by Eric Gaze, G8NKA
ONE OF MY FAVOURITE bits of kit is a handheld transceiver. If like me, you were startled at the cost of manufacturers’ bat-tery packs, you probably opted for an empty case and at the next rally bought yourself enough batteries to fill it. The most common pack holds six 1.2V batteries giving a terminal voltage of 7.2V at capacities rang-ing from 500 – 900mA/hr. The commonest is 700m/k/hrs and all reference to batteries in this article refer to this variety. .
The calculations are simple and easily altered to whatever type you have. To keep batteries in top condition it is necessary to discharge them fully before recharging. This charge / discharge facility along with a built in one or 12 hour timer is the essence of this project. Most nicads have their capacity written on them in the form current (mA) / time (hours). In a perfect world, if we take a fully charged battery and place a variable load across it, then adjust the load so we get a current of 800mA flowing, the battery would sit there supplying its 800mA at 7.2V for one hour.
The voltage and current would fall to zero, leaving the battery completely discharged. If we then remove the load and connect the battery to a constant current source of 800mA for one hour, the battery would be ‘full’. In fact it’s a bit more complicated. Batteries exhibit a characteristic called ‘memory’. If, for exam-ple, our standard battery is repeatedly dis-charged to say, half its capacity, the battery remembers and over time the effective ca-pacity falls to 400mA/hrs.
Most handhelds do not discharge the bat-tery fully or they are recharged before they are completely discharged. The practical out-come is a gradual reduction in the working life of the battery pack. To overcome this problem, the battery should be discharged to about 0.9V per cell before recharging commences.
After several cycles this reforms the battery to its rated capacity. It is also standard practice to over-charge the battery by about 10%, le 770mA for one hour. This is OK, but it is usually better to take a bit longer and do the job properly. Charging the battery at 10% of its amp/ hour rate plus 10% (77mA) for 10 to 12 hours is better, as the charge is low enough not to generate internal heat.
Also it can be kept topped up by trickle charging, ie reducing the current to approximately 3% of its Amp/hour rating (21mA). This can be maintained safely for a week or so. So now we have our charger requirements as follows:
• Discharge the batteries.
• Charge the batteries for one hour at their mA/hr rating or 12 hours at 10% of their mA/hr rating (plus 10%).
• Trickle charge the batteries (several days) at 3% of the mA/hr rating.
HOW IT WORKS
WE NOW KNOW WHAT is required, so how is it done? Referring to Fig 1 and ignoring the IC and surrounding components, look at the two transistors and presume they are both off. Insert the battery pack, nothing happens! Connect the base of TR1 to a positive supply and it will ‘turn on’. Think of it simply as a switch.
- Battery —————————— 6 x 1.2v cells 7.2 volt pack
- Charge range———————–0-1300ma
- Min Battery voltage————-2.5
- Discharge voltage—————-5.4
- Discharge time (empty)——-12 mins
- Discharge time (full)———–50 mins
- Charge Time———————–1 hour or 12 hours trickle.
When TR1 is switched on it effectively puts R17 (8R2) across the batteries. Assuming the batteries are fully charged and have a terminal voltage of 8.2V (to keep the figures simple) there will be a discharge current of lamp, (1 = VIA) discharging the battery in about one hour. In practice it will take about 50 minutes for a ‘full’ battery and about 10 to 15 minutes when my handheld ‘blanks out’.
We then remove the positive voltage to the base of TR 1, turning it off. Applying a nega-tive voltage to the base of TR2 (as it is a PNP device) turns it on. The Darlington transistor and associated components form a simple constant current source (no originality claimed), the current being limited by R16. Its value was chosen to give a maximum current of about 1300mA (this is about the limit of the transformer (Fig 2) and should be adequate for any increase in battery capacities).
That is really the heart of the charger. All we need to do is build some circuitry that will monitor the voltage of the discharging bat-tery, switch it to charge, adjust the charge current to suit a range of batteries, build a timer for one or 12 hours, then arrange for the batteries to be trickle charged until we require them. Panic not! The IC does all these jobs. Anyone interested in the full workings of the U24008 should obtain the manufacturer’s data sheet.
