Patent Application
20200003882
The field of the invention relates to a communication unit, such as a radar unit, that includes a distributed master-slave arrangement and methods for clock distribution and synchronization. The invention is applicable to, but not limited to, a radar unit for, say, an automotive application, and method therefor.
There has been an increased demand for active safety systems for vehicles. Active safety systems require multiple radar sensors per vehicle, each radar sensor typically working with a specific radar technology. In an automotive application, the radar sensors are mostly built using a number of integrated circuits (ICs), sometimes referred to as ‘chips’. The current trend is towards offering a radar system on chip (SoC), using a radio frequency (RF) complementary metal-oxide-semiconductor (CMOS) process technology) solution in order to reduce cost and power consumption.
Frequency-modulated continuous wave (FMCW) radar sensors transmit frequency modulated signals, and radar receivers substantially simultaneously receive their echo. The received echo is then mixed with the transmitted signal and results in a low frequency signal having a frequency of:
at the output of the mixer, the so-called beat frequency (Fb). By analysing the beat frequency, the range parameter (i.e. a distance to targets) can be extracted.
Commercial automotive radar sensors typically include multiple receivers and transmitters (the combination of which is referred to as transceivers (TRx)). A microcontroller (MCU) performs digital control of the transceiver circuits and digital signal processing of the digitized data (e.g. fast fourier transform (FFT) and digital signal processing) in order to output processed radar data to the MCU of the vehicle.
Next generation high performance radar solutions used for highly automated or fully autonomous driving will need to comply with stringent radar angular resolution requirements in both azimuth and elevation. Angular resolution is directly related to the radar system number of receiver antennas and their location with respect to each other. Current monolithic radar transceiver ICs are typically constrained to contain only a few transceiver channels, as the complexity and cost and heat dissipation problems of integrating more transceiver channels on the same chip increases substantially with the number of ICs that are included.
To allow increased angular resolution in both azimuth and elevation, a multi-chip set solution with a master device and several slaves may be used, as illustrated in
Each master device 110 and slave device 120 embeds several receiver and transmitter channels, and a microcontroller unit (MCU) is used to combine all received data from all receivers, as well as to control and program the master device 110 and slaves devices 120. The master device 110 is arranged to distribute the Local Oscillator (LO) signal 140 off-chip through transmission lines on the printed circuit board (PCB) 100 to other radar chips (in this case the one other TRx chip functioning as a slave device 120). The LO signal 140 is used for the different transmitters and receivers, and is typically star-routed (i.e. routed via equal length paths) to each device to guarantee the same delay and exact phase coherence between all devices (which are sometimes separate ICs). The slave device 120 (and further slave device(s) 123) contains a second set of transmitter circuits coupled to transmit antennas 122 and a second set of receiver circuits coupled to receive antennas 124.
The distribution of the LO signal 140 from the master device 110 ensures that the slave device 120 is also able to use the LO signal 140, and thereby ensure that the transmitting signal frequency and the clock frequency of the down mixer of different radar TRx are the same. The LO signal 140 is used by all devices (including the master device 110 wherein the LO signal 140 is routed out of the master device 110 and thereafter back into the master device 110. Typically, in master-slave arrangements, the LO signal 140 is routed with symmetrical PCB lengths in order to ensure that all receivers (encompassing a respective down mixer) in each master device 110 and slave device(s) 120, 123 of the system receive the same LO with same phase. Phase coherence is mandatory for cascaded systems. Other control signals may be synchronized with a lower speed clock, for example an analog-to-digital converter (ADC) clock, which may be used across multiple ICs/devices.
The master device 110 is coupled to the MCU 160, which includes various interfaces, such as a serial-parallel interface (SPI) 162, a general purpose data input-output port 170, a ramp frame start (RFS) circuit 168, as well as a MCU clock interface 164, this clock signal 165 generally provided by the master device 110.
The RFS signal can be either (i) generated and distributed by the MCU 160, through a dedicated pin on both the MCU 160 (RFS_out) and a single pin on master device(s) 110 and slave device(s) 120, 123 (RFS_in) or (ii) generated and distributed by the master device 110, where there is no pin needed for MCU 160 in this case.
