Data Pack J

Servo control IC ZN409

The RS ZN409 is a precision monolithic integrated Features

circuit designed particularly for pulse-width posi- ● Low external component count
tion servo mechanisms used in many types of ● Low quiescent current (7mA typical at 4.8V)
control applications. The low number of compo- ● Excellent voltage and temperature stability
nents required with the device, together with its ● High output drive capability
reduced length and low power consumption, make ● Consistent and repeatable performance
this integrated circuit ideal for use in compact ● Precision internal voltage stabilisation
servo applications where space, weight and battery ● Time shared error pulse expansion
life are at a premium. The amplifier will operate ● Balanced deadband control
over a wide range of repetition rates and pulse ● Schmitt trigger input shaping
widths and is therefore suitable for the majority of ● Reversing relay output
systems. The RS ZN409 can also be used in motor (dc motor speed control)
speed control circuits. PIN CONNECTIONS

Absolute maximum ratings

Supply voltage __________________________6.5 Volts
Package dissipation__________________300 Milliwatts
Operating temperature range _________-20°C to 65°C
Storage temperature range __________-65°C to 150°C
TOP VIEW

Characteristics VCC = 5V At 25°C ambient temperature unless otherwise stated.

Parameter Min. Typ. Max. Unit Conditions
Input threshold (lower) 1.15 1.25 1.35 V Pin 14
Input threshold (upper) 1.4 1.5 1.6 V Pin 14
Ratio upper/lower threshold 1.1 1.2 1.3 –10 to + 65°C
Input resistance 20 27 35 kΩ Vin≤2V (Pin 14)
Input current 350 500 650 µA Vin≥2V (Pin 14)
Regulator voltage 2.1 2.2 2.3 V –10 to + 65°C, 1.3mA load current
Regulator supply voltage rejection ratio 200 300 – VCC = 3.5 to 6.5V
dVin
RSRR =
dVout
Monostable linearity for ±45° pot rotation – 3.5 4.0 % Rp = 1.5kΩ, R1 = 12kΩ
Monostable period temperature – ±0.01 – %/°C Excluding RT, CT.
coefficient Rp = 1.5kΩ, R1=12kΩ
(potentiometer slider set
mid-way)
Output Schmitt deadband ±1 ±1.5 ±3 µS CE = 0.47µF
Minimum output pulse 2.5 3.5 4.5 mS CE = 0.47µF, RE = 180kΩ
Error pulse for drive 70 100 130 µS 15 ms repetition rate
CE = 0.47µF, RE = 180kΩ
Total deadband ±3.5 ±5 ±6.5 µS CD = 1000pF
P.N.P. drive 40 55 70 mA T = 25°C
35 50 65 mA T = –10°C
Output saturation voltage – 300 400 mV IL = 400mA
Direction bistable output 2 2.8 3.6 mA VCC = 5V max.
Supply voltage range 3.5 5 6.5 V
Supply current 4.6 6.7 10 mA Quiescent
Total external current from regulator 1.3 – – mA VCC = 3.5V
Peak voltage VC EXT (with respect – 0.7 – V T = 25°C
to 2V regulated voltage) – 0.5 – V T = –10°C
Circuit operation Circuit
Component function
reference Value
Monostable timing components RT 100kΩ
Figure 1 System diagram CT 0.1µF
Potentiometer and timing reference
components Rp 2kΩ
R1 6.2kΩ
Pulse expansion CE 0.47µF
RE 180kΩ
Deadband CD 2200pF
Dynamic feedback RF 300kΩ
RB 300kΩ
R2 1.2kΩ
Input coupling CC 2.2µF
Motor decoupling C2 0.1µF
C3 0.1µF
R.F. decoupling (see text) C1 0.1µF
Drive transistors T1, T2 BC461