I have not used all the functions available, trimming them to my specific re-quirements. The IC has three voltage comparitors which, with suitable voltage divider resistors, moni-tor the state of the battery. Pin 5 of the IC would normally go to a positive temperature coefficient resistor in close proximity to the battery, so if it became hot this would be detected and the IC would switch off the charging current. I dispensed with this be-cause I did not want to keep removing the batteries from the case and did not intend to use the rather brutal half hour charge range.
A simple potential divider R2, R3 takes care of this. Pin 6 via divider R13,R10 monitors the falling battery voltage on discharge. When the battery voltage falls to about 5.4V (0.9V x 6 cells) pin 10 goes low, turning off TR1. Pin 12 also goes low, turning on TR2 and charg-ing starts. SW2 is used to ‘trick’ the IC into thinking the battery is discharged, putting it into the charge mode, irrespective of the battery state.
Pin 4 has two functions: 1) It informs the IC that a battery is connected – provided its terminal voltage is above 2.5V. 2) It turns off the charging if it senses the battery voltage is becoming too high. This function caused me some problems. The manufacturers consider the battery overcharg-ing if its nominal 1.2V cells reach 1.6V (in a pack of six cells, this would be 9.6V). Now, perhaps because I used ‘surplus’ batteries, all of the packs I charged on a stable bench power supply. whether for one hour or twelve, reached a terminal voltage of about 10.2V.
After lots of head scratching, I decided there was not much I could do about this and elected to disable the over-voltage facility. D1 clamps the voltage at the end of R9 to 0.6V. R5, R7. form a potential divider reduc-ing this to 0.3V. This enables the IC to ‘see’ when a battery is connected, but disables the over-voltage facility. Now for the charger part. Si is set to one hour and the battery has been discharged. Now if pin 12 goes low maximum current flow, set by R16 of 1300mA occurs – too much for most batteries.
Fortunately there is more to it than this. A 200Hz oscillator is running in the IC and this via pulse-width modulation, is used to control the current flow. If we imagine TR2 as a switch that can be turned on and off rapidly, that gives an idea of how it works. If the switch is turned ‘on’ for the same length of time as it is turned ‘oft’, and this is done 200 times a second the battery would appear to integrate the pulses and we would see an average current flow – half of 1300mA. Now. if the switch turned ‘off’ for twice as long as it is turned ‘on’, we would get an apparent steady current flow of approximately 440mA – one third of 1300mA.
When RV1 is adjusted, the pulse-width of the 200Hz waveform is altered, the wider the pulse width the greater the ‘on’ time, the narrower the pulse width the greater the ‘off’ time. The result of this is, as RV1 is altered, an apparent smooth current flow, adjustable from 0 to 1300mA. Setting this for our standard battery (770mA), we switch to the 12-hour range (remembering the current requirement of approximately 10% of the battery capacity plus 10% – 77mA).
The IC does this by pulsing the battery at the previous preset current for 200 milliseconds every 1.2 seconds (you do the maths, I believe the manufacturers.) After charging for one or 12 hours the IC switches to its ‘trickle charge’ mode, by again pulsing the battery at the preset current (770mA) for 200 milliseconds, only now the frequency drops to once every 16.8 seconds. Once the one hour charge rate is set, the IC looks after everything else.
The charging times and clock frequency are default values and cannot be altered. The sharp eyed among you may be won-dering why we require D5 and C9. The dis-charge ‘high’ at pin 10 is not steady, it is also clocked at 200Hz. and varied by RV1. This results in erratic discharge times. D5 rectifies the pulses and C9 smoothes them, the time constant being long enough to turn TR1 hard on for any practical setting of RV1. Last but not least U I (LM7812) supplies the regulated 12 volt rail.
CONSTRUCTION IS QUITE straightforward, especially if using the PCB and layout shown in Figs 3 and 4. The front and rear panel layouts are not critical, but the prototype is quite pleasing to the eye. The ‘S’ shaped heatsink is really the only bit of metalwork involved in the charger (see Fig 5 for dimen-sions). TR1 and TR2 both require mica wash-ers etc to insulate them from the heatsink and R16 sits on top of it. If an in-line fuse holder is used for FS2 this simplifies construction fur-ther. As the battery capacities are changed so infrequently, I didn’t bother with any scale on RV1.
