As many commenters rightly found out, these devices are regenerative receivers. If you are interested in "real" radar sensors, I propose watching video #439
Your channel has one of the best audiences I have ever seen. So many comments are helpful and give new insights. Also, taking your time to reply to almost every comment is just wonderful. I am not even working with electronic, but I can feel this is a magical place.
Hi - as an ex-radar tech, I think @NickGrumpy and @dougcox835 nailed it. As a young ham operator in the 1960's I built super-regenerative receivers using single vacuum tubes or single transistors. They are simple, crude, and sensitive, and I think these so-called radar units are just modern "rush boxes", as we used to call these circuits. RADAR is actually RAdio Detection And Ranging, and will detect stationary objects and report their range. These new units only detect movement - any disruption of their local microwave field gets "detected" by regen transistor, and interpreted by the chip as output. They remind me of theremins, if you remember what those musical instruments are. Not actually radar, but pretty darned clever!
You are right. As said, this is an old video and in the meantime, a few viewers suggested this. It is good because if it is a receiver, it is prolly more legal ;-) BTW: I once made a video about a tube Regen receiver…
Hi! I think the evolution of radar sensors has skyrocked. Now you can get devices like HLK LD2410 with 3.3v logic, sensing movement AND precense. Programable via serial. You can program sensitivity at diferent distances. And still only about 5usd. Hlk even makes radars that give you direction of motion (x and y) of multiple people. I think a updated new video would be great!
@@ethzero is that the Bluetooth version or via serial? I have without Bluetooth, and used serial to configure. In use with only the high/low signal for detected or not. But can be used in so many ways :-)
Some time ago I built a sensor to switch security cameras if someone entered a closed area outside, rather than upgrading the DVR to add more channels. Initially it worked well' but after a few weeks the camera started randomly switching, and unusually it was also much worse at night. I replaced the module (housed in a small plastic enclosure under an eave) but it had no effect, the camera still switched randomly at night, sometimes non stop. In the workshop however both sensors behaved perfectly and after a lot of power supply filters, emi/rfi tests and general head scratching, finally one night I found the culprit. A small spider would emerge and build a web from the sensor box to the camera bracket alongside, even though the spider would sit still, the slightest breeze would shake the web and I could hear my little video relay click. After removing the offending spider the sensor became stable again, but I was intrigued that something so small could trigger these devices. A lesson learned for outdoor use there. The camera has I/R emitters that the spiders find appealing for bug attraction, and the radar sensor finds the spiders appealing...
I understand your frustration when troubleshooting this issue. Faulty Dupont wires and cheap breadboards have almost made me crazy when building and testing new projects. Based on your experience can you recommend a good radar sensor? I need a couple of them to detect birds and trigger an ESP32-Cam module.
@@oladunk9986 The versions I was using were pretty cheap and I believe the same as the XYC-WB-G1 as the switching time was the same, although you probably want a five second or less on time if you just want triggering, it will give you more captures per minute. In that case the RCWL-0516 or HW-MS03 would be good candidates, although the more expensive HFS-DC06 is likely to be more stable with less false triggers. The best part in detection with these is the daytime performance matches night, unlike PIR, and the sensor can be sealed from the elements. The downside is they can be triggered by wet foliage in the wind, insects and sometimes just nothing (may actually be flying insects...), but they are cheap. Good luck.
"This company has ONE boss, and that's me! And I tell you, if a fly as much as farts, I want to know about it. Now go engineer or whatever it is you are doing."
The glass is most likely metallised glass, to avoid that too much heat (as light) can go through them. They are known to nearly completely block Wi-Fi signals, so it wouldn't be surprising if radar wouldn't work as well.
Keep in mind that the rain radars work on ~20 GHz. Given the distance over which they are expected to work, the receivers in these are very sensitive, and 'rogue' sensors emitting a signal at those same frequencies may well cause trouble in the rain radars.
Rain radars (actually called weather radars) operate in the 5GHz wifi band. Wifi routers that operate in the range 5250-5350 and 5470-5725 MHz must use a mechanism called DFS or dynamic frequency selection to change channel if they detect a weather radar. DFS is not required in the 5150-5250 MHz band and in the 5725-5875 MHz band. Some manufacturers allows the user to disable DFS, which is is illegal and causes weather radar interference. Problems usually occur with WIFI access points installed outside, on top of a hill, or in a tall buildings.
@@atmel9077 The frequency of the rain radars depends on the intended range, not all operate around 5 GHz. Higher frequencies have a shorter range, but better sensitivity to smaller droplets. For vertical profiling, around 90-100 GHz is used, short range, but much better sensitivity for smaller droplets. But indeed, there are several (proposed) license-free channels that are of concern for observations used for weather, including passive microwave for water vapour measurements. Those can't be put on a different frequency to avoid interference.
@@MrGeekGamer The transmitter of the rain radar is not the issue. The reflected signal from a collection of raindrops is small, and drops off quickly with distance. (1/r² for the transmitted beam, another 1/r² for the reflected signal, with a strong dependence of the reflected signal on the droplet size). Even low power transmitters can easily overwhelm the reflected signal, especially for small droplets.
Thanks for the re-run. Regenerative or real radar, they are still very useful and interesting devices. This was well worth a second viewing. Have a great Summer break!
Large civil aircraft radars work in pulses and for multiple reasons they change the frequency slightly for each pulse around the base frequency and the receiver and the logic behind that know, what frequencies are to be expected in a given timeframe. This could be the reason why multiple of the devices work in the same room: because the chance of them being in sync with the exact frequency is really slim.
To understand how one transistor can perform 3 functions look at the operation of a Super-regenerative receiver. it oscillates and radiates at the TX frequency and mixes this with the reflect frequency if both frequencies are the same there will be 0 mixer output, if the reflected frequency is shifted by a moving target then it mixes with the TX frequency and the mixer outputs the difference frequency which will be a very low frequency that can be used as an output showing movement.
Don't worry. I have indeed nothing better to do than to watch your videos. Also, not being Swiss I can't afford a holiday in Switzerland. On my budget the closest I can get to Switzerland is buying a pocket knife. Thanks for your invaluable advice on radar sensors. Enjoy the summer break.
A suggestion for a video after the summer break: a second walkthrough of your solar setup, pro's and cons of your setup, would you still choose the same components today. What about the Performance of the panels and the battery.