In the first known architecture (i), as illustrated in
In the second known architecture (ii), when the RFS is generated from the master device 110: the signal to start the generation of RFS comes from a SPI command from the MCU 160. Thus, no RFS at the MCU side is needed in this case. The RFS is then generated by the master device 110. The connection from master device 110 to slave devices 120, 123 for routing that RFS signal can be performed in two different ways:
A master-slave (MS) clock signal 142 is generated by the master device and used as a time base for synchronization of the microcontroller time based events with any other master device and all slave devices 123. The connection to the MCU 160 consists of SPI control lines 163 and digital data line signals from all the ICs back to the MCU 160 for later signal processing, in a given data format (e.g. Mobile industry Processor Interface Camera Serial Interface (MIPI CSI-2), low voltage differential signalling (LVDS) or other formats).
The cascading master-slave (MS) clock signals (MS_clkP, MS_clkN) 142 are specifically used for time based synchronization of the sampling moments on the ADCs of several master and slave devices.
For optimal operation of distributed radar systems, it is important that these signals (LO 140, MS clock signal 142 and RFS signal 166) are synchronous across all receiver circuits on different devices. This means that, after calibration of the phase difference between all receiver channels in all master and slave devices, the voltage supply Vcc, temperature and aging variation between all devices distributed in the PCB should not change the initial clock alignment, for example after a one-time calibration operation is performed at radar module level.
A primary problem with such cascaded systems is clock distribution and synchronization. Problems arise due to any misalignment between clock valid and frame start timings between master device and slave device(s), due to different printed circuit board (PCB) delays between components, devices, circuits. This problem is typically overcome by adopting a MS clock distribution process in a star-connection architecture, in order to guarantee the same transmission line lengths and therefore a good clock alignment between the components, devices, circuits. The star-connection architecture is also required for the LO signal to achieve phase coherence when all the TRx channels are used as one antenna array. The requirement to adopt a star-connection architecture for all common signals (LO signal, MS clock, etc.) imposes severe constraints on the PCB design. A common and synchronised phase reference is particularly important in a distributed radar phased array design, in order to steer a beam and control a radiation pattern for the phased array system.
Synchronised clock signals are also used to sample data inside the ADCs of each device. In an FMCW radar device, the synchronized clock edges are used to sample data and control signals that are used to start the modulation. Here, the RFS signal across all devices must be synchronized. This alignment is needed to avoid sampling valid data at different times/moments in the master and slave devices, which would create phase errors between devices and consequently compromise radar system performance.
Even when adopting a star-connection architecture, a problem occurs when the RFS signal edge is close to a MS clock edge. Particularly in response to processing, voltage or temperature (PVT) variations, a master device and one or more slave device(s) may observe RFS re-timed at a different clock edges. This means that for an intermediate frequency (IF) of, say, IF=20 MHz, an error of a complete 240 MHz MS clock cycle can happen. This leads to a phase error of around +1-15 degrees, which compromises radar angular resolution. Modern imaging radar systems specifications allow a maximum of +/−3 degrees of phase error, and this considers errors due to combined ADC sampling and receiver channel variations, as well as some PCB asymmetries. This translates to a stringent required synchronization time accuracy of around 416 psec., for a maximum intermediate frequency (IF) of 20 MHz.
U.S. Pat. No. 6,775,328 B1 proposes a feedback synchronization loop using mV input-output drivers and receivers, U.S. Pat. No. 9,031,180 B2 provides synchronization via a protocol (data frame) in wireless transmitters, U.S. Pat. No. 7,876,261 B1 proposes synchronization between devices using reflected wave clock synchronization. U.S. Pat. No. 6,209,072 B1 uses a de-skewing latch technique in order to obtain a synchronous interface between master device and slave devices. Each of these known techniques is complex and/or require further components and circuits, thereby requiring valuable PCB space and increased cost.
Thus, a mechanism is needed to better support synchronization between shared clocks signals (that use a master-slave clock) and a ramp frame start (RFS) signal in radar units that use multiple radar devices or ICs.
The present invention provides a communication unit, such as a radar unit, and methods for clock distribution and synchronization in a master-slave arrangement, as described in the accompanying claims. Specific embodiments of the invention are set forth in the dependent claims. These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.
Further details, aspects and embodiments of the invention will be described, byway of example only, with reference to the drawings. In the drawings, like reference numbers are used to identify like or functionally similar elements. Elements in the figures are illustrated for simplicity and clarity and have not necessarily been drawn to scale.
Because the illustrated embodiments of the present invention may, for the most part, be implemented using electronic components and circuits known to those skilled in the art, details will not be explained in any greater extent than that considered necessary as illustrated below, for the understanding and appreciation of the underlying concepts of the present invention and in order not to obfuscate or distract from the teachings of the present invention.
