issue. As a rule of thumb, the input voltage required by the charger should be greater than three times the cell or battery (pack) voltage. This is necessary to cover the LM317’s dropout voltage and the voltage across R(n). Two final notes: the LM317 should be fitted with a small heat sink. With electrical safety in mind the use of a general-purpose mains adapter with DC output is preferred over a dedicated mains transformer/rectifier combination.