e99a022.pdf

BATTERY CHARGERS

battery

charge/refresh station (1)

automatic charging at up to 3 A

This microprocessor

controlled charger is

suitable for Nickel-Cad-

mium (NiCd) as well as

Nickel-Metal-Hydride

(NiMH) batteries and

battery packs. The

charger is capable of

topping up 1 to 10

cells using a charging

current of up to 3 A.

The end of the charg-

ing process is automat-

ically detected, as is

the adaptation of the

charging current when

the full actually available capacity is reached. Thanks to a special pulse-

driven charging process, discharging is not required before the battery is

connected. In addition to the charging function, automatically followed

by trickle charging, the station also features a charge/discharge cycle

mode and a refresh mode for ‘tired’ or ‘presumed dead’ batteries.

Design by N. Bechtloff & G. Brenner

(Conrad Technology Center, CTC)

To get things straight from the onset:

this charge/refresh station is not suit-

able for batteries consisting of ‘small’

cells like, for example, button cells, AAA

cells or 9-V PP3-style batteries. The

smallest cell size that may be used is

the ‘Mignon’ style battery capable of

‘fast’ charging. In plain words, do not

use the station with batteries having a

capacity of less than 700 mAh (at a dis-

charge rate of C/3). By contrast, there is

virtually no upper limit as to what the

station van handle: sub-C, Baby, Mono

and even larger cell sizes are okay as

long as they are Nickel-Cadmium

(NiCd) or Nickel-Metal-Hydride

Elektor Electronics

10/99

Features

Selection between NiCd and NiMH, 1 up to 10 cells

microprocessor control using intelligent charging algorithm

Sense circuits for charging current, charging voltage, charge density and temperature

Charging current automatically adapted to cell capacity (from mignon-size > 700 mAh at C/3)

Pre-discharging not required, battery is always charged to 100% of currently available capacity.

Automatic switch-off at end of charging cycle

Maximum charging current 3 A effective (8 A peak)

Maximum discharging current: 1.5 A effective

Three charging modes:

1. Charge (1 charge period for 100% charge)

2. Cycle (charge, discharge, charge)

3. Refresh (up to six cycles)

With 4 or more cells, charging data are stored and re-used after mains interruption

Storage and readout of charging and discharging capacity

LC display readout

Single-switch control for mode selection and data display

(NiMH) types. Because of its operating

principle, t his charge/refresh station is

not suitable for lead-acid or Lithium-

Ion (Li-Ion) cells or battery packs .

discharged, and precautions should be

taken to make sure the battery is never

over-charged with a high current. The

first condition is normally satisfied by

a simple discharging cycle. Two meth-

ods are often used together to meet the

second condition:

the charging efficiency is by no means

constant. Instead, it depends on many

factors including the amount of cur-

rent in relation to the battery capacity

and the cell temperature. By itself, the

‘controlled charge time’ method is

therefore too inaccurate to achieve full

charging of he battery while ensuring

that overcharging does not occur.

P RINCIPLES

OF OPERATION

Fast charging of rechargeable batteries

generally follows this simple rule: the

higher the charging current, the

shorter the charging time given a cer-

tain battery capacity (expressed in Ah

or mAh). You, the user of the battery, of

course want to be sure that the battery

is charged in the shortest possible time,

yet always be confident that the bat-

tery is reliably topped up and not dam-

aged in any way by a fast charging

process. On the contrary: the charging

process should also guarantee opti-

mum use of the battery capacity at the

longest possible battery life. Further-

more, we would like the charging

process to be as simple as possible,

allowing the battery to be connected

without having to discharge it first and

independent of the charge still con-

tained in the battery. In other words:

connect the battery to the charger,

switch on the charger and leave it to do

its work. After a short time, the battery

should be charged to its full capacity,

100 per cent, not more, not less.

