Microwave distance radar

I've used these in electronic playground equipment for proximity detection. It's COTS stuff.

has basically the frontend you'd need. It's a prox, not a distance radar, but perhaps you'll get results faster by adapting something like this.

for the raw gunn diode assemblies and horns,

Steve Roberts

Popular misconception. There are excellent ultrasonic sensors for harsh industrial applications from Siemens, Honeywell, Banner, and others.

Note that most of the short range "radar" sensors are doppler devices

-- only detect moving targets and give no range information. The ultrasonic devices can measure range to a stationary target with fairly good accuracy.

Hello Tilmann,

This frequency is popular with detectors of automatic sliding doors. The kind they have at the entrance and exit of a supermarket. Now don't go down to your local store now with a screwdriver in hand ;-)

AFAIK alarm system use it as well but probably the door hardware is easier to buy and may be less expensive.

Regards, Joerg

Hello,

I am looking for OEM components suitable for making a microwave distance radar. The preferred frequency is the 24.125 GHz ISM band, the distances that shall be detected are about 2-10m. Signal processing hardware is not mandatory, even the basic RF frontend along with the required mechanical parts (horn, reflector etc.) would be fine. It must be free to use in europe (without registration), however.

Perhaps someone has a tip where to look for such devices (makers, distributors, resellers etc.).

Thanks,

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

cs snipped-for-privacy@hotmail.com schrieb:

Yes, it is safer against noise and dirt. It's for harsh industrial environment.

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

Joerg schrieb:

Thanks for this hint, I will try to get more information about these sensors.

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

snipped-for-privacy@gmail.com schrieb:

That sounds basically like the units mentioned by Joerg.

Thanks for the note, I will look for more detailed information.

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

What's the target size?

--
Paul Hovnanian     mailto:Paul@Hovnanian.com
------------------------------------------------------------------
We are confronted with insurmountable opportunities.
                -- Walt Kelly, "Pogo"

How are you going to do this? Are you planning on getting range by pulsing CW at a single frequency and measuring the time interval before the reflected pulse returns? You may find this somewhat challenging, especially at a range of 2 meters.

In general, range-finding requires bandwidth. The more precisely you need to know the range, the more bandwidth you need.

For example, if you have a narrow band antenna, and you hit it with a sudden pulse of 24 GHz, the amplitude response of the antenna will ramp up in a leisurely fashion, making leading edge detection very difficult.

You have the same problem all along the transmit and receive signal path. Whoever said to use ultrasound was probably giving good advice.

You could check out massa.com to see if they have anything.

--Mac

snipped-for-privacy@uakron.edu schrieb:

Thanks! Good link.

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

Don Foreman schrieb:

Laser and ultrasonic distance sensors have been tested in that kind of application before, and were discarded (by several independent customers). They have been replaced by microwave detectors, which work perfectly. So it's reasonable that the customers ask for microwaves first.

However, if there really are "good" ultrasonic sensors, I would be glad to get concrete product names or a link to a product (or survey) page. The industrial ultrasonic distance sensors I found 'til now covered only the range up to max. 7 m distance, BTW.

The target is not always stationary, it might be moving. The resulting doppler effects might disturb the ultrasonic sensor more than the microwave sensor.

I know that many common microwave radars are doppler units for speed measurements (the rather cheap COTS devices are for automatic traffic lights and speed signs, for example). That's why I explicitly mentioned distance measurement.

Thanks,

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

Paul Hovnanian P.E. schrieb:

Rather large. Something like self-driving machines, vehicles, cranes and the like. However, other objects might also be around, so the sensor should measure the distance only in one explicit direction (with a viewing angle of, say, 10-20 degrees).

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

Mac schrieb:

It will work somehow like this. I think that's the way commercial microwave proximity sensors work, don't they?

The required precision is not too high, a roughly estimated distance will be sufficient (say, 10 or even 20 % tolerance).

I think it would make sense to use a pulse train and then correlate the received signal against the sent one to get the time delay as precise as possible.

I fear that it will be hard to get the ultrasound into a directed beam of about 10-20 degrees opening angle, so it will detect everything but its target...

I will surely do.

Thanks,

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

Perhaps someone knows a source (low quantity, OEM) for the 77GHz units used in collision avoidance systems in some cars these days? jj

I don't think so. I think they just use Doppler. If the target is within the detection range and has a non-zero velocity in the downrange direction, then the door opens (or whatever).

This is relatively easy, because you just transmit CW continuously and mix the receive signal back with the transmit signal. At the output of the mixer, you put a low pass filter (LPF). Any target that approaches or retreats from the antenna will show up as a sine wave at the output of the LPF.

The standard radar equation is res = C/(2*BW). C is the speed of light, and BW is bandwidth. Technically, resolution refers to the ability to discriminate between two targets separated by that distance. And the equation is correct when taken that way.

But there is no law of physics which says you can't estimate the range of a target to greater precision, as long as you know that there is no other target nearby. I don't know enough about your application to suggest whether this is the case for you.

Anyway, re-arranging, and assuming you want 2m resolution we get:

BW=C/(2*res) = 300e6/(4m) = 75 MHz of BW.

So the easiest way to get your result might be to use an FMCW (aka FM homodyne) radar. See if you can find 75 MHz of bandwidth somewhere in the microwave spectrum that you can use. Good luck!

Hmm. So you send out a series of CW pulses with a gradually increasing pulse repetition frequency (PRF). You use a narrow-band detector to detect the reflection, and you correlate this against the transmit PRF. Anywhere where the correlation function peaks up, you have a target. That might work. This is not a very sensitive technique, but at 10 m, you may not need much sensitivity. (How big is the target?)

The equation I gave above about BW would still hold true for your PRF. You will need 75 MHz of PRF change to get the resolution you want. But you can stretch the time out for quite a long way (depending on how often you need updates)

The other thing is, you may want to look into the X-band FMCW tank level detectors. They seem to have a lot of these problems solved, although they probably can't be used outside of a tank.

I still like ultrasound for this.

--Mac

Mac schrieb:

Agreed. This surely is the simplest way, and since these detectors aren't too expensive, this probably is their method.

Meanwhile I found two companies that use FMCW radar at 24.125 GHz (it's a registration-free ISM band). Seems to be possible. :-) (However, I am still looking for alternatives.)

Thanks for the note. I will give them a try, perhaps their frontends can be modified for free-air use?

But IMHO it's hard to get a directional sensitivity, isn't it?

Many thanks for your helpful comments,

--
Dipl.-Ing. Tilmann Reh
http://www.autometer.de - Elektronik nach Maß.

A wavelength of 40 KHz ultrasound is 8.4 mm, which is a wavelength of about 35 GHz RF. Air-coupled ultrasonic transducers operating well above 100 KHz are available.

The Massa E188 xdcr, operating at 215 KHz, has 3dB beamwidth of 10 degrees though it's only about 12.5 mm dia.

For a complete product, with claimed accuracy of 15mm, see: