You are on page 1of 6

I!

ANALOG Precision,
WideBandwidt
I'DEVICES Synchronized
lsolation
Amplifie
"'
ffilffiKtt*
FEATURES
tr:::lu.l
l]:l.rllill.
:

Low Nonfineatityz !O.O12% max (289L)


Frequency Response: (-3dB) dc to 20kHz
{Full Power} dc to SkHz
Gain Adlustable 1 to 10OV/V, Single Resistor
3-Port lsolation: +2500V GMV lsolation Input/Output
Low Gain Drift: 10.005o/of C max
Floating Power Output: tl5V @ 15mA
120d8 CMR at 60Hz: Fully Shielded Input Stage
Meets UL Std. 544 Leakage: 2pA rms max, @ 115V ac, 60Hz
APPLICATIONS
Multi-Channel Data Acquisition Systems
Current Shunt Measurements
ProcessSignal lsolator
High Voltage Instrumentation Amplif ier
SCR Motor Control

GENERAL DESCRIPTION
Model 289 is a wide-band, accurate, low cost isolation ampli-
DEsrcN
ffffi;ffi;';"*"",',
Isolated Power: The floating power supply section provides
fier designed for instrumenration and industrial applications. isolated +15V outputs @ +5mA. Isolated power is regulated to
Three accuracy selections are available offering guaranteed within +5%. This feature permits model 289 to excire floating
gain nonlinearity error at 10V p-p ourpur: +O.Ol2o/omax signal conditioners, front-end buffer amplifiers and remore
(289L), !O.O25o/o max (289K), !O.O5o/o max (289J). All ver- transducerssuch as thermistors or bridges,eliminating the need
sions of the 289 provide a small signal frequency response for a separate isolated dcldc converter.
from dc to 2OkHz (-3dB) and a large signal response from dc
Adjustable Gain, A single external resistor adjusts the model
to 5kHz (full power) at a gain of 1VlV. This new design offers
289's gain from 1V/V to 1O0V/V for applications in high and
true 3-port isolation, +25OOVdc between inputs and outputs
low level transducer interfacing.
(or power inputs), as well as 24OY rms between power supply
inputs and signal outputs. Using carrier modulation tech- Synchronized' The model 289 provides a synchronization
niques with transformer isolation, model 289 interrupts terminal for use in multichannel applications. Connecting the
ground loops and leakage paths and minimizes the effect of synchronization terminals of model 289s synchronizestheir
high voltage transienrs.It provides 12OdB Common Mode internal oscillators, thereby eliminating the problem of oscil-
Rejection between input and ourput common. The high CMV lator "beat frequency" interferencethat sometimesoccurs
and CMR ratings of the model 289 facilitate accurate measure- when isolation amplifiers are closely mounted.
ments in the presenceof noisy electricalequipment such as Internal Voltage Regulator: Improves power supply rejection
motors and relays. and helps prevent carrier oscillator spikes from being broad-
WHERE TO USE THE MODEL 289 cast via the isolator power terminal to the rest of the system.
The model 289 is designed to interface single and multichannel Buffered Output: Prevenrsgain errors when an isolation ampli-
data acquisition systems with dc sensorssuch as thermo- fier is followed by a resistive load of low impedance. Model
couples, strain gaugesand other low level signals in harsh in- 289 can drive a 2kSl load.
dustrial environments. Providing high accuracy with complete
Three-Port Isolation: Provides true galvanic isolation between
galvanic isolation, and protection from line transients of fault
input, output and power supply ports. Eliminates need for
voltages, model 289's performance is suitable for applications
power supply and output ports being returned through a com-
such as process controllers, current loop receivers, weighing
mon terminal.
systems, high CMV instrumentation and computer inter-
face systems. Reliability: Model 289 is conservatively designed to be capable
of reliable operation in harsh environments. Model 289 has a
Use the model 289 when data must be acquired from floating
ca.lculatedMTBF of 271,835 hours. In addition, ttre model
transducers in computerized process contiol systems. The
289 meets UL Std. 544 leakage,2trrArms @ 115V ac,6OHz.
photograph above shows a typical multichannel application
allowing potential differences or inrerrupting ground loops,
among transducers, or between transducers and local ground.

lnformationfurnishedby Analog Devicesis believedto be accurate


and reliable.However,no responsibilityis assumed by AnalogDevices P.O. Box 280; Norwood, Massachusetts
02062 U.S.A.
for its use;nor for any infringements
of patentsor otherrightiof third
partieswlrichmay resultfrom its use.No licenseis grantedby implica- Tel:617132947O0 Twx:.710/3t14-6577
tion or otherwiseunderany patentor patentrightsof AnalogDevices. Telex:9221491 Cables:
ANALOGNORWOODMAS
SPECIFICATI0NS @+2b'candvs=+14.4V
(typicat to+25V noted)
'therwise
dcunles
Model 289J 2A9L OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
GAIN (NONINVERTIN(;)
Range I to 1OoV/V
l 0 ko MAX---->1
\s1.18)
Formula t=t*n,,rkQt --2.o2 -T
Deviation from Formula i l . 5 o l om a x I
1 5 p p m / ' c t y p ( 5 0 p p m / ' Cm a x ) o.75
vs. Tempcrature (o to +70"c)I (18.95)
10.0506 max io.025olo max 1 0 . 0 1 2 7 om a x
Nonlinearit.v, (15V S$'ing)''" IVIAX
I
INPUT VOLTAGE RATINCS I
Linear Differential Range (C = lV/V) tlOV min
0 . 0 4( 1 . 0 2D) t A
Max Safe Differential InPut
Conu nuous
I Minute
l2OV rms
24OV rms -T
Max CMV (lnputs lo OutPuls) +Vs 6O
Continuous ac or tlc 1 2 5 O O Vp e a km a x
POWERCOM 7 O
I
ac,6OHz, I Minute Duration
CMR, Inputs to OutPuts 60112
2 5 O O Vr m s
.SYNC8 0 I
Rs < lkQ, Balanced Source Impedance 12 0 d B :- LOOUT 9O ,Lll,
MAX
RS< lkQ. Hl lN LeadOnlY l O 4 d Bm r n
Max Leakage Current, Input to OutPut @
l'' T1r.. . ,,,
Hr ouT 1o O
I
I | 5V rms. 6oHr ai

INPUT IMPEDANCE
Differential
,,,,,?lf
33pFllroEf)
I
I
"*?TJflY
Overload
Commun MoLle
100ksl
x 191:g
,2.9p,Flts
Yi'$-^,. --] l- o,r,uor
INPUT DIFFERENCE CURRENT
tnirial @ +25'C lOnA (75nA max)
o 11ia1lc SHIELDEDMATING SOCKET
vs. I emperarure (o to Tooc)
AClzr4
INPUT NOISE (GAIN = looV/V)
Voltage
8/rVp-p 0.08
0.O5tlz to 1OOHz (2.03)
10Hz to lkHz 3gVrms 2.7 (68.58)REF -------] MAX
Current , c o P P r R. L A D . H T E L D I I
0.05H/ (o 100llr 1Pl''l L t_ t--1 . --T
---T
o3o
FREQUENCY RESPONSE
Small Signal -3dB
r
0.093 WIRE
FEEDTHROUGH
( 1 2 . 7 )M A X
I
C = lV/V 2OkHz t2.2Sl
G = IOOV/V 5kHz
F u l l P o w e r , 10 V P - P O u r P u t
C = lV/V
G = 10oV/V
5kHz
3.5kHz
II
Full Power, 2OV P-P OutPut
2.3kHz
I
G = 1V/V
G = IOOV/V 2.3kHz
0.14V/lls
'i:li'
Slew Rate
Settling Timea 10.05o/o,1toV Step

OFFSET VOLTAGE, REFERRED TO INPUT


. 499-e'. II
tnitial, @ +25"C

200
ts tf -v -a*
.__ 100
I
vs. Temperarure (o to +70'C) max al) a--E- max tro tf lvl"c -"* BOTTOM VIEW --l
Fo1(254) GRID
W E I G H T :1 5 G R A l v l S

vs.Supplyvolmge I 5.v +20v chiTe) nt!uvx


1l :o
RATEDOUTPUT AND SHIELDING
Voltage,2kQ Load t10V min INTERCONN"il;-,
Ou!put Impedance <lQ(dc to looHz) TECHNIQUE
Ourpul RrPPle, 0 l MHz Bandwidth
5mV P-P To preserve the high CMR performance of
No Signal IN
r I OVIN
:oTvP-p. . .. model 289, care must be taken to keep the
ISOLATED POWER SUPPLY capacitance balanced about the input terminals.
t15V dc
Voltage
!\oo/o
A shield should be provided on the printed cir-
Accuracy
Current
j5mA, min cuit board under model 289 as illustrated in the
Regulation No Load to Full Load outline drawing above (screened area). The LO
Ripple, O.J.Mtlz Bandwidth, No Load 25mV P"P
IN/ISO PWR COM (pin 1) must be connected
Full Load il-v P-l ,
POWER SUPPLY, SINGLE POLARITY'
to this shield. This shield is provided with the
Voltage, Rated Performance + 1 4 . 4 Vr o + 2 5 V mounting socket, model AC1214 (solder feed-
Voltage, Operating + 8 . 5 Vt o + 2 5 V
through wire to the socket pin 1 and copper
Current, Quiescent (@ tS 11:-A
: il l"l
RANGE
foil surface). A recommended shielding tech-
TEMPERATURE
Rated Performance O to +7OoC nique using model AC1214 is illustrated in
-15oc to +75oc
Operating
_55o61q +85oC Figure 1.
Storage
-- -l.sri j,'
CASE DIMENSIONS z..i;;i'o.i Best CMR performance will be achieved by
NOTES:
using twisted, shielded cable for the input signal
'Gain temperature drift is specified as a percentage of output signal level.
'Gxin nonlinearity is s?€cified as a to reduce inductive and capacitive pickup. To
Percentageof 1ov Pk-Pk outPut span
3 whcn
iolated power output is used, nonliDesity incre.ses by IO OO2%/mA of drrent drawn
1c = lv/v; with 2-pol€, 5kHz output filter (se Figute 13).
further reduce effective cable capacitance, the
'Recommended pow€r supply, ADI model 904' i 15v @ 5omA output' cable shield should be connected to the com-
Srcrfrcations subjectto chmge without notice. mon mode signal source as close to signal low
as possible(see Figure 1).

-2-
2 5 0 0 V . m s1 M I N U T E
TRANSDUCER
2500V ok or dc CONT
MODEL289
FLOATING

2500V rms 1 MINUTE


2500V pk or dc CONT

Figure 3. Model 289 Figure 4. Model 289


Terminal Capacitance Terminal Ratings
NOTE: Terminal Ratings: CMV performance is given in both peak
GAIN RESISTOR RG, l% soppmfc METAL FILMTYPElS RECOMMENOED.
FORGAIN = lv^r', LEAV€ PIN4 OPEN pulse and continuous ac, or dc peak ratings. Continuous peak
FORGAIN> lV/V, CONNECT (RG) BEruEEN PIN4 AND PIN 1
GAIN RESISTOR
c q r r u = rH*G -I K!TS
T'
9 ratings apply from dc up to the normal full power response
frequencies.Figure 4 illustratesmodel 289 ratings between
terminals.
Figure l. Basic lsolator lnterconnection
GAIN AND OFFSET TRIM PROCEDURE
The following procedure illustratesa calibration technique
THEORY OF OPERATION which can be used to minimize output error. In this example,
The remarkableperformanceof the model 289 is derivedfrom the output span is +5V to -5V and Gain = 10V/V.
the carrierisolationtechniqueusedto ffansferboth signaland
1. Apply En = o volts and adjust Ro for Eg = o volts.
powerbetweenthe amplifier'sinput stageand the restof the
circuitry.A block diagramis shownin Figure2. 2. Apply ErN = +O.500V dc and adjust R6, for E9 =
+5.000V dc.
3. Apply EIN = -0.500V dc and measurethe output error
(see curve a.,.

4. Adjust R6 until the output error is one-half that measured


in step 3 (see curve b).
5. Apply +O.5OOVdc and adjust Rs until the ourput error is
one-half that measuredin step 4 (see curve c).

: 10
tr
SYNC
IN/OUT
E -

F
l
O 5

2 _ 1 0 + 1 + 2 + 3
OUTPUT VOLTAGE Volls
Figure 2. Model 289 Block Diagram

The input signalis filtered and appearsar the input of the non-
inverting amplifier, A1. This signal is amplified by A1, with its
gain determined by the value of resistanceconnected exter-
nally between the gain terminal and the input common termi-
nal. The output of A1 is modulated, carried acrossthe isola-
tion barrier by signal transformer T1, and demodulated. The
demodulated voltage is filtered, amplified and buffered by
+v lso
L WT
I"""
amplifier A2, znd applied to the output terminal. The voltage
applied to the V5 terminal is set by the regulator ro +12V
which powers the lOOkHz symmetrical squarewave powe r Figure 5a. RecommendedOffset and Gain Adjustment
oscillator. The oscillator drives the primary winding of trans- for Gains) |
former T2.'fhe secondarywindings of T2 energizeboth input
and output power supplies,and drives both the modulator
and demodulator.
INTERELECTRODE CAPACITANCE AND TERIVTINAL +l5V
f-wv<
RATINGS 2oka I T-E;6-
-f
I ADJUST
I
C:pacitance, Interelectrode terminal capacitance, arising from
stray coupling capacitanceeffects between the input terminals
o'u, $--'*
and the signal output terminals, are each shunted by leakage
resistancevaluesexceeding 50GS). Figure 3 illustrates model
289's caoacitance.between =
PERFORMANCE CHARACTERISTICS Gain Nonlinearity: Lineariry error is defined as the deviation
Figure 6 shows the phase shift vs. frequency. The iow phase of the output voltage from the best straight line and is speci-
shift and wide bandwidth of the model 289 make it suitable fied as a % peak-to-peakourput voltage span; e.g., nonlin-
for use in SCR Motor Controller and other hieh fre quencv earity ofmodelZS9J operaring ar an ourpur spanof 1OV pk-pk
applications. (+5V) is +0.05% or 15mV. Figure 9 illustratesgain nonlin-
eariry for any output span ro 2OV pk-pk (t1OV). Figure 10
shows the effect of gain vs. gain nonlinearity.

E 120

6 2a9r

I 6 0
= 289K

> oo1
2891
2
o
z

Figure 6. Typical 289 Phase vs. Frequency


8 1 2 1 6
Figure 7 illustratesthe effect of source impedance imbalance OUTPUTSWINGPPVOLTS

on CMR performance at 6OHz for gains of lVlV, 10V/V, and


100V/V. CMR is typically 120dB at 6OHz and a balanced
Figure 9. Typical Gain Nonlinearity vs. Output Swing
source impedance.CMR is )60dB for source impedance im-
balancesup to 1OOkO.

=
3 OO r
20
o
z 0001

Figure 7. Typical 289 Common Mode Rejection vs.


Source lmpedance
10
lnput Voltage Noise: Voltage noise, referred to input, is GAIN V/V

dependent on gain and bandwidth. Figure 8 shows rms voltage


Figure 10. Typical Gain Nonlinearity vs. Gain
noise in a bandwidth from 0.05H2 to the frequency shown on
the horizontal axis. The noise in a bandwidth from 0,05H2 to
100H2 is 8pV pk-pk at a gain of 100V/V. The peak-to-peak
Common Mode Rejection: Input-to-output CMR is dependent
value is derived by multiplying the rms value at F = 10OHz
(1.2trYrms) by 6.6. on source impedance imbalance,signal frequency and ampli-
fier gain. CMR is rated at 115V ac, 6OHz and,1kQ balanced
For best noise performance in particular applications,a low source at again of 10OV/V. Figure ll illustraresCMR per-
passfilter at the output should be used to selectivelyroll- formance as a function of signal frequency, CMR approaches
off noise and undesired signal frequenciesbeyond the band- 156dB at dc with source imbalance as high as 1kS). As gain is
rvidth of interest. Increasinggain will aiso reduce the noise, decreased,CMR is reduced. At a gain of 1VlV, CMR is typi-
referred to input. cally 6dB lower than at a gain of 1O0V/V.

-ry4\
l^\,/c'.,a' lr ' "" ' ^ . i, -."
" ^ .l,l,-:l
r. 'L''\
lt.I ]Eruj "r"*-
I
1]".

; r oo L^r, I/ -

6 V
q

a -
* l 0

Figure 8. Typical lnput Voltage Noise vs.Bandwidth Figure 11. Typical Common Mode Rejection vs. Frequency

-4-
t,

MULTICHANNEL APPLICATIONS
lsolation amplifiers containing internal oscillatorsmay exhibit
a slowly varying offset voltage at the output when used in
multichannel applications.This offset voltage is the result of
adjacent internal oscillatorsbeating together. For example, if
two adjacent isolation amplifiers have oscillator frequenciesof
100.0kH2 and 100.1kHz respectively,a portion of the dif-
ference frequency may appear as a slowly varying outpur
offset voltage error. Model 289 eliminates this problem by t5v +15V

offering a synchronization terminal (pin 8). When this terminal


is interconnected with other model 289 synchronization ter-
minals, the units are synchronized.Alte rnately, one or more ,;;',";-;;i.'," ),, ActiveFi,ter
units may be synchronized to an external 10OkHz +2o/oseu&re-
wave generator by the connection of synchronization termi- Noise Reduction in Data Acquisition Systems: Transformer
al(s) to that generator.The generator output should be coupled isolators must have a carrier to pass dc signals through
2.5V-5.0V p-p with 1kf,) source impedanceto each unit. their signal transformers. Inevitably some carrier frequency
Use an external oscillator when you need to sync to an ex- ripple passesthrough to the isolator ourpur. As the bandwidth
ternal lOOkHz source, such as a sub-multiple of a micropro- of an isolator becomes alarger fraction of its carrier frequency,
cessorclock. A differential line driver, such as SN75 158, can this ripple becomes more difficuh ro control. Despite this dif-
be used to drive large clustersof model 289. When using the ficulty, the model 289 produces very low ripple; therefore,
synchronization pin, keep leads as short as possibleand do additional filtration will usually be unnecessary.However, in
not use shieldedwire. These precautions are necessaryto avoid some applications, particularly where a fast analog-to-digital
capacitancefrom the synchronization terminal to other points. converter is used following the isolator, it may be desirable to
It should be noted that units synchronizedmust share the same add filtration; otherwise, ripple may cause inaccurate conver-
power common to ensure a return path. sions. The 2-pole low-pass shown in Figure 13 limits isolator
APPLICATIONS IN INDUSTRIAL MEASUREMENT AND bandwidth to 5kHz, which is the full power bandwidth of the
CONTROL SYSTEMS model 289. Carrier ripple is much reduced. Another beneficial
Isolated DAS: In data acquisition systemswhere multiple effect of an output filter is smoorhing of discontinuous high
transducers are powered by a single supply and the magnitude frequency waveforms.
of that supply is low enough for a multiplexer to handle the Motor Control and AC Load Control: Phaseshift and band-
voltageson all the transducers,it is economical to multiplex width are important considerationsfor motor control and ac
ahead of an isolator. The fast settling time of the model 289 Ioad control applications. The model 289 possessessufficient
makes this configuration practical where slower isolators would bandwidth and acceptable phase shift for such tasks.
not be usable.
Figure 14 shows two model 289's sensingthe armature voltage
Figure 12 shows an application where the difference in voltage and current of a motor. Faithful replicasof the waveforms
between any two terminals of any of the transducersdoes not of thesevariablesare applied to the motor control. A1 oper-
exceed 3O volts. Though the input of the model 289 is pro- ates at unity gain from divided R1-R3 to deliver an ourput
tected againstline voltage, its power terminals are not; neither that is 1/100 of the armature voltage of the motor. A2
is the multiplexer so protected. This circuit will not, therefore, operatesat a gain of 1OOV/V to deliver a voltage 100 times
withstand the differential application of line voltage. that developed acrossthe current sensingshunt.
Multiplexe r addressingis binary, an enable providing selection
of the circuit shown as a signalsource.Optical isolation is pro-
vided for digital signals.When several of these circuits are used
for severalgroups of transducers,the model 289's should be
synchronized.

MOTOR
CONTROL

CURRENT
SENSE

Figure 12. DAS with MUX Ahead of lsolator Figure 14. lsolating a Motor Controller

-5-
Figure 15 shows three model 289's sensingthe voltageson the Figure 17 shows the model 289 providing an isolated 4-to-
three phasesof an ac load. The Y network shown divides the 2OmA output from a 12-bit DAC. Al.a provides a -4Y refer-
voltages of the three phases and creates a neutral for the input ence to the DAC. The digital input causesa portion of DAC
commons of the isolators,The output of each isolator is a current to flow into OUI1, causinga positive voltage at the
faithful replica of the phase of the waveform it senses.The output of A1d. A1b producesavoltage acrossR4 proportional
isolator outputs provide the feedback necessaryfor the trig- to DAC current. Alc and associatedcircuitry sink a current
ger control to correctly fire the triacs. In other applications, which is one-fourth of the full scalecurrent of the DAC,
the outputs of the isolators might have been fed to rms-to-dc causinga positive voltage of 1 volt ar the outpur of ,A'1d.
converters. W i t h t h e c o d e 1 1 1 1 1 1 i 1 1 1 1 1 , + 5 v o l t s a p p e a r sa t t h e o u r -
put of A1d. Operation is unipolar with a positive offset. The
output voltage of A1d is reproduced at the output of the
isolator, where the circuitry shown converts it into a 4-to-
20mA current which may be applied to the load R1.

qEF
ourl
E Fr N A D 7 5 4 t
GND OUT2

t02
2N257

Figure 15. lsolating a 3 Phase Load Controller

Isolated DAG: Figure 16 shows a l?-bit DAC with 15V


isolated output. A buffered -5V reference voltage is provided
to the DAC by ,A'1a,A1b and associatedcircuitry. The digital Figure 17. l2-8it lsolated Process Current DAC
input causesa proportion of DAC current to flow into OUTl
of the DAC. The remaining DAC current flows into OUT}. Temperature Measurement: Figure 18 shows the model 289
Current flowing into OUTl causespositive voltage at rhe out- providing a ground-referred output in an application measuring
put of A1c. Current flowing into OUT2 causesa positive volt- the temperature of an object floating at a high common mode
age at the output of A1d, which in rurn causesa negarive voltage. The AD590 temperanrre sensorsinks a current of
voltage at the output of ,A,1c.Voltage appearing at the output -1pA/K. This current flows into the gain terminal of the model
of A1c is reproduced at the ourput of the model 289. R5 and 289, developing+1OmV/K acrossthe internal feedbackresisror.
R8 must be adjusted to produce lessthan 0.5mV at OUTl and This voltage also appearsat the ourput of the model 289.
OUT2 of the DAC respectively.Rl5 may be used to adjustgain
The circuitry shown connected by a dotted line may be useful
and R11 to adjust offset with the binary code 100O 0OOO0000
if an output of 1OmV/"C is desired.A currenr of +2731tA
to zero.
is sourced through the 8.66k resistor and the porentiometer
cancellingthe AD59O currenr at0"C (273K), resulting in
OmV at the output at 0"C.
voD FF OUT1
REF AD7541

C O N N E C T I O N SI O G A I N T E R M I N A L S S H O U L D
B E K E P TA S S H O R TA S P O S S I B L E ,

Figure | 8. lsolated Temperatu re Measurement

You might also like