To address the aforementioned synchronization problem between multiple devices or ICs within a communication unit, such as a radar unit, examples of the present invention propose a design to embed a frame start signal, such as a chirp start signal (RFS) of a radar unit, within the clock signal. Thereafter, the embedded signal may be shared between the master device(s) and slave device(s) (or ICs) in order to maintain synchronization that is significantly less reliant on PCB construction. Advantageously, the technique to embed a chirp start signal within the distributed master-slave clock signal (sometimes referred to as an ADC clock) enables the slave device(s) (and any other master device(s)) to readily and easily demodulate the distributed master-slave clock signal in order to obtain concurrently the chirp start (RFS). Thus, the master-slave clock distribution and the timing alignment between the master-slave clock signal and the RFS across all devices is inherently synchronized.
Such a technique benefits from the fact that the master-slave clock signal is quasi-perfectly synchronized (when using a star routed PCB) and delays of clock edges between different devices during voltage or temperature variations (once offset is calibrated) are minimal, e.g. less than one clock cycle error. Thus, in some examples, embodiments of the invention enable a designer to implement a digital cascading radar solution with increased phase aligned performance, where a single reference oscillator (e.g. phase locked loop (PLL)) may be used for all master and slave devices (supplying a reference frequency/clock and a main PLL frequency clock), thereby facilitating phase coherence, clock alignment and clock and RFS synchronization throughout the communication units.
In examples of the invention, a communication unit is described that includes a plurality of cascaded devices that includes at least one master device and at least one slave device configured in a master-slave arrangement. The at least one master device includes a modulator circuit configured to: receive a system clock signal and a frame start signal; modulate the system clock signal with the frame start signal to produce a modulated master-slave clock signal; and transmit the modulated master-slave clock signal to the at least one slave device. The at least one slave device comprises a demodulator circuit configured to: receive and demodulate the modulated master-slave clock signal; and re-create therefrom the system clock signal and the frame start signal.
Some examples of the invention enable the RFS signal (ramp frame start or chirp start) to be generated in the master device for synchronicity purposes, and not in the MCU as adopted in known communication units. This approach advantageously removes a need for a dedicated pin in the MCU and removes an additional routing constraint of this signal in the PCB, because the RFS signal is embedded within the distributed MS clock. In this example, the embedding of the RFS signal into the master-slave clock signal (thereby facilitating a distribution of the two signals with effectively no delay between the clock edges of each within the devices), may be achieved using a RFS modulator and a RFS demodulator, implemented in the master-slave clock distribution transmit and receive circuits, respectively.
Some examples of the invention may employ low-voltage differential signalling (LVDS), also known as TINEIA-644, in order to assist the modulation and demodulation of the embedded clock signal and chirp signal (i.e. low-to-high or high-to-low transitions) to provide synchronicity across different master device(s) and slave devices. LVDS is a technical standard that specifies electrical characteristics of a differential, serial communications protocol. LVDS operates at low power with programmable output amplitude of voltages and can run at very high speeds using inexpensive twisted-pair copper cables. In some examples, LVDS may be employed in order to reduce interference or signal pollution between supplies, due to current spikes, if CMOS logic were used. Some examples of the invention may take advantage of the fact that LVDS amplitude levels (Vdiff) can be programmed, for example by programming a LVDS transmitter current value (Vdiff=2*Idc_tx*Rload), assuming a constant 100 ohms termination load (Rload), placed in each LVDS receiver.
Thus, examples of the invention may facilitate digital cascading radar units with increased phase aligned performance, where a single reference PLL may be used in a clock signal generation with embedded frame alignment of frame start signals within a MS clock for distribution between all master and slave devices. This may ensure optimized phase coherence, sampling clock and chirp start signal alignment, which reduces system phase error, thereby allowing accurate range resolution.
Although examples of the invention are described with reference to synchronization in a star-connected master-slave architecture for a radar unit, it is envisaged that, in other examples, the techniques described herein may be applied to any kind of communication unit or system where several devices or ICs share the same sampling timing, and where another signal may be sent and received embedded within the clock, but synchronous to it (in this example radar scenario, this signal is the ‘chirp start’ signal).
Although examples of the invention are described with reference to a use of cascaded integrated circuits for, say, a phased array vehicular radar system with many transceiver circuits not located in a same IC, it is envisaged that the examples herein described may equally be employed in a phased array system for general wireless communication applications and units, such as base stations.
Next generation radar solutions will be based on a multi-chip transceiver configuration where a master device and several slave devices are cascaded coherently to increase the number of transmitter and receiver channels, thereby increasing sensor accuracy in angle resolution. Digital cascading is a preferable solution in order to circumvent the limitation of analog cascading in terms of misalignment of sampling clock of the ADCs and modulation ramp start signal (RFS) between different radar chips.