Those of you who have ever had a

the pleasure of studying the charging

process of batteries with at least one

Nickel electrode (NiCd or NiMH) will

confirm that it is hard to satisfy all con-

ditions mentioned above. On the one

hand, charging at relatively high cur-

rents is good to make the most of the

available capacity and at the same time

combat the very annoying ‘memory’

effect of NiCd batteries. On the other

hand, it requires the battery to be fully

1. controlled charging time

The charging current is removed after

the time calculated to fully top up the

battery. This however requires cer-

tainty about the initial amount of

charge in the battery, which is simple

to ascertain by discharging it com-

pletely. Secondly, the battery capacity

has to be known and set by the user.

Apart from having to adjust the

charger, the user often has a problem

in that the battery at hand may not

have the specified capacity any more.

A further point to keep in mind is that

2. controlled charge voltage

When the ‘full charge’ point is reached,

charging current is increasingly not

accepted by the battery but instead

turned into heat. The resulting cell

temperature increase causes the charg-

ing voltage to stop rising, then stabilise

at a certain level, and finally even drop

a little at the onset of overcharging.

This drop may serve as an indication to

initiate the switching off of the charg-

ing current. The advantage of this so-

Figure 1. Typical charging voltage curves

for NiCd and NiMH cells at 20 ºC and 1CA

(1-hour charge).

Elektor Electronics

10/99

Reliable protection against overcharging guarantees long battery life

U) switch-off method

is that it works independently of the

initial battery state, and only evaluates

the actually measured battery capacity.

In practice, however, it is not easy to

determine the exact switch-off instant

with 100% certainty just by monitoring

the charging voltage curve. Because

the relevant voltage changes are often

in the millivolts range, there is a real

risk of the evaluating circuit respond-

ing to noise and disconnecting the bat-

tery too early. Moreover, the voltage

curve as a function of charge condition

may not be the same with all batteries,

although these are of the same age and

type. The main complicating factor is

however the fact that the voltage drop

is not always as clear as we would like

it to be (or it does not occur at all). That

is why designing a reliable ∆

∆

transferred to the battery. Based on this

information, the charger does not cut

the charging current just like that.

Instead, the battery is first discharged for

about 9 seconds to ‘sense’ its internal

resistance. If nothing untoward is

detected, the charging continues at a

reduced charging current. Likewise,

the station ensures that the charging

process is terminated in accordance

with the actually loaded amount of

energy. After passing the peak in the

charging voltage (typical curves are

shown in Figure 1 ), a small battery is

quickly shut down. By contrast, larger

batteries are given some more charging

current because it takes longer for

them to reach full charge (100% of

nominal capacity).

A further problem arises when no

clear voltage drop can be detected after

the charging voltage peak. Depending

on the battery make, size and condi-

tion, that is a reality to be taken into

account. A regular delta-U control then

fails to switch off the charging current,

causing the battery to be overcharged

at a high current, and turning the

charger into an effective ‘battery killer ’.

With the present charger, however, the

switch-off routine is already activated

when the charging voltage no longer

rises. Even if the charging voltage stag-

nates, the charging process continues

for a while (depending on the amount

of energy already transferred) before

the battery is disconnected. This

switch-off routine — which adapts

itself to the battery — allows the maxi-

mum capacity to be reached with any

battery, and at the same time prevent

overcharging under all conditions.

Starting at 1.8 A, the microcontroller

also adapts the discharging current to

the response of the battery at hand. By

increasing the pauses in between dis-

charging pulses, the effective discharg-

ing current is automatically reduced

from about 1.5 A down to 0.5 A as the

battery capacity drops also.

ual control is by means of a rotary

switch for the number of cells and a

mode switch for the menu-driven

selection of the charger mode. Results

and selections appear on an LC dis-

play. The NiCd/NiMH cell type selec-

tion switch is not shown in the block

diagram.

You may not be able to spot all of

the above functions at first blush in the

circuit diagram — finding it all takes a

bit of study. The dashed lines indicate

the division of the parts between two

sub-boards. The left-hand section goes

to the smaller board, the right-hand

section, to the large board.