The system diagram in Fig. 1 illustrates the internal

structure of the RS ZN409 servo control IC Input circuit
In the standard servo application the displacement The RS ZN409 operates with positive going input
of a potentiometer control varies the pulse width of pulses which can be coupled either directly or via a
a timing circuit. The signal produced provides the capacitor to pin 14. the advantage of ac coupling
control input for the servo IC The servo shown in is that should a fault occur which causes the input
Fig. 2 consists of the RS ZN409 integrated circuit, signal to become a continuous positive level, the
several external components, a power amplifier servo will remain in its last quiescent position,
consisting of two external PNP transistors and two whereas with direct coupling the servo output arm
on-chip NPN transistors which form a bridge circuit will rotate continuously. A nominal 27kΩ on-chip
to drive the dc motor. The motor drives a reduc- resistor is shunted across the input to provide dc
tion gear box which has a potentiometer attached restoration of the input signal when ac coupling is used.
to the output shaft. The RS precision dc servo The active input circuit is a Schmitt trigger
system components are ideally suited for this ap- which allows the servo to operate consistently with
plication and the 6V motor should be used. This slow input edges and supplies the fast edge required
potentiometer in association with R1 and the timing by the trigger monostable independent of input edge
components CT and RT controls the pulse width of speed.
the timing monostable. The input pulse is com-
pared with the monostable pulse in a comparison
circuit and one output is used to enable the correct
Figure 3
phase of an on-chip power amplifier. The other
output from the pulse comparison circuit drives the Input circuit Input characteristics
pulse expansion circuit (CE, RE) via the deadband
circuit (CD). Thus any difference between the input and
monostable pulses is expanded and used to
drive the motor in such a direction as to reduce this
difference so that the servo takes up a position
which corresponds to the position of the control
potentiometer.

Figure 2 Servo system using the ZN409

2
The input circuit and its V/1 characteristic are
shown above. D1 and D1 are the parasitic substrate Figure 5 Deadband circuit and waveforms
and isolation diodes associated with the input
resistors. It is advisable that the pulse input ampli-
tude should not fall below 0V nor exceed the
supply voltage Vcc in order to prevent these diodes
from conducting, although a small amount of con-
duction will not cause the circuit to malfunction.
When ac coupling is used the value of CC should
be chosen to give a pulse droop not exceeding 0.3
volts.

Figure 4 Input waveforms

When the difference pulse is applied T1 turns off

and the base of T2 rises on an exponential wave-
form with a time constant of 4.7 kΩ × CD. If the
difference pulse is small the potential reached on
the base of T2 is insufficient to turn T2 on and no
output results.
The pulse expansion circuit has a built in deadband
of 1.5 µsec with CE = 0.47 µF and this must be
added to the deadband caused by CD to obtain the
total Td.
Assuming that the input signal swings between 0V Td = 1.5 + tdµsec
and VCC and taking the input chord resistance RIN of and td = 3.3CDµsec(CD in nF)
13kΩ the droop for a pulse of duration msec will (Taking V1 = 1.5 volts and Vbe = 0.75 volts)
be: Thus with CD equal to 2200pF(2n2)
td = 7.26µsec and Td = 8.76µsec.
Vd = VCCtp volts tp (msec) The mechanical deadband Ød depends on the
CC. RIN CC(µF) chosen sensitivity S1 of the servo and in a typical
RIN (kΩ) application at ±500µsec input pulse variation
causes ±50° rotation, ie. S1 = 10µsec per degree.
For a nominal pulse width of 1.5 msec and Vd equal
to 0.3 volts the required minimum value of CC can Thus Ød = 2.Td degrees (Td in µsec. S1 in µsec per
be found to be 1.85µF and a nominal value of 2.2µF S1 degree).
is chosen. (Nearest preferred value). Thus a value for Td of 8.76µsec provides a mechan-
4.8 × 1.5 ical deadband Ød of 1.8°.
CC = 0.3 × 13 = 1.85µF And generally:
CD in nF.
If it is required to operate the servo with reduced
input pulse amplitude the input pulse should ex-
ceed the upper Schmitt threshold voltage of 1.5
volts by a reasonable margin and a minimum input
Ød =
3 + 6.6 CD
S1
degrees
{ S1 in µsec per
degree.

pulse amplitude of 2.4 volts is recommended.

Pulse expansion
Deadband circuit Figure 6
The function of the deadband circuit is to provide a Pulse expansion circuit and characteristics
small range of output shaft positions about the
quiescent position where the difference pulse does
not drive the motor. This is necessary to eliminate
hunting around the quiescent position caused by
servo inertia and overshoot. The minimum dead-
band required is also a function of the pulse expan-
sion characteristics and dynamic feedback compo-
nent values.