I'm using such a sensor in the basement to switch on the mains powered ceiling light. It already detects me when I get close to the door (I know, because of the cat's door letting some light through). Works without issues for a few years now.
In my opinion,The HW-MS03 out of those you have is the best! Viewers should also look at HiLink HLK-LD2410C-P 5v 24GHz. What I like about the HiLink sensor, there is an app that connects to the sensor via Bluetooth to change the parameters of the sensor and detects micro-movements. The sensor will still see you if you sit still. Breathing is enough movement for the sensor detection.
I have an RCWL0516 board that turns on a light for illumination when it senses motion. It works reliably from anyplace in my two car garage. Good product!
Thank you for reminding me. I bought a couple of these (likely after watching this video the first time) and still have a couple sitting around. That'll never do...
Hiya, I think the "radar" defectors that use a single transistor are actually oscillators that change frequency due to external loads. It's loosely coupled so any human presence with add capacitance affecting the frequency. These have been around for many years and were used on some burglar alarms in the 70's. Operating frequency was sub 900MHz if I remember. I think it's related to a super-regenerative receiver.
Have a good break. Just a thought, if you are planning on doing intros for all your 'holiday' videos, could you also include the date that the archive video has come from so that we have an idea as to how far things may have moved on.
I strongly suggest that you mount the detector on a PCB earth plane, with a brass shim wall around it. This makes the device sensitive in the forward direction, screening other directions. . By soldering SIL rolled pin ic sockets to the interface and compatible pins to the carrier board (or the reverse), it is possible to make the modules pluggable onto the carrier board. By using a double sided copper carrier PCB, with a solid copper ground plane it is possible to build the screen on one side of the board, while adding the processor circuit to the other. I have played with the standoff distance between the carrier and the module, to check it does not change the sensitivity. Surprisingly, I did not see much change in sensitivity. I suspect there is an optimum stand off distance, related to creating a standing wave with the carrier board earth plane, but the microwave frequency does not appear to be that well defined.
@@AndreasSpiess the module I used appeared to have a earth plane, but was still sensitive in the reverse direction. If the frequency used is 24GHz, then one might expect an effective stand off distance from the carrier board earth plane to be 1/4 of a wavelength, approximately 0.3115 cm, as this would produce a voltage node at the surface of the earth plane. The next standing wave would be at 3/4 of a wavelength. This is the standoff distance that can be achieved using the roll pin plug and socket as the interface. In practice, I doubt if the frequency is that well controlled, and the precision of the standoff distance is not that accurate. My test equipment does not reach 24GHz,, I resorted to looking for changes in the supply current, reasoning that this might show loading of the Tx. There were small changes, but nothing conclusive. The range did not appear to be adversely affected by the earth plane of the carrier board, but it did greatly increase the directionality of the beam.
I re-used a radar Motion Sensor that was used for a automatic Door of a supermarket. To test it's function, i placed it in my Living Space, while it was directed to my Bathroom. I opened two closet doors and closed the bathroom door as obstacles. While i was sitting on the floor in the bathroom, all I had to do was move my index finger back and forth to trigger the radar sensor. I was about 5m away. The bathroom door is 4cm thick and the two closet doors are 2cm each. Since then I have been using the radar sensor to switch on the staircase lighting. It serves as a reliable replacement for the last PIR motion detector that loves to switch on randomly without logical reasons. It is used together with a self-made NE555 off-delay circuit and my self-made LED lighting (12VDC) in the form of small cable ducts with 5mm LEDs. My Radar Sensor can be used with 12-30VDC, or 12-24VAC. The required field widths are set by adjusting the lateral funnel walls of the horn antenna. It's a BBC Bircher RK 31C-P/31.
Hi Andreas, I love your reviews. Thanks for all your good work. The internet is full of IR Blasters for a/c that are not of any use. Because the makers focus on the connectabillity and not on the possibility to make the a/c smarter. .We need a smart guy like you, that provides the code for a self learning device for all the buttons of the a/c remote device, than we need the variables clearly mentioned so we can add the code to make it smarter and add sensors an use arduino/esp script to set the variables through set and if statements based on the sensor readings. And set the parameters through wifi interface via on the wireless phone. I thought that would be a wonderful project for you. Greetings, Lisa
I use one of these little sensors in the house and found that with the help of some adhesive aluminium tape, the area that they respond to can be easily modified.
I recently bought an Acconeer A121 devkit, can be mounted to a raspberry pi and it can deliver mm accuracy up to 20 meters in absolute terms, or micrometer accuracy in relative terms. The demo apps from acconeer include presence detection, breathing rate monitors, water level detections and distance measurements. Vibration of objects can also be measured. I tested these myself and found it to be incredibly effective, all on a sensor of 64mm2 !!!
Most radar sensors that are legal to use in Europe seem to run at 24GHz nowadays but in the past, they also used 2.4 GHz. I have for example a Siemens KMY-24 that works in the 2.4 GHz range.
Great video and testing. It really is amazing these simple cheap sensors work so well. Perhaps GreatScott or another engineer can explain how these work? I suspect all but the most expensive one are all based on a reference design that did NOT originate in China.
Can these be programmed to tell you how far away the detected object is? Either the standalone units, or when hooked up to a microcontroller of some sort?
Hi Andreas. I had problems using the rcwl0516 module with a ESP32. that sensor is sensitive to wifi signal. Even moving the sensor half a meter away from the ESP32 module causes false positives in the sensor. Have you ever encountered this problem? Thank you!
@guatagel2454 My experience with these cheap radar sensors: They work best with an own power supply using a linear voltage regulator + caps. Connecting to the supply of a Wemos did not work for me.
There are some posts on the internet about this issue and some suggestions to resolve it. Another possibility could be to use a more modern LD2410 sensor (video #439)
Thank you very much for your wonderful video, I really get the deep.impress of it. Can you test multiple ld241p radat please? I really have curious of it. Thank you so much
By the way. As I have understood, the radars in WW2 had a very simple schematic. Magnetron + tube as a on off switch + a monostable multivibrator + a linear line cathode ray tube as the display.
Back in the day, most electronic devices were quite simple designs, I think. They just did not have the capabilities for better equipment. However, they worked!
@@AndreasSpiess Perhaps one of the inputs will allow conclusions to be drawn about how it works. For sure PCB does not look like 24GHz 🙂It even does not look like 5.8GHz but I will check this using my SDR but that antenna length also indicates lower frequencies. Maybe a super fast transistor is able to generate harmonics. My guess is, this is just called "radar" but principle is something completely different. The human body has impact on resonance or may be even SWR.