U switch-

off control is often likened to sorcery.

Mainly because of the required ‘intelli-

gence’, a microcontroller is then called

for, besides clever design techniques

and, most of all, lots of experience in

this field.

Charging circuit

The centre tap of the mains trans-

former being connected to ground,

thyristors THR1 and THR2 act as a full-

wave rectifier under microprocessor

control. Because the charging current

is adjusted by way of phase angle con-

trol, very high pulse currents of up to

8 A may be applied to the battery or

cell. The battery charging current is

measured by means of the voltage

drop developed across sense resistors

R2 and R3. After averaging by R7-C2,

the measured voltage is amplified by

IC1a and then compared with a mains-

synhronous sawtooth voltage gener-

ated by IC2c. The sawtooth level

depends on the reference voltage

VREF and may therefore be adjusted

using P2 (at the reference source, IC3).

As soon as the voltage derived from

the charging current and available at

the output of IC1a exceeds the saw-

tooth voltage, T3 is switched off via

IC2a, preventing the thyristor from

being triggered. The phase angle con-

trol operates by itself with just these

three opamps, with the microcontroller

switching it all on and off via transistor

T1. When the controller drives T1 via

the CHARGE line, the transistor will

conduct and pull the voltage on C2 to

about 5 V. This voltage is interpreted as

a very high charging current by the

three opamps. This subcircuit responds

by not activating the thyristors, which

remain off. Once the controller re-

enables the triggering circuit, the volt-

age on C2 (5 V) drops slowly so that

the current control starts off smoothly

from zero. Also, the controller is able to

switch between 100% charging current

(3 A effective) and 33% (1 A effective)

via the CHV line. When the controller

switches on T2 via the CHV line, R8 is

connected into circuit and R7 forms a

voltage divider.

T HE DESIGN CONCEPT

The battery charge/refresh station

works without charge-time limiting

and discharging before charging. This

means that partly discharged batteries

may also be topped up. To prevent par-

tial charging from reducing the battery

capacity as a result of the dreaded

memory effect, charging takes place

with very strong current pulses of up

to 8 A. The effective charging current is

adapted to the charging behaviour of

the relevant battery by varying the

pulsewidth. The maximum effective

charging current is 3 A. The charging

current setting is achieved depending

on the voltage response. With a small

battery, the terminal voltage rises faster

than with a large one, so the charging

current is reduced accordingly to

match the smaller battery capacity (or

battery condition, if applicable).

Because the present circuit is a fast

charger by any standard, the minimum

charging current is a respectable 1 A.

That is why the station should only be

connected to batteries with a nominal

capacity of not less than 700 mAh, and

suitable for rapid charging. However,

the ‘overshoot’ effect may still occur at

the start of a charging session, in par-

ticular with small batteries (mignon

cells) or batteries suffering from

reduced capacity. Because of the high

initial current, the battery temperature

rises so quickly that the charging volt-

age drops a little after an initially

‘steep’ increase. In a normal delta-U

charger, that would mean a premature

end to the charging process. Not so

with the present charger, because its

internal microcontroller not only mon-

itors the charge response, but also

records the amount of energy already

P RACTICAL

REALISATION

The block diagram shown in Figure 2

provides an overview of the relatively

complex circuit. The actual circuit dia-

gram is given in Figure 3 . The charging

current for the batteries is supplied

directly by the secondary winding on

the mains transformer. The amount of

current is determined by the micro-

controller in combination with a half-

phase control based on thyristors. A

MOSFET is used to discharge batteries

in ‘cycle’ and ‘alive’ (refresh) mode.

