4T8R MIMO System Phase Problem

Question

Hi, I am currently using the ADF5901, ADF5904, ADF4159, and ADRF5300 to design a 4T8R MIMO radar system. Previously, I successfully verified the feasibility of a 2T4R system without using the ADRF5300. In this design, I utilized a Wilkinson power splitter to distribute the LO signal to two ADF5904 chips for achieving 8RX. Two ADRF5300 switches allow one ADF5901 to control four channels in a time-division manner. 
 During the design and testing phase, I found that the quality of the IF signals for all channels was acceptable; however, there were issues with angle-of-arrival estimation. After extensive validation, I suspect that the four signals generated by the ADF5901 may be problematic. Theoretically, for a 4T8R array, the phase trends of the 8-channel data obtained from each TX transmission should be consistent. However, testing revealed significant issues with this consistency. It&rsquo;s not that a specific TX channel is unstable; rather, the phase trends from different TXs exhibit variations across different data collection sessions. 
 I would like to inquire about the potential causes of this issue. Below are the phase results from three signal acquisitions, where 1 to 8 represent the 8 RX channels, and different curves indicate the phase changes of the received signals corresponding to different TX transmissions. In the embedded control system, both RX and LO are continuously active. TX is controlled through register 0 of the ADF5901, switching channels 1 and 2, while each channel is managed by the ADRF5300 for switching to form four transmit paths, with a channel switch occurring every ten chirps. This part of the embedded control is illustrated in the diagram. The STM32 is used for control, and the history represents the ramp output from the technical ADF4159's MUXOUT. The high and low levels of PA11 and PA10 are used to control the two ADRF5300 switches. 
 I would appreciate your assistance in understanding why different TXs produce varying phase trends. Thank you!

emrecangidik · Accepted Answer

Hi 
 I don't have much experience with RF switches and radar FEs. This might come short but I have a theory. 
 ADRF5300 is a reflective switch. Therefore, there might be disturbances while changing the load. This might create a phase mismatch. 
 How do you control chirps? It is possible to ignore the first couple of chirps and start sampling afterward. Or you can use the TXData input to start the chirp after waiting for a couple of microseconds for the switch to settle. 
 
 Additional note; I checked and found that ADRF5020 is a recommended non-relective switch for 24GHz FMCW arrays. 
 
 Thanks Emrecan)

emrecangidik · Answer

Hi, 
 Depending on the application, yes it is acceptable to discard some chirps. For example, the EV-TINYRAD24G Reference design discards some chirps while setting ADC. 
 Frankly, I don't know whether ADRF5020 is more suitable. 
 But I know that VGA or PA are susceptible to load shift and a reflective switch might change the phase. However, after the switch is settled, the phase should return to normal value. Thus, I proposed skipping a couple of chirps after the switch change or using the TXData input of the ADF4159 to wait for the switch to settle. 
 Thanks, Emrecan

emrecangidik · Answer

please refer to below thread. 
 ADF5901 output power - Q&A - RF and Microwave - EngineerZone 
 ADF5901 output power issue. - Q&A - RF and Microwave - EngineerZone 
 Thanks, Emrecan

SLA Status	Assignee	Support Status
SLA Closed. Met after 47 hours	zwang emrecangidik All mike.oshea	Assigned

Popular Forums

Product Forums

Application Forums

Design Center Forums

Highlighted Webinar

Places

Libraries

Upcoming Learning & Events

Challenge Yourself!

What's Brewing

Places

Resources

Visit ContentZone

The Latest Read

New Release

Recent Technical Insights

SLA Information

Products Mentioned

Associated Drivers

Thread Details

4T8R MIMO System Phase Problem

Thread Summary

Before You Switch