3
A schematic of the pulse expansion circuit is shown (shown dotted in Fig. 6) and if this is excessive it
in Fig. 6. In the quiescent state with no drive the can lead to the servo exhibiting an underdamped
Schmitt trigger input is biased via RE and takes up a characteristic causing jittering or hunting. Thus for
level just above the lower threshold VL. full pulse expansion the voltage on CE should
A drive pulse causes a current IE to be switched on discharge from its quiescent value of 1.5V to 0.75
for the duration of the pulse and this discharges CE volts. Thus with IE = 3mA for the current source:
linearly with time. Thus, at the end of the pulse the 1.5 - 0.75 IE
voltage on CE depends on the duration of the pulse. =
If the pulse is narrow and just causes the potential te CE
on CE to fall to VL the Schmitt trigger will switch to ∴ CE = 4.teµF(te in msec)
the upper threshold VH and at the end of the drive For te = 0.1 msec, a value of 0.47µF was chosen for
pulse CE will start to charge to VH with a time CE.
constant CERE. When the potential on CE reaches
VH the Schmitt will switch to VL and CE will dis-
charge to the quiescent level. The output drive is Figure 8 Pulse expansion waveform
taken from the Schmitt output.

Figure 7 Pulse expansion timing diagram

If tp is the maximum motor drive pulse length

required, ie. equal to the input pulse repetition
period for full pulse expansion, and the mean value
of the potential on CE is taken as 1.2 volts, then:
(tp - tmin)
dc motors need a certain amount of drive to dv = × (VCC - 1.2)
overcome static friction and the minimum output CERE
pulse obtained from this form of pulse expansion
characteristic is chosen to ensure that the motor And for the discharge period te:
will rotate when driven. IEte
The value of tmin is determined by the Schmitt dv =
CE
trigger hysteresis and the exponential waveform
on CE because VH is small the following linear ∴ RE = (tp - tmin) × (VCC - 1.2)
relationship is sufficiently accurate. IEte
(VCC - VL) For nominal values of VCC = 4.8V and IE = 3mA
VH = × tmin (tp - tmin) kΩ
CERE
RE = 1.2
For nominal operation VCC = 4.8V; VL = 1.5V; VH = te
0.12V and: and for tp = 20msec, tmin = 3.5msec, te=0.1msec,

tmin .
CERE
30
msec {
CE in µF

R in kµ
E
RE = 180kΩ. (Nearest preferred value).

Monostable timing
and for CE = 0.47µF and RE = 180kΩ, tmin =
3.5msec. Figure 9
It can be seen from the simple equation that tmin is Monostable timing circuit and waveform
dependent on VCC, and tmin will increase with
reducing VCC. This variation is put to good use to
maintain the initial motor drive, VCC × tmin reason-
ably constant over the operating voltage range of
3.5 to 6.5 volts.
When the pulse expansion drive is increased above
the minimum value the output pulse increases
from tmin almost linearly until full pulse expansion
is reached, ie. when the ouput pulse width equals
the input pulse repetition rate. The pulse expansion
will be almost linear provided that the current The leading edge of the input waveform triggers
source IE does not saturate, ie. provided that CE is the timing monostable by opening switch S1. CT
not discharged to almost zero volts. Ideally the then charges until the differential amplifier detects
current source should saturate when full motor that the timing waveform potential has fallen to Vp,
drive is obtained but due to component tolerances the potential on the potentiometer wiper and
it is usual to allow some margin to ensure that full switch S1 is closed to terminate the timing pulse. Thus
motor drive can be obtained. If a margin is allowed, an the monostable period is determined by the
extended pulse expansion characteristic results setting of the potentiometer wiper. In the standard
4
application the servo centre position pulse width is Vm = 2.0. tmono(max)
1.5 msec with a range of ±50° rotation at 10 µsec CTRT
per degree. Thus the 2.0 msec maximum monost-
since CT = 0.1µF (nearest preferred value) was
able period tmono (max) corresponds to a potentio- chosen with RT = 100kΩ, Vm = 0.4V and hence: R1
meter setting of 200° (for a linear relationship) and = 3.1 Rp. If Rp = 2kΩ then R1 = 6.2kΩ.
since the potentiometer has a total rotation of
approximately 270° and the maximum allowable Dynamic feedback
swing on pin 3 is specified as 0.5 volt, the value of Without dynamic feedback in the standard applica-
CTRT can be calculated as follows: tion the inertia of the motor and gearbox causes
0.5 2 the servo output shaft to overshoot the set position
tmono (MAX) . CTRT which results in the servo 'hunting'. If the dead-
∴CTRT = 4tmono (max) band was widened to stop this effect an unaccept-
ably large deadband would result and the servo
Thus if tmono (max) = 2 msec, CTRT = 8msec.
would still be underdamped. The dynamic feed-
The optimum value of RT is 100kΩ due to the back circuit utilises the motor back emf (which is
design of the on-chip monostable circuit giving CT proportional to motor speed) and feeds back a
= 0.1µF (nearest preferred value). proportion of this signal to the wiper of the poten-
RT = 100kΩ CT = 0.1µF tiometer. The phase of the feedback signal is
The value of R1 can now be calculated from the chosen to modify the potential on the wiper so that
actual voltage swing with a potentiometer setting the monostable period is dynamically varied to
of Øp = 200° and Ømax = 270°. reduce the motor drive as the servo output shaft
approaches the set position and the values of the
feedback resistors are chosen to achieve optimum
setting characteristics.
Figure 10 The value for RF and RB of 330kΩ will suit the
Potentiometer set at 200°C Equivalent circuit normal type of servo mechanism, however if the
servo is fairly fast this can be decreased to 300kΩ
to minimise any tendency to overshoot. Where the
servo is slow RF and RB can be increased to 360kΩ
or 390kΩ.
RF Decoupling
C1 (typical value 0.1µF) is only necessary where
strong RF fields may affect the operation of the circuit.
Transistors T1 and T2
Thus from the equivalent circuit: The external PNP transistors are usually selected
for a low VCE(sat) to obtain maximum output drive.
Vm = (VCC - Vm) The recommended type is BC461.
200 70
270 Rp R1 + 270 Rp
Motor speed control
where Vm is calculated from the actual values of CT Introduction
and RT chosen using the relationship: In the motor speed control application the

Figure 11
Proportional motor speed control circuit

5
RS ZN409 is used as a linear pulse width amplifier. Pulse width generator
The dc motor is driven via a power amplifier with A pulse width generator circuit is shown in Figure
a train of pulses whose mark/space ratio can vary 12 suitable for driving the RS ZN409 in servo or
between zero and one to control the motor speed motor speed applications.
from zero to maximum. The device operates with
fixed timing components and a fixed resistor re- the frame rate is generated by the first timer and
places the position feedback potentiometer. The the frequency is adjusted by VR1 so that the time
nominal monostable period represents zero motor between pulses is 18mS. VR2 is then used to
speed and input pulses less than or greater than control the ouput pulse width over the range
nominal drive the motor in the forward and reverse 1-2mS. VR3 is set so that the midpoint of VR2
direction respectively. The motor direction is usual- corresponds to an output pulse of 1.5mS.
ly controlled by a relay operated from pin 4, the VR4 can be included should an output amplitude
direction output. Pulse expansion components CE control be required.
and RE are chosen to obtain the required relation
between control potentiometer and motor speed
and it is usual to operate with a much larger Figure 12 Pulse width generator
deadband than that used in the servo application.
The outputs from pins 9 and 5 of the RS ZN409
integrated circuit are combined using two ZTX500
PNP transistors to provide a pulsed output whose
mark/space ratio varies from 0 to 1 depending on
the deflection of the control potentiometer.
This signal is then used to drive the motor via the
power amplifier.
The RS ZN409 has additional circuitry which per-
forms the motor reversing function by taking the
output from the direction bistable and provides
either zero current or approximately 3mA sink
current at pin 4, depending on the state of the
direction bistable. This current is amplified and
used to drive the relay coil (100mA) via the 2N3053
transistor thus controlling the motor direction via
the relay changeover contacts.
It is usual to have a relatively wide deadband and
CD = 22nF provides a deadband of about 14% (±7
degrees).
A 1kΩ potentiometer (Rp) can be used to set up the
zero output condition with the control pot in its
central position.