@@greekstraycats You will see a signal in the GHz range in the air. As many comments suggest, these devices are regenerative receivers, and they emit a small signal via the antenna.
An LDO has a minimum dropout voltage, if it’s being used to generate 3.3V for the controller chip maybe you’d be better off connecting 3.3V to both the Vin and 3.3Vout pins.
I do not see why I should connect anything to the input if I connect power to the output. As you say, there should be a difference between the two pins. So the input will be cut off by the LDO.
Andreas, in your other radar video #439 you are stating at 4:01 that the cheap sensors of this video interfere with the 2.4 Ghz wifi frequency... But in this (summer break) radar video you are telling us that it does not interfere with the 2.4 Ghz frequency. So, what is the correct statement? Enjoy the summer Andreas!
Would they encode the radio pulse with a binary code and only accept returns with the same code? If that's the case that is pretty amazing for a device that is under $10
Hey Andreas, love your videos, I have learnt so much from you! Thank you. I wonder if you can help me, I'm trying to find a way of sharing the data connection (4g) from my mobile phone over bluetooth to my esp32 project so that the esp32 can send and receive data from the internet using the Arduino IDE. I have searched for months but not found a way of doing this, are you able to help with this?
@@AndreasSpiess I do already have wifi working but for my application a bluetooth internet connection would be so much better for battery life and end user experience. thank you anyway
@@AndreasSpiess I'm on solar 12v input. Can usb to 5v. This device puts out 3v so that's my issue. In not electronics educated, I generally watch utube and do understand. But I'm a novice. So some doorbells run of 3v. But probably more amps again. So all I'm trying to do is step up v and a. The 5seconds is ok, for an alarm. It's some 100m away from me, the setup. So audio alarm is good. And if there's continuous movement it will trigger repeatedly. So that's ok too. But ideally I want to trigger a 12v output. So I've seen there are 5v relay boards out there. Just wondered if you have the patience to demonstrate something along those lines. Many thanks.
@@earthlingthings Your questions are far too big for a RUclips comment. Maybe you search for a device that runs on 10 volts and is capable of switching at least 12-volt loads. Or you start with simple things like a voltage regulator to convert 12V to 3.3V and so on and learn all the stuff needed to build such a device.
@@AndreasSpiess Mate, you've the hardware there, how about giving it a crack and report the findings? If you find an easy/cheap option to do so, you might have a best seller video. I'm sure it find an wider interest than more than just our regular electronics cohort.
Would be interesting to try tapping off the IF signal to send to a microcontroller and do more advanced analysis on it. There's an opamp in the chip which acts as an IF amplifier.
Getting speed information is trivial, just do an FFT to extract the beat frequency. Things like distance measurement is a different story and it's better to use radar sensors dedicated to that.
Not sure if presented, but instead of having these devices interfere on my WiFi, I've had (my) WiFi interfere on these modules. Using an ESP8266 jogged a lot of fake readings from for instance the RCWL-0516 module. I've come to realise that although they are cheap, they cannot fulfill their role in a crowded 2.4Ghz spectrum. Switching to devices that make use of other, higher, frequency bands is more reliable as far as I have experienced at the moment. Also, the lack of directivity from these cheap modules make them trigger on people not in the same room as well, through hollow walls and in some cases even through masonry bricks. That's the issue with not getting an analog strength signal but only a binary on/off signal.
I'm almost sure that one of the pins on the chip on the cheapest sensor can be tapped for more information. It was a while ago and I didn't finish playing with it, so I'm fuzzy about just what the extra information is and that I wasn't just fooling myself, and I don't remember which pin. Still you might want to mess around to see what you can find. I think it was the beat frequency between TX and RX so you could get a speed measurement. If I remember correctly, the chip is actually the same or similar to one used on inexpensive PIR sensors. I think it has one or two op amps built in to the chip and the signal of interest was found on the output of one or both. I think I cut the trace to the voltage out pin so I could use it for the signal..
The IC has nothing to do with the RF. It is only a logic chip. They’re is Alyson no TX and RX because it only has one transistor. It seems to be a regenerative receiver (once famous in the valve age).
@@AndreasSpiess Presumably the IC isn't there just for decoration? Somehow it must be in touch with the RF section to know what is going on so it can do useful things like turn on and off. I'm not a radiohead, so technically RX and TX may be the wrong terms. My intended meaning though is a mixture of the transmitted signal and the reflected signal have been combined to make a much lower frequency beat signal. I believe that goes into opamps integrated into the IC for amplification and that is where I tapped in. How all that happens is black witchcraft as far as my understanding goes but I believe I'm at least approximately right about what shows up after amplification. One way I wonder if I could have been tricked by the witch is perhaps I was seeing into the middle of some magic the IC does to do what it does and it just happened to look a hell of a lot like it responding in an analog like way to speed or distance. For instance, maybe if what it is working with from the RF section isn't strong enough, it tries again sooner and sooner the weaker it is. received I'm in the middle of a project at the moment but I'll try to make time tonight to repeat what I did and get back to you with better details.
Antenna beam width is determined by the aperture. Larger aperture, narrower beam. These things are tiny so are a fraction of the wavelength and essentially omnidirectional. I am also baffled about how it works with only one transistor. I think the magic is in the antenna shape which is both the resonator and receiver/transmitter. One think about radar that really helps is that the transmit frequency is known so mixing with the RX is simple. I think in this case the entire room is part of the resonant field so changes in the environment change the frequency of oscillation. How it self mixes must be tricky. I think the S shape might have something to do with that. Perhaps the opposing elements create a phase difference and it's that small difference that gets passed on to the chip. Normally radar is pulsed to determine range but this is cw so all you are interested in is doppler return. I'm just speculating about this though.
Same transistor used for transmitter and receiver is called regenerative receiver.The transmitter is disturbed by an external object or signal and create a difference which can be demodulated. Very old way of building radios for pagers and remote control for boats and airplanes. Also built with tubes before WW2.
if your windows are 'modern', they probably have a metallic coating the reflects IR, but also, natürlich, RF... I wonder if these things are mostly UWB, not 'tuned', narrow-band RF...
I agree with your statement about Glas. Our windows are simple and old… As many others stated, these devices are regenerative receivers, not real radars. But they have a similar effect for a ver low price.