Referring to
The radar unit 200 contains one or several antennas 202 for receiving radar signals 221, and one or several antennas 203 for transmitting radar signals, with one shown for each for simplicity reasons only. The number of antennas 202, 203 used may depend on the number of radar receiver and transmitter channels that are implemented in a given radar device. One or more receiver chains, as known in the art, include receiver front-end circuitry 206, effectively providing reception, frequency conversion, filtering and intermediate or base-band amplification, and finally an analog-to-digital conversion. In some examples, a number of such circuits or components may reside in signal processing module 208, dependent upon the specific selected architecture. The receiver front-end circuitry 206 is coupled to the signal processing module 208 (generally realized by a digital signal processor (DSP)). A skilled artisan will appreciate that the level of integration of receiver circuits or components may be, in some instances, implementation-dependent.
The microcontroller unit (MCU) 214 maintains overall operational control of the radar device 200, and in some examples may comprise time-based digital functions (not shown) to control the timing of operations (e.g. transmission or reception of time-dependent signals, FMCW modulation generation, etc.) within the radar unit 200. The MCU 214 is also coupled to the receiver front-end circuitry 206 and the signal processing module 208. In some examples, the MCU 214 is also coupled to a memory device 216 that selectively stores operating regimes, such as decoding/encoding functions, and the like.
As regards the transmit chain, this essentially comprises a power amplifier (PA) 224 coupled to the transmitter's one or several antennas 203, antenna array, or plurality of antennas. In radar unit 200, radar transceiver topology is different from traditional wireless communication architectures (e.g. Bluetooth™, WiFin™, etc.), as modulation occurs within a phase locked loop (PLL) (typically via a fractional-N divider), and is applied directly to the PA 224. Therefore, in some examples, the receiver front-end circuitry 206 and transmitter PA 224 are coupled to frequency generation circuit 230 arranged to provide radio frequency (RF) local oscillator (LO) signals. The generated RF LO signals are thus modulated directly to generate transmit radar signals, and also used to down-convert received modulated radar signals to a final intermediate or baseband frequency or digital signal for processing in a receive operation.
In examples of the invention, digital cascading of multiple master and slave devices is achieved by sending a reference clock signal generated in a master device to both other master device(s) and one or more slave device(s). In examples of the invention, the reference clock signal has been adapted by embedding a start of a frame (RFS) indication in the clock signal, which in some examples is a master-slave clock signal. In some examples, the generation and distribution of the MS clock signal with an embedded start of a frame indication may be implemented via a star connected LVDS link, for example to reduce PCB skew and reduce signal reflections that could compromise signal integrity and increase jitter.
To remove any misalignment between RFS and the MS clock signal, as well as relax PCB requirements on the RFS signal, examples of the invention embed the RFS (sometimes referred to as a modulation chirp start) inside the master-slave clock signal, by adding an RFS modulator in a LVDS transmitter for generating a reference clock, and by adding an RFS demodulator at a LVDS receiver side. Thereafter, for example, amplitude modulation/demodulation may be used to retrieve the RFS signal, advantageously without a need for one or more dedicated RFS pin(s).
In accordance with examples of the invention, the use of improved synchronization between respective master device(s) and one or more slave device(s) may then allow correct operation of the communication unit, e.g., that implements an N-unit phased array FMCW imaging radar system. For identical units, the use of a phased array increases the sensitivity by ‘N’ times, as compared to that of a single unit. In a radar transmitter mode of operation, beam steering can also be implemented by shifting the phase of the transmitting signals of an N-unit phased array radar system. For multiple mid-range radar units, the inputs to an N-unit phased array radar system can be combined in order to increase the radar transmit power, and thereby allow longer range applications. Such benefits may result from the use of improved synchronization between respective master device(s) and one or more slave device(s).
In one example of the invention, the embedding of a chirp start signal within a clock signal, which is shared between the master device(s) and slave device(s) (or ICs) in order to maintain synchronization, may result in an enabling of more flexible PCB design without the constraints of a strict transmission line or connection length between the master device(s) and slave device(s). Furthermore, a smaller MCU for, say, a phased array FMCW imaging radar system, may be achieved by eliminating a need for an additional one pin on MCU (when the RFS is provided by the MCU in the known architecture (ii)) or two dedicated RFS pin(s) on the master device (when the RFS is provided by the master device in known architecture (i)).
Some examples of the invention employ synchronization in order to minimize any delay mismatch in either the LO/frequency generation circuitry and/or analog-to-digital converter (ADC) sampling instants (which would ordinarily have led to angle estimation errors) as well as transmit control, frequency chirp start control signals. Examples of the invention employ synchronization in an FMCW radar unit, particularly to align the frequency chirp start and end frequencies.
Furthermore, examples of the invention allow multiple radar transceiver chips to be located on different PCBs or modules in order to reduce installation complexity in a constructing of an adaptive phased array FMCW imaging radar system on a non-planar vehicle surface. Clearly, the various components within the radar unit 200 can be realized in discrete or integrated component form, with an ultimate structure therefore being an application-specific or design selection. A skilled artisan will appreciate that the level of integration of circuits or components may be, in some instances, implementation-dependent.
Referring now to
In this example, the master device 310 incorporates a radar transceiver formed by at least a frequency generation circuit, which in this example includes a reference phase locked loop (PLL) 350 and a main PLL 317. An output from the main PLL 317 provides a local oscillator (LO) signal to a local oscillator output pin 318, which is routed via a LO_out path 315 to a RF splitter 307 that is coupled to each master and slave LO input pin (LO_in) 319, 329 via, say, an equal transmission line length (e.g. via a star configuration). In some examples, by using equal transmission line lengths, it is possible to ensure equal delay and phase aligned signals for the LO, applied across all master device(s) 310 and slave devices 320, 323 for the generation and reception of radar signals.
The master device 310 further includes a digital controller 316 and a transmitter circuit 312 comprising one to several transmitter channels (TX_CH) and a receiver circuit 314 comprising one to several receiver channels (RX_CH). In a transmitter mode of operation in the master device 310, the digital controller 316 may provide a transmit radar signal to the main PLL 317 to generate a modulated transmit signal. The modulated transmit signal is then optionally passed to a buffer or frequency multiplier in the transmitter circuits 312, 322 (if the generated signal is not at the operating frequency of the radar unit 305) of the respective master device(s) 310 and slave device(s) 322. A high-frequency output of the transmitter circuit 312 is passed to a power amplifier, say, via a phase shifter circuit (both not shown), where it is amplified within transmitter circuit 312 and routed to the one or more transmitter antenna(e) 311 (and in some examples routed to the one or more transmitter antenna(e) 325 in the one or more slave device(s) 320.
In a receiver mode of operation in the master device 310 and slave device 320, a received radar signal may be received at the one or more receiver antenna(s) 313, 321 and passed to master and slave receiver circuits 314, 324 that include a low noise amplifier (LNA) configured to amplify the received radar signal. The amplified received radar signal is passed to a down-mixer, where it is mixed with the received high-frequency LO signal 315 received from the master device 310.
The master device 310 further includes one or more programmable bandpass filter(s) and one or more gain amplifiers (not shown), as well as one or several ADC(s) 342 that is/are coupled to the one to several receiver channels (RX_CH) 314 as well as a serial interface 344. The one or several ADC(s) 342 process a relative narrow band signal, for example in a range between a few kHz up to tenths of MHz.
In accordance with examples of the invention, the master device 310 further includes a master-slave clock distribution and synchronization circuit 360. In this example, the reference clocks are generated on the master device 310 using the reference PLL 350, tied to an external crystal oscillator 352, and sent as a first clock signal (clk1) to the main PLL 317 and a master-slave clock signal (clk2) 382 to master-slave clock distribution and synchronization circuit 360.
In this example, the master-slave distribution and synchronization circuit 360 is configured to embed a frame start signal, such as a chirp start (RFS_in) signal 380 in a radar unit, into the master-slave clock signal (clk2) 382 received from reference PLL 350. Embedding the chirp start (RFS_in) signal 380 into the master-slave clock signal (clk2) 382 allows the two signals to be distributed across all master slave devices with inherent synchronization and effectively no delay between devices. In this example, the result of the embedding operation is a (distributed and received modulated) differential master-slave clock signal 384. In this example, master-slave distribution and synchronization circuit 360 uses a RFS modulator 362 (implemented in, say, a LVDS transmit circuit) to embed the chirp start (RFS_in) signal 380 into the master-slave clock signal (clk2) 382, and a RFS demodulator 364 (implemented in, say, a LVDS receive circuit) to respectively de-embed (e.g. demodulate) the differential master-slave clock signal 384 into a slave RFS_out signal 390 and a master-slave slave clock signal 388. Similarly, in the master device, a fed back (distributed and received modulated) differential master-slave clock signal 384 is also received in master-slave clock distribution and synchronization circuit 360.
In response to a serial-parallel interface (SPI) command coming from the MCU 214, the digital controller 316 of the master device 310 provides a chirp start signal (rfs_in) 380 that will be used by each master device 310 and each slave device(s) 320, 323 of the radar unit 305. The rfs_in signal 380 is provided to the master-slave clock distribution and synchronization circuit 360, and it is embedded on the master-slave clock signal (clk2) 382 by the RFS modulator 362. In this example, the RFS modulator 362 is an LVDS transmitter circuit that produces a modulated/embedded output LVDS signal. In one example, as illustrated, the modulated/embedded output LVDS signal may be a differential master-slave clock synchronization signal with embedded RFS signal 384 (MSclk_n & MSclk_p). In this example, the RFS modulator 362 (and RFS demodulator 364) may be terminated by a 100 ohms differential resistor (T), as well as all other slave LVDS clock RFS demodulators, in order to reduce any reflections due to LVDS link asymmetry, which may corrupt the signal integrity and cause false zero crossings.
In accordance with examples of the invention, the differential master-slave clock synchronization signal with embedded RFS_in signal 384 (i.e. containing both clock & chirp start) may then be propagated to a number/all devices of the radar unit 305 via an equal length (L2) star connection and splitter 387, in order to limit any printed circuit board (PCB) skew to very low values (e.g. <100 psec).
In this example, each radar slave device 320, 323 includes a digital controller 326 and a transmitter circuit 322 comprising one to several transmitter channels (TX_CH) and a receiver circuit 324 comprising one to several receiver channels (RX_CH). The slave device 320 further includes one or several ADC(s) 341 that is/are coupled to the one to several receiver channels (RX_CH) 324 as well as a serial interface 334. In accordance with examples of the invention, the slave device 320 further includes a slave-based master-slave clock distribution and synchronization circuit 361, where only the LVDS receiver is enabled. In this example, the slave-based master-slave distribution and synchronization circuit 361 is configured to demodulate and extract the chirp start (RFS) signal 380 from the differential master-slave clock synchronization signal with embedded RFS_in signal 384. In this example, slave-based master-slave distribution and synchronization circuit 361 uses an RFS demodulator 364 implemented in LVDS.
In some examples, as appreciated by a skilled person, the slave device 320 may be configurable as a master device 310, but with a number of circuits or components or functionality disabled. For example, in such a slave device 320, since the clock and sync signal as well as LO signals are received from the master device 310, a slave reference PLL, main PLL and the MS_CLK transmit circuits are disabled. Many other blocks may be present inside the slave device, but are not mentioned herein to avoid obfuscating concepts of the invention.
The respective RFS slave demodulator 364 is configured to extract respective slave clock signals (clk_rxS 388) in a CMOS format with very low delay variation between that clock and the master clock signal (clk_rxMS 385). In addition, the respective RFS demodulators 364, 365 are configured to demodulate the transmitted chirp start signal (rfs_in 380) in each respective device. Since the respective RFS master modulator 385, generates the clock signal (clk_rxM) and RFS master demodulator 365 and RFS slave demodulator 364 demodulate the respective master clock signals (clk_rxM 385) and slave clock signals (clk_rxS 388) that are very closely aligned, they are used to re-sample the respective master (rfs_outM) 386 and slave (rfs_outS) 390 device chirp output signals. In this manner, the RFS master demodulator 365 and RFS slave demodulator 364 create synchronized chirp start signals (rfsSync_M 392, rfsSync_S 394) at respective inputs at time engines 370, 372. In this example, in the slave device, this sampling operation is performed in the digital domain, inside respective digital controller 326, by flip-flop 332. In the master device, this sampling is performed inside digital controller 316, by flip-flop 330.
A skilled artisan will appreciate that many other circuits, components and blocks may be present inside a master device 310 and slave device 320, but that these have been omitted purely for clarity purposes. In this manner, a system, such as a phased array FMCW imaging radar system, employing example embodiments of the invention relating to a master-slave arrangement, may be able to benefit from improved clock synchronization accuracy in a multiple chip configuration.
In some examples, the master device 310 is configured to generate and distribute a master-slave clock 384 from a MS_clock LVDS transmit modulator 326 path to the MS_clock LVDS receive path inside slave devices, and back to master device LVDS RX demodulator 364, in, say, a star configuration. In this manner, in some examples, an ADC clock applied to both master device(s) and slave devices may be perfectly aligned, so long as the transmission line length to each of the receivers is matched. Furthermore, in some examples, the clock signal with an embedded start of a frame indication may be used to sample the data inside different ADCs belonging to different ICs, in order to reduce phase error and consequently angle resolution at sensor level.
Also for example, in one embodiment, the illustrated examples of a master device 310 or slave device 320 may be implemented as circuitry located on a single integrated circuit 398, 399. Alternatively, the circuit and/or component examples may be implemented as any number of separate integrated circuits 398, 399 interconnected with each other in a suitable manner. The specifications and drawings are, accordingly, to be regarded in an illustrative rather than in a restrictive sense.
Referring now to
The reference clocks are generated on master device 310 using the reference PLL clock signal (clk2) 382, tied to an external crystal oscillator 352, and sent to RFS modulator 360. Under, say, a SPI command from the MCU 214, the digital controller 316 of the master device 310 provides a chirp start (RFS_in) signal 380 to the RFS modulator circuit 360, which is configured to embed (e.g. encapsulates) the chirp start (RFS_in) signal 380 into the master-slave clock signal (clk2) 382 received from reference PLL 350. In this example, the embedded chirp start (RFS_in) signal 380 takes a form of differential master-slave clock signal (MS_clkdiff) 384, which is then used by all the master device(s) 310 and slave devices 320 of the system. In
In some examples, the use of LVDS facilitates programmability of output voltages and supports very high data speeds. In some examples, LVDS may be employed in order to reduce interference or signal pollution between supplies due to current spikes if CMOS logic were used.
This differential master-slave clock signal (MS_clkdiff) 384 containing clock & chirp start is then transmitted by the master device 310 to all devices of the system, and is received by both master device(s) 310 and slave devices 320. Master-slave clock distribution and synchronization circuits 360, 361 are configured to extract the clock signals (clk_rxM 385 & clk_rxS 388) in a CMOS format with very low delay variation between both clocks, as well as to demodulate the transmitted chirp start signal (RFS_in) 380 in both the master device(s) 310 (to produce a reproduced chirp start signal rfs_outM 386) and slave devices 320 (to produce a reproduced chirp start signal rfs_outS 390). Since the respective clock signals present a very good alignment, as illustrated, advantageously with very low absolute delay and delay variations, they are used to re-sample the reproduced chirp start signals rfs_outM 386 & rfs_outS 390. In this manner, synchronized chirp start signals (rfsSync_M 392 and rfsSync_S 394) are created at inputs to the respective time engines 370, 372. In some examples, this sampling is performed in the digital domain, inside digital controller 316, 326 by flip-flops 330, 332.
In some examples, in order to obtain this re-sampling in a digital domain, at least one of the at least one master device clock receiver and demodulator 360, and at least one slave device clock receiver and demodulator 361 may be arranged such that clock extraction circuitry that generate signals (clk_rxM, clkrxS) do not introduce a problematic delay (GM, GS 420) due to, say, any of process, voltage, temperature (PVT) variations, for example where the variations during a communication unit design phase are constrained to be smaller than one clock cycle (Tclk) 410, to avoid introducing re-sampling errors on RFS_signals (rfs_outM, rfs_outS). In some examples, a clock extraction circuitry generates signals (clk_rxM, clkrxS) that are only allowed to introduce a delay (GM, GS<Tclk 420) of up to half of a clock period of the system clock signal (388, 385) due to any of process, voltage, temperature, PVT variations. The inventors of the present invention have identified that a more acceptable ratio is one eighth of a clock period delay of the system clock signal 388, 385. Any variability is well controlled to guarantee that the datavalid signals (datavalidM 396, datavalidS 398), which are used to set sampling time of data at ADC side, are accurately time aligned on all master device(s) and slave devices, with a maximum delay (d) defined by the overall radar phase variation specification (d=<R*Tclk) 430.
In one example operation, it is envisaged that the radar units that include a built-in inter-chip localization feature may be employed in radar units in vehicles. Examples of the invention may be additionally employed to support beam steering control in a phased array radar system across multiple sensors, in some examples having multiple chips.
Examples of the invention propose a method for clock distribution and synchronization in a communication unit having at least one master device and at least one slave device configured in a master-slave arrangement and configured to process at least one of: transmit signals, and receive signals. The method includes, at a master device: generating a system clock signal and generating a frame start signal. The method further includes embedding the frame start signal into the system clock signal to produce a modulated embedded master-slave clock signal. The modulated embedded master-slave clock signal is then transmitted to at least one slave device to synchronise the system clock signal and a frame start signal between the at least one master device and at least one slave device.
Referring now to
In examples of the invention, at 508, a master device starts a FMCW ramp signal and embeds a RFS_in signal within a MS clock signal. In some examples, the MS clock signal is then sent to all other master device(s) and slave device(s). At 510, the MS clock signal with an embedded RFS_in signal is received and demodulated in each of the master device(s) and slave device(s). At 512, the ramp (chirp) signal starts in each of the master device(s) and slave device(s). At 514, each of the master device(s) and slave device(s) receive a representation of an echo signal and send validated, demodulated data to the MCU based on the synchronised clock signal with an embedded RFS_in signal. In this manner, synchronization between the master device(s) and slave device(s) is achieved, after which, all the master device(s) and slave device(s) have the same frequency chirp start and chirp end frequencies, as well as the same sampling clock. The data sampled by all of the devices (or ICs) including the ADC is now substantially synchronous. The flowchart ends at 516.
Thus, examples of the invention describe a communication unit (such as a radar unit) that includes a plurality of cascaded devices that comprise at least one master device and at least one slave device configured in a master-slave arrangement and configured to process at least one of: transmit signals, and receive signals. The at least one master device includes: a clock generation circuit; a modulator circuit coupled to the clock generation circuit and configured to receive a system clock signal and a frame start signal. The at least one master device is configured to embed the frame start signal (for example a chirp signal) into the system clock signal to produce a modulated embedded master-slave clock signal. The at least one master device transmits the modulated embedded master-slave clock signal to the at least one slave device to synchronise the system clock signal and a frame start signal between the at least one master device and the at least one slave device. In other examples, at least one master device may be employed with any number of slave devices. In this situation, the at least one master device and one or more slave device(s) may present a different number of receiver and transmitters channels. The slave devices can be programmed to have only receive channels on, whilst the transmit channels are enabled by the master device.
Although examples of the invention are described with reference to a radar unit suitable for an automotive application, it is envisaged that the concepts herein described may be applicable to other applications, such as radar for robotics or drones.
In the foregoing specification, the invention has been described with reference to specific examples of embodiments of the invention. It will, however, be evident that various modifications and changes may be made therein without departing from the scope of the invention as set forth in the appended claims and that the claims are not limited to the specific examples described above. The connections as discussed herein may be any type of connection suitable to transfer signals from or to the respective nodes, units or integrated circuit devices. Accordingly, unless implied or stated otherwise, the connections may for example be direct connections or indirect connections. Also, plurality of connections may be replaced with a single connection that transfers multiple signals serially or in a time multiplexed manner. Likewise, single connections carrying multiple signals may be separated out into various different connections carrying subsets of these signals. Therefore, many options exist for transferring signals.
Those skilled in the art will recognize that the architectures depicted herein are merely exemplary, and that in fact many other architectures can be implemented which achieve the same functionality. Any arrangement of components to achieve the same functionality is effectively ‘associated’ such that the desired functionality is achieved. Hence, any two components herein combined to achieve a particular functionality can be seen as ‘associated with’ each other such that the desired functionality is achieved, irrespective of architectures or intermediary components. Likewise, any two components so associated can also be viewed as being ‘operably connected,’ or ‘operably coupled,’ to each other to achieve the desired functionality.
Furthermore, those skilled in the art will recognize that boundaries between the above described operations merely illustrative. The multiple operations may be combined into a single operation, a single operation may be distributed in additional operations and operations may be executed at least partially overlapping in time. Moreover, alternative embodiments may include multiple instances of a particular operation, and the order of operations may be altered in various other embodiments.
In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word ‘comprising’ does not exclude the presence of other elements or steps then those listed in a claim. Furthermore, the terms ‘a’ or ‘an,’ as used herein, are defined as one, or more than one. Also, the use of introductory phrases such as ‘at least one’ and ‘one or more’ in the claims should not be construed to imply that the introduction of another claim element by the indefinite articles ‘a’ or ‘an’ limits any particular claim containing such introduced claim element to inventions containing only one such element, even when the same claim includes the introductory phrases ‘one or more’ or ‘at least one’ and indefinite articles such as ‘a’ or ‘an.’ The same holds true for the use of definite articles. Unless stated otherwise, terms such as ‘first’ and ‘second’ are used to arbitrarily distinguish between the elements such terms describe. Thus, these terms are not necessarily intended to indicate temporal or other prioritization of such elements. The mere fact that certain measures are recited in mutually different claims does not indicate that a combination of these measures cannot be used to advantage.
| Number | Date | Country | Kind |
|---|---|---|---|
| 18305856.9 | Jul 2018 | EP | regional |
