Download Application_Note_III_-_web.pdf

APPLICATION NOTE III
IPM-165 – a universal Low Cost K-Band Transceiver
for Motion Detection in various Applications
www.innosent.de
Application Note III
Editorial
Copying /publishing this note or parts of it requires the endorsement of the author
InnoSenT GmbH want provide to beginners and
first-time users an easy start in radar technology
without big disappointments and surprises and to
customers a helpful lead based on many years of own
experience.
In 1999 InooSenT was the first producer worldwide of
planar and highly integrated low cost radar modules
based on FET oscillator technology. This technique
replaced the at that time common waveguide
modules with Gunn oscillators. Seven years later
InnoSenT again was the first manufacturer who used an integrated SiGe chip,
a co-called MMIC for the transmitter part in high quantities.
Today InnoSenT produce more than 1.5 million radar modules per year for
commercial, industrial and automotive applications with a significant growth
rate and of increasing functionality and complexity.
The author Wolfgang Weidmann, Dr.Ing. and co - founder of InnoSenT
GmbH has worked for more than 35 years on radar sensing technology. His
activities in engineering and sales and marketing and the existing contacts
with customers, users and colleagues encouraged him to offer his acquired
experience to other individuals. Besides these application notes he has just
recently published a book (in German) „Rardarsensorik – schwarze Magie
oder faszinierende Technik?“, publisher Roell, ISBN 978-3-89754-411-6. It
desribes in a easily understandable way basics of radar sensing techniques
and the plurality of applications.
2
Application Note III
Table of contents
1. Introduction
2. Description of the IPM-165
Copying /publishing this note or parts of it requires the endorsement of the author
3. Handling precautions – ESD sensitivity
4. LF-amplifier circuitry
5. Pulsing the transceiver
5.1 Oscillator starting performance – pulse length
5.2 Circuitry to pulse the IPM-165
6. Frequency hopping method by pulsing the supply voltage
6.1 Frequency pushing by supply voltage
6.2 Frequency shift keying (FSK)
7. First time installation of a IPM-165
8. Trouble shooting – just in case …..
3
Application Note III
APPLICATION NOTE III
IPM-165 – a universal Low-Cost K-Band Transceiver for
Motion Detection in various Applications
1. Introduction
The mono transceiver IPM-165 from InnoSenT has been designed into various applications
and markets and is highly appreciated for its attractive price, its small dimensions, its high
sensitivity and its possibility of universal applications.
9.96
dimensions IPM-165
12.70
10.50
PIN 1
6.60 +1.00
- 0.20
antenna
25.00 +- 0.20
0.20
backside
20.20 +- 0.20
0.20
1.25
PIN 1
1.90
Copying /publishing this note or parts of it requires the endorsement of the author
2.54
PIN 1
all dimensions in mm
20.20 +- 0.20
0.20
25.00 +- 0.20
0.20
Fig 1: Photograph and dimensions of K-band transceiver IPM-165
This note shall facilitate the user, to design this
component into various applications. Besides the
usual CW (continuous wave) - mode the possibility
of pulsing is described. Signal processing circuitry
becomes pretty simple because of the straight-forward
architecture and enables the user to build cost efficient
and compact radar detectors.
block diagram of the
IPM-165
~
5V
I-Signal
Fig 2: Block diagram of the IPM 165
The circuitries discussed here should meet most of
the requirements as they occur in applications like door opening, intrusion alarm and security,
machine and equipment control, sanitary equipment and sports and toys applications. Basic
rule is that the sensor has to detect and monitor the movement or motion of an object. These
so-called “objects” might be passive items, vehicles, animals or human beings which are
moving more or less quickly. The correct circuitry will detect a motion down to quasi stillstanding.
Virtually stationary objects cannot be detected by the IPM-165, as also the direction of the
motion can’t be determined by this simple module without additional circuitry offered by
sensors with stereo (dual channel) architecture.
On the other side the sensor impresses by outstanding sensitivity. A human being i.e. can
easily be seen in a range up to 15 or 20m or even beyond. Therefore this sensor is perfectly
suited for so-called dual-technology solutions in security applications, where the advantages
of PIR (passive infrared) and radar detectors are supporting each other.
Just as an example radar and PIR detectors perform actually vice versa when an object is
moving towards or around the sensor. While a PIR detector is rather insensitive to motions
direct and straight towards or away from the sensor, the radar detector performs best there
and with highest available sensitivity. If the object is moving pretty much on a circle around the
sensor with a constant distance it is the other way round. The radar sensor loses sensitivity
4
Application Note III
because of the missing Doppler signal, while the PIR sensor shows its best sensitivity
because of the rapid change of the temperature image.
The IPM-165 can be perfectly operated with very short pulses and high pulse pause to pulse
length ratio. Therefore options become attractive like fast amplitude modulation and saving
in current consumption by keeping the average operating current low (operation with battery
and/or solar panel buffered supplies), see more information about that in paragraph 5.1.
2. Description of the IPM-165
The IPM-165 represents a highly integrated radar sensor including transmit and receive
antennae, a transmit and a receiver part. It requires one single polarity power supply only. It is
available as a 3V (IPM-365) and 5V (IPM-165) operating voltage version.
Copying /publishing this note or parts of it requires the endorsement of the author
The outstanding sensitivity is possible by two design features:
•
•
the usage of separate transmit and receive paths and antennae respectively
the usage of a balanced mixer
Sophisticated circuitry design and the selection of proper components enable the user to
operate the IPM-165 without additional temperature compensation methods and meeting
the ETSI frequency standards at the same time. Therefore the IPM-165 has got a generic CE
approval and certification.
Extracts from the IPM-165 data sheet:
Parameter
Symbol
Min.
Typ.
Max.
Units
Comment
data sheet IPM-165
Oscillator
transmit frequency
fstandard
24.050
24.250
GHz
fF
24.075
24.175
GHz
US-frequency band
fUK
24.150
24.250
GHz
UK-frequency band
output power
Pout
16
dBm
temperature drift
Δf
-1
MHz/°C
Receiver
IF output
Signal level
Noise level
voltage offset
-300
300
mV
category A
563
855
mVP-P
category B
856
1177
mVP-P
category C
1177
1819
mVP-P
116
mV
R
Antenna pattern
full beam width @ -3dB
side-lobe suppression
horizontal
80
°
azimuth
vertical
35
°
elevation
horizontal
12
dB
azimuth
vertical
13
dB
elevation
Power supply
Supply voltage
VCC
Supply current
ICC
4.75
5.00
5.25
V
30
40
mA
5
Application Note III
IPM-165 data sheet (cont‘d)
Parameter
Symbol
Min.
Typ.
Max.
Units
+60
°C
Comment
Environment
Operating temperature
TOP
-20
Mechanical Outlines
Outline dimensions
height
length
width
25
25
7 (12.7)
mm
A few more remarks regarding the IPM-165:
Copying /publishing this note or parts of it requires the endorsement of the author
The antenna patterns of the transmit and the receive antennae are identical and rather broad
to be able to detect within a pretty large angle.
The total current consumption happens exclusively in the transmit part. The given minimum
operating current cannot be lowered decreasing the supply voltage without taking the risk of
a malfunction at certain temperatures. Therefore other methods have to be used when trying
to lower the current consumption (see paragraph 5).
The signal available at the unit output is sinusodial for a monotonously moving object and will
provide very low signal amplitude (in the dimension of 300 µV). Therefore it must be amplified
immediately with high input impedance and lowest noise contribution. The load impedance
of the amplifier can be high because the internal and integrated load resistance is middleohmic. Without additional external amplification a radarsensor cannot be simply checked with
a scope since no signal can be seen on the screen because of missing scope sensitivity for
such signal levels.
3. Handling precautions – ESD sensitivity
ESD-sensitivity
Attention please! Transceivers of this architecture with direct access to the mixer output
are definitely ESD sensitive. Make sure personnel who is handling an individual unit, not
yet mounted onto a motherboard and equipment like soldering irons are protected properly
according to ESD recommendations. It starts already when taking the units out of the sealed
package. Never touch or grap the sensor at the connector pins, but only at the edges or
corners of the device.
As soon as the device has been assembled and soldered into the surrounding circuitry the
danger is gone except for stressing the mixer part directly with higher voltage than 3 kV.
External components like varistors don’t provide proper protection since the mixer diodes are
the fastest fuse you can find!
4. LF-amplifier circuitry
LF-amplifier circuitry
The load of the signal output and the following amplification of the mixer output signal is
usually done by operational amplifier stages, which are providing both – amplification and
bandwidth limitation.
Depending on the application the total required amount of external amplification can be
somewhat around 70 to 80 dB, in order to get the mixer output signal into an amplitude range
of 1V.
Generally the bandwidth of the receiver mixer of the IPM-165 is pretty high – at least 100
MHz. However in order to build a highly sensitive detector for a certain application it is heavily
recommended to limit the bandwidth of the amplified frequency band and therefore avoid
injection of unwanted noise.
6
Application Note III
The frequencies of the signals expected at the mixer output can be easily estimated by the
following famous “Doppler” formula:
Equation (1)
fDopp
f0 v
c0
α
Doppler formula
v
fDopp = 2f0 $ $ cos a
c0
fDopp = 2f0 $
v
$ cos a
c0
Doppler- or differential frequency
transmit frequency of the radar
velocity of the moving object
speed of light
angle of the direction of the object motion with the direct connecting straight line between sensor and object
Copying /publishing this note or parts of it requires the endorsement of the author
For the nominal transmit frequency of 24.125 GHz the following rule of thumb applies
fDopp = 44
Equation (2)
Hz
km
h
$ v $ cos a
When detecting humans only, the filter bandwidth may be limited to 6 to 600 Hz.
Inside rooms with light emitting lamps using discharge like neon lights in Europe a sharp filter
is required for the 100 Hz spectral line.
10µ
input
4k7
150k
1n
1n
10µ
+
Vcc=5.0 V
MC33078
10k
10µ
150k
4k7
recommended schematic
of a LF amplifier
100
+
output
MC33078
1k
10k
Fig 3: recommended schematic of a LF amplifier with 6…600 Hz bandwidth and 60dB gain
It has to be noted that if the opportunity exists to increase the capacitance values of the
capacitors used by let’s say a factor of 10, the Ohm values of the relevant resistors might be
decreased by a factor of 10, which will help to improve the noise performance and therefore
sensitivity. In this recommendation a AC coupling of the signals is used. This avoids any
problems of variations in production.
Basically the mixer output will show a DC offset, which can be used for a self-test of the unit.
However you have to be aware of the following:
Ideally the DC offset at the mixer output would be ZERO for an IPM-165, since we are
using anti-parallel mixer diodes. As discrete mixer diodes are never exactly equal in DC
characteristics, a differential DC level can be found, which can be negative or positive and
may vary up to 200 mV. Just this side effect can be used for the self-test.
To take advantage of that you add a first DC-coupled amplifier stage of low gain like 20 dB or
factor 10, de-couple this signal and then continue to amplify with AC-coupled stages with the
balance of 50 to 60 dB of gain.
7
Application Note III
5. Pulsing of the transceiver
5.1 Oscillator starting performance – pulse length
It is virtually simple enough to pulse the IPM-165 by its supply voltage. There are 2 reasons
for pulsing a sensor
•
•
to generate an amplitude modulation with 100% modulation depth
to save average current
Very fast pulsing or modulation of an oscillator is only achievable, if the oscillator starts
oscillating fast enough when applying the supply voltage. The following screen shots prove,
that the transmit oscillator of the IPM-165 is really starting rapidly.
Copying /publishing this note or parts of it requires the endorsement of the author
supply voltage ON
starting performance
IPM-165
supply voltage OFF
Dopplersignal
time scale 500 nsec/div
supply voltage ON
supply voltage OFF
Dopplersignal
time scale 50 nsec/div
Fig 4: Starting performance of an IPM-165 by fast pulsing the supply voltage with 1µsec pulse
Obviously transmitter and receiver both start within 100 nsec, while the switch-OFF time may
take 200 nsec. In order to be on the safe side (variations in production) we recommend not to
decrease the pulse length below 1 µsec.
If pulsing the voltage supply is used for current saving, the pulsing must be selected carefully
to re-generate the Doppler signal. According to Shannon’s sampling theory a signal of
frequency f has to be sampled at least with double the frequency for correct re-construction.
8
Application Note III
Example:
In an intrusion alarm application sampling the frequency range (i.e. 5….500 Hz) of the Doppler
signal generated by a human being requires at least 1000 Hz as sampling rate corresponding
to 1 msec. A pulse / pause ratio of 1:1000 results in a reduction of the average current by a
factor of 1000. This leads to the above mentioned minimum pulse length of 1 µsec.
5.2 Pulse circuitry for IPM-165
Pulsing the supply voltage can be done simply by switching a MOSFET.
5V
pulse circuitry for the
IPM-165
FDV304P
PLS
Vcc
Fig 5: pulse circuitry for the IPM-165
The pulse can be applied by a CMOS or TTL gate circuit.
6. Frequency hopping method by switching of the supply voltage
6. 1 Frequency pushing by supply voltage
Each oscillator will change its oscillation frequency when its supply voltage changes. The
change depends on how the frequency is generated. Since the frequency stabilisation of the
IPM 165 is rather simple, the change of operating frequency is in the area of a few MHz when
changing the supply voltage up to 10%.
By the way this effect has to be considered when stabilizing the supply voltage by an
integrated voltage regulator, which provides the output voltage within certain specified limits.
Decreasing the supply voltage beyond 10% to lower values is not recommended since the
start of oscillation at temperature extremes might be influenced or even prohibited.
Frequency Pushing
24,140
24,130
frequency [GHz]
Copying /publishing this note or parts of it requires the endorsement of the author
10k
24,120
24,110
24,100
4,5
4,6
4,7
4,8
4,9
5
5,1
5,2
5,3
5,4
Vcc [V]
Fig 6: Oscillation frequency change as function of supply voltage for an IPM-165
5,5
9
Application Note III
As an example a change in supply voltage by 5% causes a change in oscillation frequency of
about 6 MHz.
We must point out that the frequency pushing effect varies from module to module. The
variation of oscillation frequency has its root cause in the variation of the transmit transistor
impedance. Therefore InnoSenT cannot guarantee a specification for frequency pushing. It is
therefore recommended to use this effect for frequency variation only if the absolute value of
frequency change over voltage supply is uncritical.
6.2
Frequency shift keying (FSK)
Copying /publishing this note or parts of it requires the endorsement of the author
The effect of frequency pushing mentioned can be
used for frequency shift keying (FSK) modulation,
a way of frequency hopping. Again the comment
that the modulation change varies from module
to module. Fig. 7 shows the very clean spectrum
of a FSK-modulated IPM-165, where a modulation
change of about 2.5 MHz was generated by pulsing
the supply voltage.
FSK with the IPM-165
Fig 7:
frequency spectrum of FSKmodulated IPM-165, generated by
switching the supply voltage.
We recommend the following switching circuitry:
5V
10k
recommendation for
switching circuitry for the
IPM-165
10
FDV304P
PLS
Vcc
Fig 8: Recommended circuitry for FSK modulation of an IPM 165 by switching the supply voltage. Pulse
generation can be done by a CMOS or TTL gate circuit.
In this schematic the supply voltage is switched between the full voltage available from the
supply and a value reduced by the voltage drop generated by the operating current (about
30mA). Since this operating current varies from module to module, also this voltage variation
is subject to production variations.
In the literature this method is mentioned to detect the direction of a motion by FSK. In this
case the phases of the two Doppler signals generated by two different transmit frequencies
have to be evaluated for its sign. In our opinion the effort for that is much higher and will be
dependent on production variation compared with the solution using a stereo (dual channel)
module (i.e. IPS-154 from InnoSenT).
There is an opportunity to frequency-modulate the module with a triangle or sawtooth signal
applied to the supply voltage. Some sort of range gating can be achieved, always keeping
in mind that the production variations will be limiting the performance. An introduction to
frequency modulation of radar modules is available as our application note AN2 “Detection of
moving and stationary objects”-
10
Application Note III
7. Operating module IPM 165 for the first time!
moving object or
Dopplersimulator
(e.g. IDS-208)
operating module
IPM-165
IF - amplifier
2
1. 5
1
Copying /publishing this note or parts of it requires the endorsement of the author
Scope
0. 5
0
−0. 5
−1
−1. 5
−2
0
0. 1
0. 2
0. 3
0. 4
0. 5
0. 6
0. 7
0. 8
0. 9
1
Power-Supply
Fig 9: Circuitry for functional test of a IPM 165
•
•
•
•
•
•
•
•
Be sure you apply the correct supply voltage to the module (plus 5 or plus 3V)
Connect the module output with a low frequency amplifier with at least 60 dB
gain before feeding the input of a scope with a sensitivity of at least 50 mV/
division.
Move your hand in about 60 cm to 1m (2 to 3 feet) distance in front of the
antenna.
You should see a sinusoidal, not frequency-constant, but clean signal on the
scope screen providing a good signal/noise ratio
For a more professional investigation and analysis you may purchase InnoSenT’s
Dopplersimulator IDS-208 (contact by tel. 49-9528-9518-84).
This device simulates a monotonously moving object in one direction by
electronic means, with which you can test the radar module. Other simple test
methods may also be fine for a first glance like a turning fan.
You may optimize your radar sensor by changing the gain of your LF amplifier
and by adjusting the amplifier bandwidth.
Operating the radar sensor close to neon lights the generated 100 Hz
interference signal must be thoroughly filtered out.
8. Trouble shooting – just in case ………..
We presume you have connected your IPM-165 module according to fig. 9.
1. case: an output signal is definitely not existing – zero amplitude!
•
•
•
Troubleshooting!
The well-known answer: did you connect the supply voltage correctly – correct
polarity? The module hasn’t got any polarity protection!
Is your low frequency amplifier operating perfectly?
Is the scope input sensitivity set to the correct, while lowest value?
2. case: The output shows a pretty noisy signal while your movements in front of the sensor
just generate no signals at all or only a weak sinusodial signal
•
In this case the probability is pretty high, that the internal mixer diodes have
suffered from an exposure by ESD and have been damaged. Repair impossible,
11
Application Note III
a new sensor is due! Next time, please pay more attention when unpacking
and mounting.
3. case: the output does show a clean sinusodial signal, but with very low amplitude
•
•
Are gain and bandwidth of your LF-amplifier correct?
You are moving your hands outside the antenna pattern of the sensor.
4. case: The output shows a strong, constant signal with single frequency
•
•
Copying /publishing this note or parts of it requires the endorsement of the author
•
Try to calculate the frequency of the signal. Do you find neon lights nearby if the
signal is 100 Hz and you haven’t implemented special filtering?
Any other electrical interferer? Connecting cables to power supply, LF-amplifier
and scope too long and causing instability?
Mechanical interference by rotating parts, i.e. a fan in summertime, which might
be operating as a reflecting and moving object generating a Doppler signal?
12
Don’t hesitate to contact us directly if you got further questions!
Tel: +49 (0) 9528 / 95 18 0 | E-Mail: [email protected] | www.innosent.de