@@AndreasSpiess BTW what do you think is the best solution today to use lipos with microcontrollers? There is a lot of noise out there but we need some signal here. We trust you :)
Glass is no problem, as long as it has no metal. I use a 516 with a esp, for over a year now, with no problem. Through a wall. I use it to check for movement on my driveway.
The PWB antennas are not directional (as the tests demonstrated), except for the one which has a groundplane/cover mounted behind it. You could make them directional by mounting them in front of a metallic groundplane or corner reflector. You could even mount them at the focal point of a paraboloidal reflector. Make sure the mounting is rigid; any relative motion of the sensor and structure will trigger it. You're essentially making this the feed for an antenna of your own; ideally you would determine the operating frequency and design the antenna for the pattern you want. But the analysis is fairly involved; if you're not too particular, you can get significant directivity from ad-hoc approaches.
As many commenters rightly found out, these devices are regenerative receivers. If you are interested in "real" radar sensors, I propose watching video #439
Your channel has one of the best audiences I have ever seen. So many comments are helpful and give new insights. Also, taking your time to reply to almost every comment is just wonderful. I am not even working with electronic, but I can feel this is a magical place.
Video link for those interested - ruclips.net/video/dAzHXpP3FcI/видео.html
Cheers, enjoy the break, Andreas
Thank you, Peter!
Hi - as an ex-radar tech, I think @NickGrumpy and @dougcox835 nailed it. As a young ham operator in the 1960's I built super-regenerative receivers using single vacuum tubes or single transistors. They are simple, crude, and sensitive, and I think these so-called radar units are just modern "rush boxes", as we used to call these circuits. RADAR is actually RAdio Detection And Ranging, and will detect stationary objects and report their range. These new units only detect movement - any disruption of their local microwave field gets "detected" by regen transistor, and interpreted by the chip as output. They remind me of theremins, if you remember what those musical instruments are. Not actually radar, but pretty darned clever!
You are right. As said, this is an old video and in the meantime, a few viewers suggested this. It is good because if it is a receiver, it is prolly more legal ;-)
BTW: I once made a video about a tube Regen receiver…
Hi! I think the evolution of radar sensors has skyrocked. Now you can get devices like HLK LD2410 with 3.3v logic, sensing movement AND precense. Programable via serial. You can program sensitivity at diferent distances. And still only about 5usd.
Hlk even makes radars that give you direction of motion (x and y) of multiple people.
I think a updated new video would be great!
The LD2410 is also supported in ESPHome that exposes it's sensing data to Home Assistant.
@@ethzero is that the Bluetooth version or via serial?
I have without Bluetooth, and used serial to configure. In use with only the high/low signal for detected or not. But can be used in so many ways :-)
I think LD2410 needs 5V
@@saninnsalas it uses 5V Power Supply, but have 3.3V logic. I think it has its own 3.3V reg onboard for stability.
You are right. I covered the LD2410 in a later video.
Some time ago I built a sensor to switch security cameras if someone entered a closed area outside, rather than upgrading the DVR to add more channels. Initially it worked well'
but after a few weeks the camera started randomly switching, and unusually it was also much worse at night. I replaced the module (housed in a small plastic enclosure under an eave)
but it had no effect, the camera still switched randomly at night, sometimes non stop. In the workshop however both sensors behaved perfectly and after a lot of power supply filters, emi/rfi tests and general head scratching, finally one night I found the culprit.
A small spider would emerge and build a web from the sensor box to the camera bracket alongside, even though the spider would sit still, the slightest breeze would shake the web and I could hear my little video relay click.
After removing the offending spider the sensor became stable again, but I was intrigued that something so small could trigger these devices. A lesson learned for outdoor use there. The camera has I/R emitters that the spiders find appealing for bug attraction, and the radar sensor finds the spiders appealing...
I understand your frustration when troubleshooting this issue. Faulty Dupont wires and cheap breadboards have almost made me crazy when building and testing new projects. Based on your experience can you recommend a good radar sensor? I need a couple of them to detect birds and trigger an ESP32-Cam module.
@@oladunk9986 The versions I was using were pretty cheap and I believe the same as the XYC-WB-G1 as the switching time was the same, although you probably want a five second or less on time if you just want triggering, it will give you more captures per minute. In that case the RCWL-0516 or HW-MS03 would be good candidates, although the more expensive HFS-DC06 is likely to be more stable with less false triggers. The best part in detection with these is the daytime performance matches night, unlike PIR, and the sensor can be sealed from the elements. The downside is they can be triggered by wet foliage in the wind, insects and sometimes just nothing (may actually be flying insects...), but they are cheap. Good luck.
Wonderful story; and although spiders are not classified as bugs, this is definitely....real debugging.
Awesome engineering have done. Simply wow!
"This company has ONE boss, and that's me! And I tell you, if a fly as much as farts, I want to know about it. Now go engineer or whatever it is you are doing."
The glass is most likely metallised glass, to avoid that too much heat (as light) can go through them. They are known to nearly completely block Wi-Fi signals, so it wouldn't be surprising if radar wouldn't work as well.
I do not know, but our home is quite old and in Switzerland, they only recently started with this metallized glass.
Keep in mind that the rain radars work on ~20 GHz. Given the distance over which they are expected to work, the receivers in these are very sensitive, and 'rogue' sensors emitting a signal at those same frequencies may well cause trouble in the rain radars.
These devices are not real radars, they are regenerative receivers (mentioned in other comments). Do there is no worry about interferences.
I don't anything this size will emit anything near the power required to drown out such a high powered radar.
Rain radars (actually called weather radars) operate in the 5GHz wifi band. Wifi routers that operate in the range 5250-5350 and 5470-5725 MHz must use a mechanism called DFS or dynamic frequency selection to change channel if they detect a weather radar. DFS is not required in the 5150-5250 MHz band and in the 5725-5875 MHz band.
Some manufacturers allows the user to disable DFS, which is is illegal and causes weather radar interference.
Problems usually occur with WIFI access points installed outside, on top of a hill, or in a tall buildings.
@@atmel9077 The frequency of the rain radars depends on the intended range, not all operate around 5 GHz. Higher frequencies have a shorter range, but better sensitivity to smaller droplets. For vertical profiling, around 90-100 GHz is used, short range, but much better sensitivity for smaller droplets.
But indeed, there are several (proposed) license-free channels that are of concern for observations used for weather, including passive microwave for water vapour measurements. Those can't be put on a different frequency to avoid interference.
