ultra low power 433MHz detector?

If you can't use power reduction by using on/off cycling, it will be hard to do.

The received power will be in the 30nW range (optimum antenna orientation and good RX antenna). In real world it will be well below this level.

With a LC resonator you can convert this 30nW to say 10 kOhms resulting in about 25mVp RF. When using a real microwave schottky diode, this would result in around 5..20 mV rectified EMF.

Note that when you lose 6 dB received power, the output voltage will drop with 12 dB. If your application requires detection of moving transmitters (or the receiver is moving), you may get some constructive interference helping you. In case of a stationary setup, passive detection will not give you reasonable reliability as the antenna may be in a notch due to multi path effects.

The low power amplifier should be able to detect voltage levels of some mV, and should have relative high input impedance 100k..1MOhm to get most of the diode EMF at the input of the amplifier.

You can use more resonance to get higher impedance (hence RF voltage across a micrwave diode), but for example temperature effects (think of diode capacitance) and mechanical size will put a limit on the maximum Q-factor.

A diode voltage doubling circuit will very likely not help you in combination with high Q factor RF input circuitry.

You may lose significant power if you don't have sufficient space to construct your resonanting antenna.

In one sentence: It isn't impossible from a theoretical standpont of view, but practicallhy spoken it depends on many other factors.

"Wimpie" ha scritto nel messaggio news:a387a$52b04a32$5353a19c$ snipped-for-privacy@cache1.tilbu.nb.home.nl...

Thank you for your accurate and realistic point of view. At least now i know that it is not easy at all,and this will avoid me to wasting time, trying to obtain state of art performances ,relying on my very basic skills. Many thanks Diego

There was an application circuit that did exactly that, a design idea in one of the electronic magazines IIRC. I vaguely recall it used a resonant antenna loop, a schottky diode detector, then a (Maxim?) ultra-low power comparator.

HTH,

James Arthur

Have a look at the design of the dsrc tolling tags you put on the cars front window.

Br, Rune

The Q of a small resonator is not very good, so you are going to get a lot of false triggerings from nearby transmitters, such as amateur radio transmitters in the 3-100 W power range, which is located at or within a few hundred kilohertz from your 433.920 MHz nominal frequency.

This sounds like a remote-doorbell kind of receiver; it ought to be possible with discrete components (or with ICs, if you can spend long times in sleep mode, and wake with a 50 uA budget for a few milliseconds). Disambiguating a real signal from noise is likely to be more easy if you modulate the trigger signal, and maybe the microcontroller can decide if the modulation is right, or if it should just go back to sleep...

ha scritto nel messaggio news: snipped-for-privacy@4ax.com...

..thanks,upsidedown i was not considering this even worst case

thanks for all your answers I tried unsuccessfully to find that design idea that dagmar was talking of

yes,it is something like this,the receiver waits for the closeness of the Tx,the 433 MHz is OOK modulated(i forgot to say it,sorry),the detector should wake the ic able to detect pramble,sync word etc.,it is a well known CC1101,that draws 15 20 mA when activated in Rx. It has a mode called Wake On Radio,that draws "only" 8 uA,I stil don't understand if in this mode it can detect a preamble and then turn RX fully on,anyway 8 uA are a bit too much for the battery it should mount

I couldn't find the original either--too many false hits--but this is the basic idea:

See fig. 10. In section 5.3.3 they say it triggers at 8 meters with a key-fob transmitter.

Quiescent draw of 1.5uA--could be improved.

Cheers, James Arthur

Thank you , this is described precisely,so that i can build something similiar to test it

Sorry to join this a bit late. Have a look at this:

It is similar to the CargoNet idea posted elsewhere.

Are you sure you only want to detect devices that have amplitude modulation? The solution in the article is for AM (there is capacitor in series with the 1kOhm resistor).

Other thing, I don't believe the 1:2 autotransformer as with -65 dBm input, this would generate just some uV and then the gain of the opamp stage is too low (1000) to wake-up the uC. You need the resonance (increase of sensitivity and out of band rejection).

You can be happy that you may have some more input power. Your design challence will be in the RF parts left of the schottky detector. use a real microwave diode (such as BAT 15), not a hybrd (guard ring) schottky rectifier.

Can you accept false-alarms as mentioned by others (due to strong near or in band transmissions)?

Study that idea. I've used that idea with different op-amps and it works.

You could experiment a bit with the 433Mhz.

Jamie

If you are going to transfer some information (such as a device ID) across your link, you need to consider also blocking issues with such simple receivers.

We have a club building on top of a hill with a 434 MHz 30 W repeater on the roof. We had problems with illegally parked cars outside the building. One way to scare these cars away was to look when the car owner approached his car, was to fire up the transmitter and the car owner tried frantically to open the car doors with his key transmitter at 433.92 MHz. The receiver in the car had so bad selectivity, that it was completely blocked by the strong signal nearly 1 MHz away.

If you are going to use some simple LC front head for RF detection, you need much more selectivity for any actual data extraction to reliably work in the RF polluted world.

I would suggest using a proper receiver with passive LC front end and proper IF filtering or zero-IF (I/Q) system which is activated for very short periods of time. Here is the critical issue, how fast will the local oscillator start. Some PLL constructs are notoriously (1 s) slow to stabilize, forcing to use very long activation periods and hence large average consumption.

However, if the crystal oscillator and some DDS construct starts in less than 1 ms, you should be able to power up the receiver for 1 ms every 1 s, the duty cycle is 1:1000, compensating for the high current consumption of a decent receiver. You may have to take the RF-sense after some 16 kHz FM filters, if you are going to use only 1 ms sampling periods, even if you use 100Hz final filters for data extraction when the RF appears to be strong and hence constantly turn on the receiver.

"Wimpie" ha scritto nel messaggio news:619e5$52b1ee18$5353a19c$ snipped-for-privacy@cache1.tilbu.nb.home.nl...

Thanks once more for the hints i think that i can accept false alarms,this circuit,while receiving a preamble should only wake up a microntroller and a receiver IC, I can decide the preamble elngth, hopefully i should have anough time to syncronize on the remaining pramble and to detect the sync word,in case of false alarm wasted power would be the only damage . Of course if the circuit will not be able,in the worst case, to discriminate powerful and close radio ham Tx,the consumption due to false alarms would be an issue Now i'm not able to predict

"piglet" ha scritto nel messaggio news: snipped-for-privacy@googlegroups.com...

Thank you, another simple and interesting circuit tha i could arrange in one hour or less i've used the LTC1540 in a previous project(audio detection)and it worked fine,i should have one right here

I' let you know about the results Thanks once more

On a sunny day (Thu, 19 Dec 2013 11:30:41 +0200) it happened snipped-for-privacy@downunder.com wrote in :

You can do a replay attack too, and open or lock his doors: