CO1000A/00

System Overwiew

The CO1000A/00 is an interface for evaluating Sivers IMA FMCW modules of the type RS3400. The evaluation system is intended to enable a user to perform high quality measurements and to evaluate the capability of the RS3400 front-end in a very short time. Using a PC, the user can fully control the RS3400 and perform measurements. Typical accuracy for range measurement is 2mm (standard deviation of measurement on target at a distance of 10m from the RADAR). The unit can also be used for accurate velocity measurements, having the RF output frequency locked. Depending on the intended application, an antenna to transmit and receive the RF signal will be necessary. For simple evaluations, SiversIMA can provide also the antenna. The evaluation system key components are the RS3400 FMCW RADAR front-end (called: RS3400) and a digital control and acquisition board (called: control board). The RS3400 is a medium-cost, synthesized, X-band, FMCW radar front-end. It is designed for intrinsic safety giving a reliable unit with low power consumption. The RF output is capable of a 1500MHz sweep, which is controlled digitally via a standard, 3-wire serial interface. Frequency accuracy is better than +-35ppm over the temperature range and settling time is less than 50 μs for the output signal frequency to settle within 5% of the frequency step. The digital control and acquisition board provides correct bias for the RS3400 and conditions the analog IF output signal. It is based on a micro controller responsible for communication with the RS3400 synthesizer and the host PC. It also controls a high accuracy ADC for data acquisition synchronized with the frequency setting of the RS3400. Measurements can be synchronized using both a trigger input and a trigger output port.

Assembling an Evaluation System

The evaluation system is delivered with the RS3400 mounted on the control board as shown in Figure 1. Power is connected through the 2.5/5.5mm DC connector. Positive voltage, 10 12V, should be applied on the center pin. The control board is protected against reverse voltage. In order to prevent undesired noise, a linear power supply is recommended, but for simple tests, a switched supply may suffice. Communication between the PC and the control board is done using a RS232 serial connection. Connect a straight serial line between the female DSUB9 of the control board and the PC. If the PC does not have an RS232 interface, USB to RS232 adapters are available in regular computer stores. Installing this adapter is usually very simple. If requested, the adapter can be included with the system.

To test the communication with the system, start a terminal window on the PC (Hyper terminal on Windows works well) and configure the communications port where the control board is connected to the PC. The port can usually be identified using Windows device manager, select “Ports (COM & LPT)”. Native ports are usually labelled COM1, COM2, COM3 or COM4. For USB-adapters, the COM-number is somewhat higher. The configuration of the port should be as follows: • Bits per second: 115200 • Data bits: 8 • Parity: None • Stop bits: 1 • Flow control: None Make sure that the terminal program started, the port configured correctly and the control board is connected to the serial line. Connect power to the control board. A green LED on the control board should light up. If the communication is working, a message should be displayed looking similar to the following: SiversIMA AB FMCW Eval board initialized Software version: B RS3400 If the message is not displayed, try pressing the RETURN key a few times. The character “?” should be displayed. If this is not the case, see the troubleshooting guide at the end of this section. With the communication established between the computer and the control board, controlling of the RS3400 is possible and measurements can be performed. Some kind of waveguide needs to be attached. This is typically an antenna, but may be other devices depending on application. For ranging purposes, a 15-20dB standard gain horn is recommended. Bearing in mind that the sweep bandwidth is 1500 MHz, the antenna needs to have a fairly high bandwidth. It is very important that the antenna is connected with high quality cabling, capable of the operating frequency of the RS3400 since otherwise high reflections and attenuation will occur that will impede the quality of the measurements.

Trouble shooting guide for communication between the control board and a PC

Symptom	Possible Cause	Solution
The LED of the control board is not lit	Power is not supplied to the control board.	Connect 10-12V on the center pin of the 2.5/5.5mm DC connector
The terminal program or the control board is not responding when the enter key is pressed	The serial cable between the PC and the control board is not properly connected	Connect cable
	The terminal program has not connected to the remote system.	Initiate communication within the terminal. For Hyper terminal, select menu Call and click on Call (alt+C, C)
	The communications port is not configured correctly	From within the terminal program, verify that the correct communications port is selected and that it is configured as described above.
	The cable between the PC and the control board is incorrect.	Try shifting the lines going to the DSUB9 connector pin 2 and 3.

The FMCW RADAR measurement principle

The FMCW RADAR differs from classical pulsed RADAR systems in that an RF signal is continuously output. Consequently, time of flight to a reflecting object can not be measured directly. Instead, the FMCW RADAR emits an RF signal that is typically swept linearly in frequency. The received signal is then mixed with the emitted signal and due to the delay caused by the time of flight for the reflected signal, there will be a frequency difference that can be detected as a signal in the low frequency range. A schematic presentation is shown in Figure 2.

Figure 2. Schematic presentation showing how a low frequency signal is generated by mixing the received RF signal with the output RF signal. Due to the delay Δt caused by the distance travelled by the emitted signal to the reflector and back to the receiver, there will be a small difference in signal frequency between the two RF signals. This is output as an IF-signal with frequency Δf. A simplified derivation of the intermediate frequency (IF) signal with the frequency Δf can be made in the following way. Assume that the RF signal generator will output a frequency that is changing linearly in time as:

where f RF0 is the starting frequency, T is the frequency sweep time and kf is the slope of the frequency change, i.e., the sweep rate:

where BW is the frequency sweep bandwidth. The delay caused by the roundtrip of the emitted signal to the reflector is calculated as:

where d is the distance between the RADAR antenna and the reflector and c is the speed of light. Due to the delay, the frequency of the received signal compared with the emitted signal will be:

The difference in frequency, Δf, between f RFReceived and f RFOut is thus:

This is the signal that is output from the detector. The minus sign can be omitted since the real signal frequency output from the RADAR detector is wrapped to a positive frequency. Thus the expression can be written as:

Typical values for the RS3400 would be a frequency sweep of BW=1500MHz in T=75ms corresponding to a sweep rate of kf = 20 000 MHz/s. A distance between the RADAR and a reflector of d=15m would give a delay of Δt = 0.1 μs and the IF signal frequency would then be Δf = 2 000 Hz. This signal is easily sampled with a high resolution ADC and detected. If several reflectors are appearing in the measurement setup, the resulting IF signal will contain a superposition of the individual IF-signals from the echoes. The different echoes are distinguished by their unique IF signal frequency and a Fourier transform of the sampled signal can be used to extract the distances to the different echoes. The measurement range that the system is capable of is limited by the sensitivity of the detector and the sampling rate of the ADC. For the RS3400 a sampling rate of 20 kHz gives a maximum detectable IF signal frequency of 10 kHz, which corresponds to a range of 75m. Higher ranges are easily achievable by either increasing the sample rate or lowering the sweep rate. In addition, antenna gain needs to be fairly high in order to provide sufficient signal levels for the detector. The fundamental range measurement resolution of the system can be estimated as follows. The Fourier transform of a time limited signal can only detect IF signal frequency with a resolution of 1/T, keeping in mind here that Δt<<1/T and thus the sampling time can be approximated by T. Using eq. 6, this gives that minimum change in d, Δd, is found to be:

which can be transformed to:

thus, the resolution is only limited by the sweep bandwidth. This is an important observation since it is says that resolution is not dependent on the frequency of the RF signal itself, but rather only on the sweep bandwidth. There are methods of increasing the resolution of the measurements by a factor of 10 to 100 using fitting algorithms to find a peak in the IF signal spectrum that is not at an integer frequency point defined by the sampling rate and sweep bandwidth. The range detection and FMCW RADAR principle may also be derived using a characterization of the IF signal phase rather than the frequency. This is recommended in order to understand the possibilities of a discrete system where the frequency sweep really is generated by a discrete set of frequencies. This derivation also lends itself more directly for high resolution range measurements. For simplicity of understanding the measurement principle it is however not necessary and is thus left out of the main text, but is included as an appendix.

Performing an FMCW RADAR measurement

The following shows a typical measurement sequence and some trivial data manipulation in order to retrieve a distance measurement. The unit is connected to a standard gain horn antenna using a one meter RF cable. A metal sheet, used as a radar reflector, is positioned approximately zero, one and two meters away from the horn during the three measurements. Starting the units shows: SiversIMA AB FMCW Eval board initialized Software version: B RS3400 Start by initiating the equipment. Setup measurement, use default for most parameters. Position the radar reflector close to the horn antenna. Send the following commands to initialise the system, enable measurement and set the number of frequency sweeps to one: INIT SWEEP:MEASURE ON SWEEP:NUMBERS 1 Initiate a measurement sweep: TRIG:ARM Measurement completes in approximately 85 ms and data can be retrieved: TRACE:DATA ? Download the 1501 integers and store in a file s0.txt. Position the radar reflector at approximately 1m from the horn. Initiate a new sweep. TRIG:ARM Download the measurement data and store in a file s1.txt. TRACE:DATA ? Position reflector at approximately 2m from the horn. Initiate a new sweep. TRIG:ARM Download the measurement data and store in a file s2.txt. TRACE:DATA ? Using some math package (e.g., Matlab), data can be plotted and processed. Plotting the three data files will give the following result. As can be seen from Figure 3 to Figure 5, the data is typically an oscillating signal with an amplitude of approximately 10e3 digital units. It appears as if the closest echo gives the highest amplitude. Performing a Fourier transformations of the data sets give spectral results as is shown in Figure 6. With a frequency span of 1500MHz, the resolution of the frequency bins will correspond to a distance of 0.1m (accounting for a two way path). As can be seen in Figure 7, the peaks are separated by approximately 10 bins, which corresponds to a shift in position of the reflecting sheet of 1m. It is also evident that all signals contain an echo positioned approximately at bin 16. This is most likely caused by the connection between the RF cable and the horn antenna. The 16 bins passed from the RF connector of the FMCW module and the connector of the horn relates to the one meter RF cable containing a teflon dielectric, which gives an equivalent length in vacuum of approximately 1.6 meters.

Figure 3. Plot of s0. Signal corresponds to a very close echo and possible reflections in the measurement equipment.

Figure 4. Plot of s1. Signal corresponds to an echo positioned approximately one meter away from the horn and possible reflections in the measurement equipment.

Figure 5. Plot of s2. Signal corresponds to an echo positioned approximately two meter away from the horn and possible reflections in the measurement equipment.

Figure 6. Plot of spectrum (absolute magnitude in dB).

Figure 7. Close up of Figure 6, peaks indicated with markers.

Controlling the FMCW unit

The following section describes the commands that are available for controlling the evaluation kit. Communication can either be performed using a terminal program for manual control or by sending and receiving characters over the serial port using a program for automated measurements. When automated measurements are performed it is important to make sure that the input and output buffers are properly emptied when reading and sending characters.

Startup

When the unit is started a message is displayed, i.e., transmitted on the serial line: SiversIMA AB FMCW Eval board initialized Software version: B RS3400K/02 2008-06-29 This means that the unit is ready to control the module and to perform measurements. The following text will briefly describe the commands that are available. Please note that this message is sent on the serial line regardless of if there is a terminal window attached or not. If an automated system is configured, this has to discard the message from the input buffer before reading any data from the unit.

Commands available on the control board

All commands are built up in a hierarchical structure where there are seven main categories: • FREQUENCY (control frequency setting parameters) • HELP (provides a simple list of available commands) • INIT (initialises the unit) • MEASURE (controls measurement parameters) • SWEEP (controls sweep parameters) • TRACE (handles data generated by the unit) • TRIGGER (controls trigger parameters) Some main-categories have under-categories. These are reached by entering the main category and a colon “:” followed by the under category. All categories where data can be entered can also return data, typically the previously entered data. Some categories can only return data, typically measurement data. Data is returned by entering the full category name (main:under) followed by a question mark “?”. Please note that there should be a white space between the category name and the question mark. All data, where applicable, is entered and returned in SI units, e.g., frequency is entered in Hz and time is entered in seconds. The commands are not case sensitive.

Main categories

FREQUENCY

Controls frequency setting parameters. Note: several parameters are related and the setting of one parameter may change the setting of others. As an example, changing SPAN will affect both START and STOP. Under categories. CENTER Sets the center frequency of the sweep. Usage: FREQUENCY:CENTER 24.7e9 FREQUENCY:CENTER ? Default: 24.75e9 (24.75GHz) SPAN Sets the frequency span of the sweep. Usage: FREQUENCY:SPAN 500e6 FREQUENCY:SPAN ? Default: 1.5e9 (1.5GHz) START Sets the start frequency of the sweep. Usage: FREQUENCY:START 24.4e9 FREQUENCY:START ? Default: 24.0e9 (24.0GHz) STOP Sets the stop frequency of the sweep. Usage: FREQUENCY:STOP 25.4e9 FREQUENCY:STOP ? Default: 25.5e9 (25.5GHz) STEP Sets the frequency step, i.e., frequency separation between two contiguous frequency points of the sweep. Usage: FREQUENCY:STEP 0.5e6 FREQUENCY:STEP ? Default: 1e6 (1MHz ) Max: - Min: 8e3 (8kHz) POINTS Sets the number of frequency points to use during a sweep. Usage: FREQUENCY:POINTS 301 FREQUENCY:POINTS ? Default: 1501 Max: 1501

HELP

Provides a simple list of available commands. Under categories None.