@@MrGeekGamer The transmitter of the rain radar is not the issue. The reflected signal from a collection of raindrops is small, and drops off quickly with distance. (1/r² for the transmitted beam, another 1/r² for the reflected signal, with a strong dependence of the reflected signal on the droplet size). Even low power transmitters can easily overwhelm the reflected signal, especially for small droplets.
Thanks for the re-run. Regenerative or real radar, they are still very useful and interesting devices. This was well worth a second viewing. Have a great Summer break!
Indeed, for me, it is still unbelievable how sensitive they are!
Large civil aircraft radars work in pulses and for multiple reasons they change the frequency slightly for each pulse around the base frequency and the receiver and the logic behind that know, what frequencies are to be expected in a given timeframe. This could be the reason why multiple of the devices work in the same room: because the chance of them being in sync with the exact frequency is really slim.
These sensors work very different to the „real“ radars you describe. These sensors only have one transistor which has to do all functions…
Have a great break & thanks for all your efforts to bring us all unique content 😊
Thank you for your kind words!
To understand how one transistor can perform 3 functions look at the operation of a Super-regenerative receiver. it oscillates and radiates at the TX frequency and mixes this with the reflect frequency if both frequencies are the same there will be 0 mixer output, if the reflected frequency is shifted by a moving target then it mixes with the TX frequency and the mixer outputs the difference frequency which will be a very low frequency that can be used as an output showing movement.
You are right, they are regenerative receivers!
Andreas, nature is blessing, we want you fresh ans clear after the deserved rest!!! Cheers!!!!
Thank you!
Just in time, I was about to buy some radar sensors to play with, now I can know which ones better, thanks!!
Maybe you have also a look at the LD2410 (I also made a video about them)
Don't worry. I have indeed nothing better to do than to watch your videos.
Also, not being Swiss I can't afford a holiday in Switzerland. On my budget the closest I can get to Switzerland is buying a pocket knife.
Thanks for your invaluable advice on radar sensors. Enjoy the summer break.
@jacquesmertens3369 Switzerland is very versatile: Some Swiss chocolates or candies might be even cheaper 🙂
Indeed, Switzerland is expensive. But I experienced that other "holiday" countries in Europe catch up quickly :-(
If you unsolder the PCBs on the HFS-DC06 you can see the RF 'black magic' used in the trace design. It's positively astonishing.
I agree.
A suggestion for a video after the summer break: a second walkthrough of your solar setup, pro's and cons of your setup, would you still choose the same components today. What about the Performance of the panels and the battery.
The solar videos were not too successful. So I was not sure if I wanted to du an update. But maybe I will do it now…
Your videos are always interesting.🙂🎩
Thank you!
Great video as always.
Thank you!
I'm using such a sensor in the basement to switch on the mains powered ceiling light. It already detects me when I get close to the door (I know, because of the cat's door letting some light through). Works without issues for a few years now.
Thank you for sharing your experience!
In my opinion,The HW-MS03 out of those you have is the best! Viewers should also look at HiLink HLK-LD2410C-P 5v 24GHz. What I like about the HiLink sensor, there is an app that connects to the sensor via Bluetooth to change the parameters of the sensor and detects micro-movements. The sensor will still see you if you sit still. Breathing is enough movement for the sensor detection.
I agree that these sensors are. Good choice. I already made a video about them…
Enjoy your break and have a creative summer
I will! Thank you.
RCWL 0516 works on 3.3V. Mine has a dedicated 3.3V pin and I use several of them hooked to ESP32 3.3v Pin.
Thank you for the info!
Best wishes & good to meet you In Friedrichshafen. Rgds, Steve
Thank you! Was my pleasure, too.
I have an RCWL0516 board that turns on a light for illumination when it senses motion.
It works reliably from anyplace in my two car garage.
Good product!
Thank you for sharing your experience!
Thank you for reminding me. I bought a couple of these (likely after watching this video the first time) and still have a couple sitting around. That'll never do...
Maybe you now find a project for them ;-)
❤ Tks again and enjoy summer!
I will. Thanks!
I have a few of the RCWL-0516 sensors and modified the time delay by adding a capacitor across C-TM and changed the delay to about 30 sec
Thank you for sharing this enhancement!
Hiya,
I think the "radar" defectors that use a single transistor are actually oscillators that change frequency due to external loads. It's loosely coupled so any human presence with add capacitance affecting the frequency. These have been around for many years and were used on some burglar alarms in the 70's. Operating frequency was sub 900MHz if I remember. I think it's related to a super-regenerative receiver.
Just saw your pinned post on regenerative receivers.
Indeed, when I made the video, it was not clear. Only after reading some comments, it became clear that these are "only" receivers...
it looks lovely over there hope your having a fantastic time
I will try ;-)
I’ve played with a couple these. Cheap and they work amazingly well.
Thanks for sharing your experience!
Check out the LD1125H. I'm getting motion detection 8-10 meters away through two layers of sheetrock. DigiblurDIY has a great video about it.
Thanks for the info!
Have a good break.
Just a thought, if you are planning on doing intros for all your 'holiday' videos, could you also include the date that the archive video has come from so that we have an idea as to how far things may have moved on.
I do Not plan intros for the remaining videos, but maybe add the original number in the description.
I strongly suggest that you mount the detector on a PCB earth plane, with a brass shim wall around it. This makes the device sensitive in the forward direction, screening other directions. . By soldering SIL rolled pin ic sockets to the interface and compatible pins to the carrier board (or the reverse), it is possible to make the modules pluggable onto the carrier board. By using a double sided copper carrier PCB, with a solid copper ground plane it is possible to build the screen on one side of the board, while adding the processor circuit to the other.
I have played with the standoff distance between the carrier and the module, to check it does not change the sensitivity. Surprisingly, I did not see much change in sensitivity. I suspect there is an optimum stand off distance, related to creating a standing wave with the carrier board earth plane, but the microwave frequency does not appear to be that well defined.
Thank you for sharing your learnings. Interesting!
@@AndreasSpiess the module I used appeared to have a earth plane, but was still sensitive in the reverse direction. If the frequency used is 24GHz, then one might expect an effective stand off distance from the carrier board earth plane to be 1/4 of a wavelength, approximately 0.3115 cm, as this would produce a voltage node at the surface of the earth plane. The next standing wave would be at 3/4 of a wavelength.
This is the standoff distance that can be achieved using the roll pin plug and socket as the interface.