Using an A-D converter the microcon-

troller continuously monitors the

charging current as well as the charg-

ing voltage. In addition, a temperature

signal is processed, where either the

battery temperature or the charger

temperature may be measured. Man-

Discharging circuit

Batteries are discharged by passing

their current through MOSFETs T4

and T5. This, too, is an autonomous

subcircuit which is simply switched on

and off by the microcontroller. Resis-

tors R35 and R36 ensure equal distrib-

Elektor Electronics

10/99

called delta-U (

ution of the discharging current

between the two FETs. The full dis-

charging current flowing through R37,

this resistor supplies the voltage that

enables opamp IC1 to monitor the

amount of current. The opamp com-

pares the measured voltage with a ref-

erence level set with preset P1. Here,

too, the controller is able to switch it all

on and off by means of a voltage level.

When the DIS line is pulled to 5 V, this

high voltage reaches the opamp’s

measurement input via R29 and D3.

This level is taken to mean a high dis-

charging current, and causes the

opamp to switch off the discharging

circuit. As in the charging control, the

switching off is gradual thanks to tim-

ing element C7 in the opamp feedback

circuit. For security’s sake, the dis-

charging circuit includes a Polyfuse,

which is a kind of super-PTC. In the

inactive (‘cold’) state, this component

represents a very small resistance in

the region of a few tenths of an ohm.

If the nominal current is exceeded, the

internal resistance of the Polyfuse rises

as a result of the higher temperature.

When the current drops again, the

fuse returns to its low-resistance state

once it has cooled down. The advan-

tage: a ‘self-healing’ ability; the disad-

vantage: much slower response than a

conventional wire fuse.

power

supply

FAN

Vref

V+

RESET

LCD

charge

µP

MODE

discharge

A / D

overtemp

# cells

990070-12

Figure 2. Block diagram of the battery

charge/refresh station. The charging current

flows directly from the transformer secondary

inti the battery.

at an almost ‘current-less’ point in time.

In this way, the (considerable) voltage

loss caused by contact resistance and

wires is avoided, and the actual battery

terminal voltage is reliably measured.

The measurement occurs just after the

zero crossing, with the controller

receiving a signal from IC2b telling it if

the current is large or very small. This

is achieved by IC2b weighing the cur-

rent measurement signal (amplified by

IC1b) against a very small direct volt-

age. In the circuit diagram, this is

labelled ‘ZEROREF’.

type used).

D15 and D16 generate an auxiliary

voltage to ensure proper triggering

of the thyristors. When no battery is

connected, D17 and R71 apply a

high direct voltage across the bat-

tery connector K5, to enable the

microcontroller to detect, via the A-

D converter, whether or not a bat-

tery is connected.

D14 carries the current for the voltage

regulators, to wit

IC4 followed by diodes D11 and D12

for the two 5-V supply lines CPU-

VDD (microcontroller) and VDD

(remaining 5-V electronics).

Via R69, IC3 is connected to the 6 V at

the output of IC4, and so generates

the reference voltage VREF (typ.

2.8 V) adjusted with P2.

ZEROREF (typ. 60 mV) is derived from

VREF via R67 and R68. Its function

is to aid the zero-crossing detection

of the mains transformer voltage.

Note however that ZEROREF

always has to exceed the maximum

offset voltage of the opamps used in

the circuit.

A-D converter

This is built in quasi-discrete fashion

with opamp IC6d acting as a simple

single-slope converter. Initially, T6 is

driven hard by the microcontroller so

that the voltage across C10 is (almost)

zero. When a measurement is

required, the controller first switches

off T6, allowing C10 to be charged via

R51 by a stable reference voltage. As

soon as this comparison voltage is

equal to the (scaled-down) battery volt-

age applied to the other opamp input,

the output of IC6d toggles. The time

between the enabling of the capacitor

charging until the toggling of IC6d is

measured by the microcontroller, and

the resulting value enables it to com-

pute the measured current. The

absolute accuracy of the measurement

is not terribly important because we’re

looking at capturing a measurement

value during a period of several hours,

rather than an absolute value over a

longer period.

To enable NiMH batteries or cells to

be charged, switch S1 modifies the

composition of voltage divider at the

converter input. The result is that the

A-D converter is made slightly more

sensitive to enable it to reliably follow

the smaller voltage changes in the

charging voltage curve (as compared

with NiCd, see the curves in Figure 1).

The charging voltage measurement

always takes place at the same instant

after a charging current pulse, that is,

Battery/cell polarity

The battery or cell polarity is checked

by means of IC6a comparing with

ZEROREF. When the battery is con-

nected the wrong way around, LED

D6 lights.

Temperature guard

This function consists of a simple cir-

cuit around IC6b monitoring the volt-

age produced by a measurement

bridge based on NTC R70. If the mea-

sured temperature exceeds the thresh-

old set by bridge resistor R80, the

opamp output drops low and causes

the microcontroller to interrupt the rel-

evant process (charging or discharg-

ing). Once the temperature has

dropped below the threshold level, the

process is continued. If, however, the

temperature guard is activated three

times in succession, the process is ter-

minated.

MPU backup voltage

The backup supply voltage for the

microcontroller to work from during a

mains outage (or when the station is

briefly switched off) comes from the

external battery rather than from an

internal power source. Of course, this

will only work as long as the battery

voltage is high enough. To ensure the

guaranteed switchover to emergency

supply by the external battery, the volt-

age at the input of voltage regulator

IC4 is sensed via T10 and D10. As long

as this voltage is higher than 6 V, tran-

Power supply

This is relatively complex sub-circuit.

D18 and R72 supply the ventilator

supply voltage (depending on the

Elektor Electronics

10/99

+ ACCU

R11

VREF

POLYFUSE

1µ

16V

R13

215k

R27

1A6

100n

BATTERY

4V4

R12

9k09

0V18

4V5

R33

100 Ω

R34

100 Ω

BUZ11

33k2

70mV

IC1a

R31

390 Ω

IC1c

CHGI

1V5

100 Ω

R28

R35

R36

R20

R10

22n

CHV

R30

2k2

1µ 16V

BC547B

0V19

R32

R37

1N4148

R29

100n

R72

Ω 2W

D18

R71

10k

DIS

1N4001

C26

C27

FAN

10µ 63V

220µ

35V

R21

22k

ZEROREF

THR2

IC2b

6A3

R22

C25

D 15

D17

IC1b

TIC116A

1V2

100n

R19

1N4148

14V

10n

C24

D16

TIC116A

100n

2x 18V

3A33

C21

C19

Tr1

6A3

IC1

IC2

100n

47µ

16V

R83

THR1

R16

R15

R14

D 14

S10K275

IC2d

IC1 = LM324

IC2 = LM339

1N4001

VOUT

T3 BC557B

BUZ11

TIC116A

1N4148

C23

R17

4k7

R18

10k

100n

VREF

IC2a

VREF

R25

R23

R24

IC2c

1N4148

22n

ZEROREF

C22

22µ

35V

Figure 3. The two dashed boxes represent the division of

the circuit in two sub-boards.

sistor T9 conducts, while T8 is held

switched off so that it has no effect on

the controller ’s supply voltage (CPU-

VDD). However, as soon the level

drops below 6 V, FET T8 starts to con-

duct, thereby passing the battery volt-

age to the microcontroller. In this

setup, D8 ensures that a maximum

level of 6.8 V can not be exceeded. As

long as the battery powers the con-

troller, all values and settings remain

safely stored. In this way, a charging or

discharging process continues where it

left off after the mains voltage disap-

peared, as if nothing had happened.

The return of the mains voltage is

detected via C12 and R7. If the voltage

rises again, T7 is shortly opened by

C12, causing C11 to discharge across

the transistor and so generate a reset

for the microcontroller. Next, the reset

line returns to logic high, and the con-

troller starts from a defined state. Via

its IRQ pin, the controller is informed

Elektor Electronics

10/99

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: