What is the average gain, maximum gain, polarization efficiency?

What are the left and right rotation components, the axial ratio, and the phase angle?

What is the relationship between test distance and accuracy?

What is the relationship between throughput and TRPTIS, test distance, and path loss?

How to choose multi-probe, single-probe, plane near field, spherical near field and compact field?

With the questions of these antenna R&amp;D partners, we arranged a technical interview with Ren Xiangshun, general manager of Tianjin Feitu Technology. Mr. Ren has been engaged in RF automation for more than 20 years and is one of the leading antenna testing experts in China. The following is a transcript based on the recording.

Edit:Okay, let's talk about something else. Now, when you make a router, you need to test the throughput and simulate the test distance. Does this have any conversion relationship?

appoint:Double the distance by 6dB, it's that simple.

Edit:Specifically?

appoint:For example, if you are testing a throughput or an EIRP, and a value is measured at a spatial distance of 3m, another value will be measured when the test distance is extended to 6m, and the strength difference between the two values ​​will be 6dB. To 12m will add 6dB. Converted back to throughput is that for every 6dB of attenuation added, the simulated test distance is doubled. The EIRP test means that the test distance is doubled, the air loss is increased by 6dB, and the EIRP seen from the spectrum analyzer will be reduced by 6dB, and the EIRP will remain unchanged. This is only theoretical. In reality, it is also affected by the surrounding environment. This is what it looks like in outer space.

Edit:Isn't it 3dB to double the distance? How is 6dB?

appoint:Go online and look up the Frisian transmission formula. In terms of the relationship between distance and intensity, you can think of it this way. There is a reservoir that releases fish in a ray shape. You use a net to catch fish, and you get a lot of fish. If you double the distance now, you need four times the size of the fish. You can catch as much fish as you go with the net, right? This quadruple is 6dB. This is the same as that spherical integral, as well as the closed spherical space integral.

Edit:So what does throughput have to do with TRPTIS?

appoint:There is a certain relationship, I am not an expert in WiFi development, I just know what to say, throughput is a thing, if you continue to increase attenuation, you can see that there is a downward trend from 0 attenuation to a certain degree, and then Until it is completely disconnected, it is equivalent to placing it far away and unable to communicate. You can't feel it when you are close. I have tried to unplug the router antenna before and it can connect to WiFi. It's amazing, and I don't know where it comes from. Doing a good job in TRPTIS is helpful for communication under the condition of a certain distance. In fact, many communication systems are like this. The quality of the antenna research and development can only be seen when the signal is poor. What is it called, it can be used in difficult times. In the past, the boss of Huaqiangbei put his mobile phone in the safe to make a call. It was simple and rude but very effective.

Edit:We have talked about so many antenna parameters, so how do you choose a test site?

appoint:This depends on what our testing needs are. You can't have both, and it all depends on the input-output ratio. For example, if you are engaged in active testing, you can get more probes to test quickly, so you can do this. For example, to develop small antennas, what to do with glue stick antennas, navigation antennas, etc., it is good to get a single probe test. If you want to make a large antenna, you have to consider a large darkroom or time-consuming near-field testing. , or simply engage in one on the roof. If you are developing a car antenna, you have to consider whether to get a car to drive in or near-field testing at an outdoor temporary site. This is only for the selection of development sites for commonly used civilian antennas. If the frequency is extended to above the C-band, it is the frequency above 8G If it is used for millimeter wave communication, military products and automotive radar, etc., there are various kinds of things. Because of the existence of the far-field test distance formula, the test of this frequency band with short wavelengths often cannot reach the ideal test distance, so it is required to use other The technical means are adapted to the new requirements.

Edit:What technical means? I haven't touched these.

appoint:Such as plane near field, spherical near field, compression field, PWC, lens field and so on.

Edit:Is there any difference?

appoint:The near field means that it occupies a small space and has a good quiet zone, which can be tested indoors. But the time is too long. Many times we only care about the information of a few special slices, such as the gain of the horizontal slice, the gain pointing angle, the beam width, the sidelobe and so on. However, in order to obtain the information of these slices, it is necessary to measure the data of the entire plane or the entire sphere. For example, the base station antenna test is concerned with the horizontal section and the vertical section, and 3D data must be tested in the near field of the multi-probe, but only two turns are required in the far field. As a result, the final test time is almost the same.

Edit:What are the other ones?

appoint:The compression field is to use the reflecting surface to convert spherical waves into approximate plane waves. PWC is to use a method similar to phased array to transmit plane waves with large-area radiation units. Lenses similar to telescopes can also convert spherical waves into plane waves.

Edit:What is a lens?

appoint:Have you heard of Lumberg Lens? It is something like this. We have also done related research. It may be because of our materials. The phase centers of different frequencies are different. Yes, in fact, this thing is still very useful, save money and the test speed is very fast, but because of this phase center problem, the frequency sweep test cannot be performed.

appoint:Plane wave conversion, the problem with this thing is that it is not easy to widen the frequency band, and the matching is not easy. If each unit is equipped with a phase shifter, the cost will be unbearable. I have only seen one type of RS push, and other companies have not seen it. I also rarely participate in exhibitions, and I don't know how the latest technology is.

Edit:What is the match?

appoint:Matching is very important. It is required for the development of all phased array antennas. Simply put, the radiation characteristics of all the elements of the phased array are exactly the same. It is useful for far-field matching, and useful for plane near-field direct matching. There is also the field configuration for the near-field inversion caliber, each village has its own strategy.

Edit:Does it match the phase?

appoint:Amplitude and phase are equivalent to the array element calibration of the array antenna. For example, we generally use a plane near-field probe to align each array element, and then control the phased array antenna to open the transmission or reception of this array element, and use the network to test the amplitude. and phase, just get a meter and input it back to the antenna, and some need to do it again to see if it is correct after calibration. After this is done, other controls of the phased array are purely software issues. Otherwise it might be messed up.

Edit:That means all array antennas have to be calibrated, right?

appoint:Not always. For example, the antennas of automotive radars are basically put out of the chip and directly into the PCB antenna, or directly into the antenna module of the millimeter wave type. The length, impedance, etc. of the microstrip inside are all designed, as long as SMT No problem, just do it right away. There are also those frequency swept antennas that look like arrays and don't need to be calibrated, and the kind of endless circle of a flat cylinder on the oldest warships, and the kind of circles on airports, which are also array antennas , but it is a machine scan and does not need matching calibration.

Edit:How many types of array antennas are there?

appoint:There are many ways to divide this thing. From my point of view, it is phase sweep, frequency sweep and machine sweep, and some are a combination. For example, the latest Starlink ground station is a combination of phase sweep and machine sweep.

Edit:We often hear phased arrays, and this machine scan can also understand. What is a frequency scan?

appoint:I haven't seen much of this, but I did test it. I don't know the design principle behind the antenna. Anyway, after it is designed, a passive antenna is like a stick. As the frequency changes, the direction of the main lobe also changes. For example, I am talking about a frequency, X The frequency band, 10-11G, 10G, the horizontal plane beam of this frequency point points 40 degrees to the left, at 10.5G, it points to 0 degrees, and at 11G, it becomes 40 degrees to the right. You can see that there is a sloping line when you look at the change of gain pointing angle and frequency. What's the use of it depends on the imagination.

Edit:How to choose these venues?

appoint:If it is a high-gain antenna with a very narrow beam, the planar near-field is preferred because the speed can be tolerated and the accuracy is high. But the plane near field is not suitable for low gain antenna. Generally, the selection of the site depends on the size of the antenna to be tested, the test frequency, and the test requirements to decide what kind of test site to choose.

Edit:So how are these near-field conversions to far-field calculated?

appoint:This is a bit complicated, like the plane near field is to intercept the limited data on a plane for plane spectrum expansion, plus probe compensation and finally re-FFT in the far field. If it is a spherical near field, it is to sample the vector data of a closed sphere and expand it and then map it, what is the Hankel function, Legendre transform a series of messy resynthesis FFT. Many papers on the Internet talk about this fast algorithm, and anyone with a certain microwave theoretical foundation can do it by themselves.

Edit:Going back to the venue we had just now, what are the characteristics of the tight venue?

appoint:I especially like to use the crunch field, the test speed is fast.

Edit:Besides the test speed?

appoint:It seems to be the same, but for me, to build a tight field venue, as long as it passes the quiet zone acceptance, the rest is the far field set, and the ready-made one can work. Very easy. Of course, the higher the frequency of the compression field, the better. The lower the frequency, the worse it gets. Our lowest crunch field can only go to 3G, and the phase is difficult to get lower, and the spikes are not easy to do, let alone curling.

Edit:How to verify the quiet zone of this tightening field?

appoint:Our approach is to set up a scanning gantry in the face of the compressed field, and then perform 3D scanning in XYZ. The ideal test field site is like this. Assuming that the longitudinal axis of the darkroom is defined as Z, then the quiet zone should be within the XY plane at any Z point. Equal amplitude and equal phase. The YZ plane at any X or the XZ plane at any Y has the same amplitude and phase distribution as a sine wave of frequency. This is an ideal site, but in fact the feed can be stable in phase center but must be unstable in amplitude. In the end, there will be a change in amplitude in the XY plane, which we call amplitude taper. Of course the smaller the better.

Edit:You keep talking about phases, what's the use of this?

appoint:Hi, to tell you the truth, I didn't know that it could measure phase and delay until three years after I started using the vector network analyzer, hehe.

Edit:Proper black history.

appoint:Yes, in the beginning, it was to measure the amplitude and gain gain. How can we know that we can measure the phase, but this is a learning process. The tasks given to us by leaders and customers are sometimes challenging, so we need to be here. In the process of continuous learning, you will grow from a young bird to an old bird. If you cannot continue to learn, you will be eliminated. That far-field formula is actually the inverse of the phase difference of the Eighth School. I don't know why the division of eights is used. I haven't looked into what the old gentlemen stipulated. As engineers, we are the first to achieve the target. I do not have the scientific attitude of doing research, so I have been mixing products in the industry, and I cannot publish papers.

Edit:Okay, thank you Mr. Ren for sharing, it took up too much time today, thank you very much.

appoint:Where, it is also a pleasure to communicate with you, and I am very grateful for this opportunity to share. thanks.

Editor: Huang Fei