In practice, I doubt if the frequency is that well controlled, and the precision of the standoff distance is not that accurate. My test equipment does not reach 24GHz,, I resorted to looking for changes in the supply current, reasoning that this might show loading of the Tx. There were small changes, but nothing conclusive. The range did not appear to be adversely affected by the earth plane of the carrier board, but it did greatly increase the directionality of the beam.
I would be checking those cheaper ones for RF interference, the shielding on the expensive one is there for a reason.
I agree
I re-used a radar Motion Sensor that was used for a automatic Door of a supermarket. To test it's function, i placed it in my Living Space, while it was directed to my Bathroom. I opened two closet doors and closed the bathroom door as obstacles. While i was sitting on the floor in the bathroom, all I had to do was move my index finger back and forth to trigger the radar sensor. I was about 5m away. The bathroom door is 4cm thick and the two closet doors are 2cm each.
Since then I have been using the radar sensor to switch on the staircase lighting. It serves as a reliable replacement for the last PIR motion detector that loves to switch on randomly without logical reasons.
It is used together with a self-made NE555 off-delay circuit and my self-made LED lighting (12VDC) in the form of small cable ducts with 5mm LEDs.
My Radar Sensor can be used with 12-30VDC, or 12-24VAC. The required field widths are set by adjusting the lateral funnel walls of the horn antenna. It's a BBC Bircher RK 31C-P/31.
I am also puzzled by how sensitive these cheap devices are!
That blue sports car was a Renault Alpine A110
You are right!
Hi Andreas, I love your reviews. Thanks for all your good work. The internet is full of IR Blasters for a/c that are not of any use. Because the makers focus on the connectabillity and not on the possibility to make the a/c smarter. .We need a smart guy like you, that provides the code for a self learning device for all the buttons of the a/c remote device, than we need the variables clearly mentioned so we can add the code to make it smarter and add sensors an use arduino/esp script to set the variables through set and if statements based on the sensor readings. And set the parameters through wifi interface via on the wireless phone. I thought that would be a wonderful project for you. Greetings, Lisa
I once made a video on how to hack IR devices. But it was not as easy as you describe ;-)
Results from measurements:
XYC-WB-D1 - 2.906 GHz
HFS-DC06H - 5.721 GHz
RCWL-0516 - ca. 3 GHz
HW-MS03 - ca. 3 GHz
SK-600 - 5.739 GHz
Thanks for sharing! So the HFS-DC06 and the SK-600 are in the 5GHz Wi-Fi band!
Felices vacaciones!!!!🎉🏞️🏝️
Gracias!
No, I can't 'like' at the end because I liked at the beginning! 😄
Enjoy the summer! 73
I will! Thank you.
I use one of these little sensors in the house and found that with the help of some adhesive aluminium tape, the area that they respond to can be easily modified.
Cool. Thank you for sharing your trick!
Viel Spaß und Erholung im Urlaub, gönn dir Andreas :) hoffe das ich dann einige Projekte nachbauen kann ;)
Vielen Dank! Der Urlaub wird hauptsächlich „ums Haus“ stattfinden…
I recently bought an Acconeer A121 devkit, can be mounted to a raspberry pi and it can deliver mm accuracy up to 20 meters in absolute terms, or micrometer accuracy in relative terms. The demo apps from acconeer include presence detection, breathing rate monitors, water level detections and distance measurements. Vibration of objects can also be measured. I tested these myself and found it to be incredibly effective, all on a sensor of 64mm2 !!!
Thank you for sharing your experience. This sensor seems to be very good. However, it is a bit pricey if you do not need its particular features...
Enjoy the summer Andreas!
Thank you!
Nice,
Thanks for researching this,
Hope to pick one myself.
They should be widely available.
Enjoy your summer break Andreas!
I will. Thank you!
Most radar sensors that are legal to use in Europe seem to run at 24GHz nowadays but in the past, they also used 2.4 GHz. I have for example a Siemens KMY-24 that works in the 2.4 GHz range.
The ones shown here vary a lot in frequency. They can be classified as receivers an so are legal even if they transmit a bit.
@@AndreasSpiess Okay, understand. So the principle is at bit different than with dopper radar sensors like RSM-1650 for example..
@@EinzigfreierName It is completely different. They even cannot be named "radars". But I did not know that when I made the video...
Thanks and have a nice holiday and safe cycling.
Thanks, you too!
@@AndreasSpiess Thanks!
It great to go back and see some of your old videos thanks brother 🧑 de 2e0ree 73
:-) 73 de HB9BLA
Genieß deine Sommerpause :-)
Mach ich ;-)
Great video and testing. It really is amazing these simple cheap sensors work so well. Perhaps GreatScott or another engineer can explain how these work? I suspect all but the most expensive one are all based on a reference design that did NOT originate in China.
I pinned a comment on how they work. They are simple receivers like back in the old days...
Hi. Great video and usefull components that i think I will use in the future. Have you ever thought of revisiting the car radar speed detector again?
Another video on radars is planned. But I still did not find a cheap radar capable of detecting on a long-range :-(
Can these be programmed to tell you how far away the detected object is? Either the standalone units, or when hooked up to a microcontroller of some sort?
No. Maybe you have a look at the video about the LD2410?
Enjoy the summer holiday ! ❤
Thank you! You too!
Hi Andreas. I had problems using the rcwl0516 module with a ESP32. that sensor is sensitive to wifi signal. Even moving the sensor half a meter away from the ESP32 module causes false positives in the sensor. Have you ever encountered this problem? Thank you!
@guatagel2454 My experience with these cheap radar sensors: They work best with an own power supply using a linear voltage regulator + caps. Connecting to the supply of a Wemos did not work for me.
I also had false positives. I found adding a capacitor to the power rails and a resistor on the Vcc line to create a simple rc filter fixed it
@@fluteplayerify congrats! I my specific case RC did not help, But in both cases an improvement of the supply path was the key.
There are some posts on the internet about this issue and some suggestions to resolve it. Another possibility could be to use a more modern LD2410 sensor (video #439)
Thank you very much for your wonderful video, I really get the deep.impress of it.
Can you test multiple ld241p radat please? I really have curious of it. Thank you so much
I made a video about the ld2410
Are these modules from the Automobile area. My VW Golf has a front collision detector, when on detects cars etc when 20m in front.???
No. Cars have different sensors (more sophisicated)
Have a great summer break!
Thank you! You to.
Nice Radar 😃
Schönen Urlaub ✌️
Danke!
hi Andreas! Do you know if this devices are secure for install on top of gasoil deposits?
Thanks for your inspiring videos!
I do not know the laws around this topic :-(
By the way. As I have understood, the radars in WW2 had a very simple schematic. Magnetron + tube as a on off switch + a monostable multivibrator + a linear line cathode ray tube as the display.
Back in the day, most electronic devices were quite simple designs, I think. They just did not have the capabilities for better equipment. However, they worked!
Maybe your window has a metalic coating that block the radar signal? Modern window panes often have that...
Maybe. But our windows are quite old...
It would be cool to get the raw signal from PCB to calculate distance speed etc
I agree. But these sensors are not made for that.
Have a wonderful holiday Andreas😊
Thank you! 😃
Interesting topic. I will try to order some of these devices and measure them with the use of SA up to 26 GHz at work. I will let you know.
Please do! They seem to work on different frequencies.
No interesting signals to measure at the pins of that IC?
Also bought one of those some time ago but did not play with it yet.
The IC is well known it contains nothing special (no RF function)
@@AndreasSpiess Perhaps one of the inputs will allow conclusions to be drawn about how it works. For sure PCB does not look like 24GHz 🙂It even does not look like 5.8GHz but I will check this using my SDR but that antenna length also indicates lower frequencies.
Maybe a super fast transistor is able to generate harmonics. My guess is, this is just called "radar" but principle is something completely different. The human body has impact on resonance or may be even SWR.
@@greekstraycats You will see a signal in the GHz range in the air. As many comments suggest, these devices are regenerative receivers, and they emit a small signal via the antenna.
Have a great vacation, love the radar stuff
Are their similar radars that have longer ranges that are available for hobbyists?
Thx
As many others stated, they are not real radars.
I had real radars ( they are more complex) on the channel. But their range is not much longer.
Is works well in same room?? Wow.haha amazing
???
i really hope youtube can have pause notifcation for certen time option like this for creaters like you to take a break
I do not know if they have such a feature in their plans...
An LDO has a minimum dropout voltage, if it’s being used to generate 3.3V for the controller chip maybe you’d be better off connecting 3.3V to both the Vin and 3.3Vout pins.
I do not see why I should connect anything to the input if I connect power to the output. As you say, there should be a difference between the two pins. So the input will be cut off by the LDO.
Stupid RUclips has not been updating me on your new videos. Now I see just in time for summer break!
Indeed, I enjoy already beautiful weather ;-) Happy summer for you, too!
Andreas, in your other radar video #439 you are stating at 4:01 that the cheap sensors of this video interfere with the 2.4 Ghz wifi frequency... But in this (summer break) radar video you are telling us that it does not interfere with the 2.4 Ghz frequency. So, what is the correct statement?
Enjoy the summer Andreas!
This video initially had #135, and I learned these problems till I made video #439...
@@AndreasSpiess Ok, thanks Andreas.
Would they encode the radio pulse with a binary code and only accept returns with the same code? If that's the case that is pretty amazing for a device that is under $10
Keep in mind: They only have one transistor. So such functions are not possible.
Do you not wonder what frequency is using? Whether it’s microwaves? And is it dangerous being exposed to these microwaves?
Look at the power they consume an assume, that most of the power is wasted. Then you decide if it is dangerous for you. I have no problems with them
Seeing Roger Clark's www the schematic reminds me of the old audio receivers.
Indeed, these devices are regenerative receivers, not radars.
Enjoy summer, sunshine, beer and blue skies
I will. Today, we had 36 degrees in Switzerland…
Hey Andreas, love your videos, I have learnt so much from you! Thank you.
I wonder if you can help me, I'm trying to find a way of sharing the data connection (4g) from my mobile phone over bluetooth to my esp32 project so that the esp32 can send and receive data from the internet using the Arduino IDE. I have searched for months but not found a way of doing this, are you able to help with this?
I share the ESP32 internet using the Wi-Fi hotspot functionality of the smartphone. I do not think that doing it via Bluetooth is a good idea...
@@AndreasSpiess I do already have wifi working but for my application a bluetooth internet connection would be so much better for battery life and end user experience. thank you anyway
I want to connect rcwl (first one in list) to a 12v doorbell. Will i need some inbetween board, say, a delay timer relay board?
Where I live, this is available. "xcluma 5V-30V Delay Relay Timer Module Trigger Delay Switch Micro USB Power Adjustable Relay Module"
Generally, you need a microcontroller to get the behavior you want. Only if the sensor does exactly what you need, you can do without it.
@@AndreasSpiess I'm on solar 12v input. Can usb to 5v. This device puts out 3v so that's my issue. In not electronics educated, I generally watch utube and do understand. But I'm a novice. So some doorbells run of 3v. But probably more amps again. So all I'm trying to do is step up v and a. The 5seconds is ok, for an alarm. It's some 100m away from me, the setup. So audio alarm is good. And if there's continuous movement it will trigger repeatedly. So that's ok too. But ideally I want to trigger a 12v output. So I've seen there are 5v relay boards out there. Just wondered if you have the patience to demonstrate something along those lines. Many thanks.
@@earthlingthings Your questions are far too big for a RUclips comment. Maybe you search for a device that runs on 10 volts and is capable of switching at least 12-volt loads.
Or you start with simple things like a voltage regulator to convert 12V to 3.3V and so on and learn all the stuff needed to build such a device.
Great review!!
You haven't checked the ability to detect the movement of people behind concrete and reinforced concrete walls.
Reinforced most probably will not work because of the metal.
Hi, Would you suggest radar sensors for higher frequency applications above 10GHz.
The newer ones work on 24GHz (I made a video about one)
Could these be used in automotive uses such as blind spot detection of cars?
I do not know. The car contains a lot of metal and maybe it interacts. You have to try.
@@AndreasSpiess Mate, you've the hardware there, how about giving it a crack and report the findings? If you find an easy/cheap option to do so, you might have a best seller video. I'm sure it find an wider interest than more than just our regular electronics cohort.
I was thinking about design of powerfull radar from old satelite dish, will be interesting experiment.
That would be cool!
Would be interesting to try tapping off the IF signal to send to a microcontroller and do more advanced analysis on it. There's an opamp in the chip which acts as an IF amplifier.
Agree, or at least look at the demodulated IF on an Oscilloscope. Do you know some good and easy to use software to analyze signals ?
Getting speed information is trivial, just do an FFT to extract the beat frequency. Things like distance measurement is a different story and it's better to use radar sensors dedicated to that.
There is no IF. These seem to be „regenerative“ receivers where the only transistor acts as a transmitter, too.
Schönen Urlaub und Grüße aus dem außergewöhnlich sonnigen und heißen hohen Norden
Vielen Dank. Heute war es auch 36 Grad hier…
Not sure if presented, but instead of having these devices interfere on my WiFi, I've had (my) WiFi interfere on these modules. Using an ESP8266 jogged a lot of fake readings from for instance the RCWL-0516 module. I've come to realise that although they are cheap, they cannot fulfill their role in a crowded 2.4Ghz spectrum. Switching to devices that make use of other, higher, frequency bands is more reliable as far as I have experienced at the moment.
Also, the lack of directivity from these cheap modules make them trigger on people not in the same room as well, through hollow walls and in some cases even through masonry bricks. That's the issue with not getting an analog strength signal but only a binary on/off signal.
I agree with your statements. These sensors can be triggered by close enough WiFi antennas.
Enjoy the summer!
I will! Thanks.
I'm almost sure that one of the pins on the chip on the cheapest sensor can be tapped for more information. It was a while ago and I didn't finish playing with it, so I'm fuzzy about just what the extra information is and that I wasn't just fooling myself, and I don't remember which pin. Still you might want to mess around to see what you can find. I think it was the beat frequency between TX and RX so you could get a speed measurement. If I remember correctly, the chip is actually the same or similar to one used on inexpensive PIR sensors. I think it has one or two op amps built in to the chip and the signal of interest was found on the output of one or both. I think I cut the trace to the voltage out pin so I could use it for the signal..
The IC has nothing to do with the RF. It is only a logic chip. They’re is Alyson no TX and RX because it only has one transistor. It seems to be a regenerative receiver (once famous in the valve age).
@@AndreasSpiess Presumably the IC isn't there just for decoration? Somehow it must be in touch with the RF section to know what is going on so it can do useful things like turn on and off. I'm not a radiohead, so technically RX and TX may be the wrong terms. My intended meaning though is a mixture of the transmitted signal and the reflected signal have been combined to make a much lower frequency beat signal. I believe that goes into opamps integrated into the IC for amplification and that is where I tapped in.
How all that happens is black witchcraft as far as my understanding goes but I believe I'm at least approximately right about what shows up after amplification.
One way I wonder if I could have been tricked by the witch is perhaps I was seeing into the middle of some magic the IC does to do what it does and it just happened to look a hell of a lot like it responding in an analog like way to speed or distance. For instance, maybe if what it is working with from the RF section isn't strong enough, it tries again sooner and sooner the weaker it is. received
I'm in the middle of a project at the moment but I'll try to make time tonight to repeat what I did and get back to you with better details.
@@karlharvymarx2650 The IC has a function. It detects a threshold and does the timing. That is all. It is also used in most PIR sensors
Have a lovely time off, see you soon.
Thank you! 😃
Nice to see you again ```````
:-)
I don;t know that a 30 second on time makes it useless. It depends how the sofwtare in your microcontroller is organised.
I agree.
Antenna beam width is determined by the aperture. Larger aperture, narrower beam. These things are tiny so are a fraction of the wavelength and essentially omnidirectional. I am also baffled about how it works with only one transistor. I think the magic is in the antenna shape which is both the resonator and receiver/transmitter. One think about radar that really helps is that the transmit frequency is known so mixing with the RX is simple. I think in this case the entire room is part of the resonant field so changes in the environment change the frequency of oscillation. How it self mixes must be tricky. I think the S shape might have something to do with that. Perhaps the opposing elements create a phase difference and it's that small difference that gets passed on to the chip. Normally radar is pulsed to determine range but this is cw so all you are interested in is doppler return. I'm just speculating about this though.
Same transistor used for transmitter and receiver is called regenerative receiver.The transmitter is disturbed by an external object or signal and create a difference which can be demodulated. Very old way of building radios for pagers and remote control for boats and airplanes. Also built with tubes before WW2.
Oladunk is right. These are regenerative receivers. They are no more used these days. The chip does only the threshold and the timing.
Is this a reuploaded video?
I am sure I have seen a video regarding radar something pretty close to this last night. Or am I wrong?
yes, ive seen this video a year ago at least
He does mention, that these are reruns.
Indeed, this is a rerun of video #135 from December 2016...
if your windows are 'modern', they probably have a metallic coating the reflects IR, but also, natürlich, RF... I wonder if these things are mostly UWB, not 'tuned', narrow-band RF...
I agree with your statement about Glas. Our windows are simple and old…
As many others stated, these devices are regenerative receivers, not real radars. But they have a similar effect for a ver low price.
can they be used to measure material thickness?
I do not think so.
Is there a minimum distance with these radars? And also, can it be used for a very narrow detection angle?
Read the comment about the spider ;-)
they dont interfere thanks to that "around 3Ghz", thankt to the tolerances of the components they all emit in different frequencies
I agree.
@@AndreasSpiess BTW what do you think is the best solution today to use lipos with microcontrollers? There is a lot of noise out there but we need some signal here. We trust you :)
@@raguaviva I use LiFePo4 directly to the 3.3V MCU if I do not want to have switching noise.
My guess (I have not done any math):
It's the timing, they are just unlikely to listen while a signal from another returns.
Maybe. But also read my pinned comment.
I believe most glass these days has a microns thick metallic coat for trapping IR.
I agree. However, ours is pretty old.
Glass is no problem, as long as it has no metal. I use a 516 with a esp, for over a year now, with no problem. Through a wall. I use it to check for movement on my driveway.
Thank you for sharing your experience!
Modern glazing contains thin metal covering to improve thermal reflection. Probably it is also shielding radio waves.
Any suggestion to make these sensors directional?
The PWB antennas are not directional (as the tests demonstrated), except for the one which has a groundplane/cover mounted behind it. You could make them directional by mounting them in front of a metallic groundplane or corner reflector. You could even mount them at the focal point of a paraboloidal reflector. Make sure the mounting is rigid; any relative motion of the sensor and structure will trigger it.
You're essentially making this the feed for an antenna of your own; ideally you would determine the operating frequency and design the antenna for the pattern you want. But the analysis is fairly involved; if you're not too particular, you can get significant directivity from ad-hoc approaches.
There are other comments where people experimented with metal surfaces and were able to get different patterns.