INIT

Initialises the RS3400 unit. Does not restore variables to default. Under categories None

MEASURE

Controls measurement parameters. Under categories CHANNEL Sets the channel to be used when measuring during a frequency sweep. This feature is currently not active.

SWEEP

Controls sweep parameters. Note: several parameters are related and the setting of one parameter may change the setting of others. As an example, changing TYPE will affect the number of data points acquired when a measurement is performed and TIME will affect IDLE. Under categories

TYPE

Sets the type of sweep to be performed. Available types are SAWTOOTH and TRIANGULAR. The type SAWTOOTH is a sweep of linearly increasing frequency. When the highest frequency is reached, the sweep restarts at the lowest frequency. The type TRIANGULAR is a sweep of first linearly increasing frequency and then linearly decreasing frequency. The sweep then restarts with an increasing frequency. When a measurement is made during a TRIANGULAR sweep, data is recorded both on the positive and the negative frequency slope. Thus the number of data points is twice that of SAWTOOTH and also twice of what is entered at FREQ:POINTS. Usage: SWEEP:TYPE TRIANGULAR SWEEP:TYPE ? Default: SAWTOOTH

NUMBERS

Defines the number of frequency sweeps to be performed when triggered. The measurement of the IF signal is averaged over the number of sweeps. If 0 is entered, the sweep continues indefinitely, but can be aborted by sending a capital Q. Please note that the currently started sweep will be finished and communication with the unit is not possible before that. Usage: SWEEP:NUMBERS 5 SWEEP:NUMBERS ? Default: 0 (infinite number of sweeps, finish with “Q”)

MEASURE

Defines whether a measurement of the IF signal should be performed during the sweep or not. Usage SWEEP:MEASURE ON SWEEP:MEASURE ? Default: OFF

TIME

Sets the total time for a complete sweep. This time should be chosen long enough to allow a sufficient time at each frequency point, see SWEEP:IDLE. Usage: SWEEP:TIME 1.0 SWEEP:TIME ? Default 0.075 (75ms)

IDLE

Defines the time spent at each frequency point. In principle, the RF frequency is set at the beginning of this time and a measurement of the IF signal is performed at the end of this time. Usage: SWEEP:IDLE 10e-3 SWEEP:IDLE ? Default: 50e-6 (50 μs) Min: 50e-6 Max: 1 Please note that a high setting of IDLE time will give a very long frequency sweep when the number of frequency points is big.

TRACE

Controls data generated by the unit. Under categories DATA Returns measurement data. Data is an array of floating point numbers. The array length is controlled by SWEEP:NUMBERS (see SWEEP). Usage: TRACE:DATA ? Please note that TRACE:DATA ? will return data for a complete measurement. The number of lines that are returned is controlled by FREQUENCY:POINTS and SWEEP:TYPE. If SWEEP:TYPE is set to TRIANGULAR, twice the number of data points is returned. This may overflow the input buffer of the serial connection of the computer if it is not set up correctly. The output will be ended with a separate line containing OK indicating that transmission is completed.

TRIGGER

Controls trigger parameters. Under categories.

SOURCE

Defines the source that will trigger the start of a frequency sweep. Available settings are IMMEDIATE, meaning that no trigger signal is necessary, and EXT0. The EXT0 is connected to the Trigger input pin of the control board. A transition from low (0V) to high (3.3V) will trigger a measurement. Usage: TRIGGER:SOURCE IMMEDIATE TRIGGER:SOURCE ? Default: IMMEDIATE.

ARM

Puts the unit in a state ready to start the defined frequency sweep. Certain parameters are calculated and the module is set to the start frequency. If TRIGGER:SOURCE is set to IMMEDIATE, the frequency sweep is started as soon as the unit has completed the necessary preparations. Usage: TRIGGER:ARM

OUTPUT

Defines whether or not a trig signal should be output when a frequency sweep starts. Trigger output is available as a 0/3.3V signal on the Trigger out on the control board.

DELAY

Sets the length of output trigger pulse. The actual frequency sweep does not start until the trigger pulse has completed and consequently this can be used as a delay between input trigger and actual start of the frequency sweep. If a trigger output is enabled, it will be issued for each single sweep. Usage: TRIGGER:DELAY 10e-3 TRIGGER:DELAY ? Default: 10e-3 (10ms) Min: 50e-6 Max: 1

Appendix

The FMCW RADAR measurement principle based on phase measurements (time discrete version). The FMCW RADAR works by outputting a continuous RF signal, whose frequency is swept over a specific frequency band. Synthesized modules, like RS3400, are in fact not sweeping the frequency continuously, but are rather stepping the frequency with a set of discrete frequency points. Thus, these systems are also called Stepped Frequency Continuous Wave (SFCW) RADAR. The synthesized signal source assures a very precise frequency control, which is very important for the accuracy and repeatability of the measurements. The RF signal will be radiated and reflected against different objects. The echo is then received and compared (mixed) with the radiated RF signal. Had the system been measuring time-of-flight for a pulsed signal, the sensor output could be linear with distance. In an FMCW, however, the sensor output corresponds to the cosine of the phase difference between the echo signal and the radiated signal.

where s is the output signal from the sensor and Φ denotes phase difference between the echo RF signal and the radiated signal. Put in other words, the measurement signal from the sensor will be a cosine signal indicating the electrical distance, roundtrip, which the radiated signal has travelled.

where d is the distance to the reflecting object and λ is the electrical wavelength of the RF signal. The multiplication by 2 accounts for the roundtrip. The expression for Φ can also be written

where λ has been substituted with

, c is the speed of light and fRF is the frequency of the RF signal. For any measurement, but ones with very short distances, the electrical distance will exceed one wavelength and there will be ambiguities about the measurement result. Fortunately, the RF signal can be stepped in frequency and several measurements can be performed. From eq. 12, it is clear that Φ will increase linearly with f RF and thus the detector output will be a cosine shaped signal. A small value of d, meaning a close echo, will create a slowly varying detector signal and a distant echo will create a quickly varying detector signal. The frequency of the RF output signal, fRF, is stepped over the available band (BW) For the RS3400X, this band is from 9250MHz to 10750MHz and for the SY3400K, this band if from 24000MHz to 25500MHZ, i.e., BW=1500MHz. The expression for Φ then looks like:

where n indicates each unique measurement, n = 0, 1, ..., N-1 and N is the number of frequency points used for the measurement sequence. The term fRF0 denotes the starting frequency. Recalling that the detector output is the cosine of Φ, it will have the following appearance:

Since the cosine function only is unique over a range of [0, π], the expression can be simplified to:

where Φ0 accounts for the phase value at the starting RF signal frequency of the sweep. It can be noticed that Φ0 can be limited between 0 and π. To extract the value of the distance to the reflector, d, one needs to estimate how much Φ changes over the frequency sweep. One simple way is to take the Fourier transform of the signal s(n).

here, m denotes the normalized index in the transformed domain, m=0, 1, ..., (N-1). With the detector signal being a time varying signal, m can be seen as the index in the frequency domain for the detector signal. As an alternative, this domain may be seen as a distance domain. For simplicity, this domain will be called frequency domain or spectrum in this text since it in most cases is the typical interpretation of a Fourier transformed signal. Recalling that the Fourier transform of a cosine yields two Dirac-delta functions, the transform of s(n) becomes:

The second term on the right hand side refers to a peak at a negative value of m, this can be easily converted to a positive value by adding N, but is of no further interest here. The first term on the right hand side will have a peak at m=(2d*BW/c). Conversely, if a peak is identified at m=m0, the corresponding distance to the reflector can be calculated as:

Thus, a ranging function is achieved. It is worth noticing that eq. 18 is completely independent of the center frequency of the RF signal. In fact, to get an impression of the resolution that is available from only extracting a measurement of the distance based on the maximum peak found in the frequency domain, the difference between two integer values of m is c/2BW. This shows that the resolution is only depending on the sweep bandwidth of the RF signal and not of its specific frequency. A frequency sweep from 9250 to 10750 MHz will give the same resolution as a frequency sweep from 24000 to 25500MHz. For the specified frequency sweep of BW=1500MHz, the integer range resolution will be 0.10m. In order to achieve higher resolution in the range measurements, a weighted average of several frequency points can be used to find a peak location that is positioned between integer points in the spectrum. Also, when a single echo is available in a local part of the spectrum, it is possible to estimate d based on the slope of the phase angle Φ. Using only the first term in the right hand side of eq. 17, recall that the inverse Fourier transform of a Dirac-delta function is a complex exponential:

Here, the right hand side is a complex series of N points. In a real-life measurement, the signal will not be an ideal complex exponential like in eq. 19. However, an inverse Fourier transform of only a section from the spectrum around a peak of interest will give a complex signal whose phase angle may be extracted. Using this phase angle, the slope may be found using for example a least square fit to a linear expression.