This, together with an ammeter, can easily be added if you wish. Component val-ues are fairly critical and should be kept as shown. The two LEDs require a mention, as the IC will only supply a current of 5mA. R4, R5 have been chosen to limit the current to 4.5mA.
Using standard LEDs gives a satis-factory light output. Lastly do not leave out any decoupling capacitors – they are required. C4 is mounted on RV1 • positive to the center (slider), negative to the tag that is also connected to P5.
I ALWAYS THINK a lot of good projects are spoiled by the case and poor lettering, mak-ing it obvious that the equipment is home-made. Bearing in mind the cost of the trans-ceiver and this charger, it is surely worth spending more time and money to produce a finished article that you don’t feel the need to throw a cloth over every time someone visits your shack. The Nicharge is built in standard ‘Verobox’ and finished with ‘Letraset’ or similar.
Mark out, drill all holes, clean, draw faint pencil lines where appropriate, count the letters to find the centre of the word, rub on the letters. carefully clean off the pencil marks and finish off with a clear plastic spray, which you can get from most art shops along with the ‘Letraset’. The cost of both is about £6, but it will do for several projects. I know this adds a few hours to the completion of your projects, but t feel it is well worth it. WARNING: It must always be remembered that there are lethal voltages on any equip-ment connected to the mains. The trans-former. fuses and switches etc. Disconnect the Nicharge from the mains before working on it. Heat shrink sleeving or similar on ex-posed ‘live’ terminals etc is good working practice. A bench power supply set at about 18V (2A) connected from the negative rail and C10 positive will enable you to test/set-up the unit in safety.
TESTING / SETTING UP.
BEFORE COMMENCING TESTING and set-ting up note the WARNING above. • Remove IC2 – Remember, when han-dling, this is a CMOS device. II Fit a 500mA fuse in FS1, switch on. You should have approximately. 18V on FS1, and 12V on pin 1 and 8 of IC2. LED1 should be lit. There should be no voltage on the output terminals.
• If all is well, switch off, insert IC2 into its holder.
• Set RV1 to half-way, and S1 to one hour.
• Connect an analogue ammeter (1A scale) in the positive lead of the battery to be charged, (negative lead to the charger: positive lead to battery.)
• Turn on. Both the ‘On’ LED and the ‘Dis-charge’ LED should glow.
• After approximately two seconds the ‘Dis-charge’ LED should flash. The meter should show a discharge current of about 850mA – this depends on the battery’s charge condition.
• Switch off, reverse the meter leads be-tween the battery and the charger.
• Switch on. press S2. ‘Discharge’ LED goes out, and the ‘Charge’ LED flashes.
• Adjust RV1 to give a reading correspond-ing to your battery capacity +10°,/.. (770mA). • Switch off, remove meter, Replace the fuse with a 1A5. That’s it. No need to touch RV1 again unless you change the batteries for different capacity types.
[1 ] If you wish to enable the overvoltage facility, change R7 to a 100k, replace R9 with a shorting link, and remove diode D2. This will set an overvoltage level of about 10V.
[2) As the charger always goes into the discharge mode first, Si should be switched to charge (1 hour), so that the current level can be set without having to wait anything up to 50 minutes until the battery is discharged.
THE CHARGER PLACES a small drain on the battery in the ‘off’ condition. This is more than compensated for in the trickle charge mode. However, when the charger is turned off the batteries should be removed. When the batteries are being charged in the 12 hour or trickle charge mode there is a faint, but audible ‘beep’ every time they are pulsed. The ‘beep’ comes from the batteries themselves (I wonder what is resonating in there?) This, whilst faintly amusing, also gives a reassuring indication that the battery is indeed charging. Using an analogue meter the pulses of charging current (12 hour or trickle) can just be seen as a slight ‘flick’ of the pointer. My digital meter did not register at all. Finally I should like to thank the regulars on my local repeater GB3HG for putting up with me talking about it and trying/buying the prototypes. With special thanks to Brian, GORI-11. for editing the text.
Fig 1 : Circuit diagram of NiCad charger
.A kit. including PCB but excluding the box, can be obtained from JAB Electronic Components, 1180 Aldridge Road. Great Bar Birmingham 844 8PB