Re: RF Detector not detecting cellular phones
I have to agree with Brian.
I also find it hard where to begin.
Like this comment or not: it is no HF design. I can see any design at all.
A PCB layout is not simply joining some pins of electronic parts.
"Joining" in the meaning of electrical connection, but even this is not true for all pins of your PCB. (Connector, diode)
You think you now have a HF suitable GND plane. But it is not. In many cases your "GND plane" is not better than traces.
You need to take care about current flow, current loops (mind that a singal is not from A to B you always need to consider the return path very carefully. If the return path is bad, your signal is bad and thus the performance of the whole circuit is bad.
Sadly a forum can't replace school, reading books, experience...
You can't expect to learn a HF design within a forum thread in a couple of hours where other need years of school.
I design industrial electronics for decades now, always touching HF design, but after all the years I'm far away to be an HF expert.
I just know some basics.
But I don't want to discourage you. Go on. Maybe with another - not that high frequency - circuit.
Maybe with this circuit ... but read some books about HF design first..
Klaus
@M84AB3, being not so familiar with RF, I would consider before proceeding with the design of the PCB Layout, making computational simulations; there are tools suited for that. Note that the smallest λ/4 lenght for the frequency spectrum in question, Wifi in the case, may lie in the same magnitude of the board sizings, and it would not be improbable to have any kind of reflection canceling/attenuating the signal injected from the antenna. The spectrum you are using is broad, ranging from 0.027GHz up to 2.4GHz, therefore what works "exceptionally well" for one carrier is not necessarily supposed to work fine for another.
That spec is misleading. This is a simple narrow band antenna (monopole).
On your PCB, trace length is an issue at 2.45GHz, and having GND doesn't solve that. For a quick test, take all components up to the diode off the PCB, and wire them as short as possible at the SMA. You can do some creative 3D wiring, it just needs to be SHORT.
Actually this is not such bad. You went from no RF experience, to build a 2.4GHz circuit. I've seen recently people with zero experience at RF jumping to 77GHz or even to 110GHz circuits. And they never made even simple 10MHz circuits.
1. Start learning about behavior of electronic components at high-frequencies, especially the behavior of capacitors at RF.
You are using through hole capacitors at 2.4GHz, which is very inappropriate. You have access to a milling machine. Make the circuit using SMD components.
https://www.murata.com/~/media/webre...ow/12to14.ashx
2. Read about high-frequency PCB design (which basically starts making very short signal traces and thicker as possible ground planes).
True, but the long traces have a MUCH bigger influence here.
They don't
3G cellular uses spread spectrum, so it's quite difficult to pick up a signal from it as the TX signal aka RF power is spread over a wide bandwidth
4G uses OFDMA multi-carrier transmission, which again can be difficult to get a strong signal from because of the way the signal is transmitted
Additionally, the amount of power transmitted form a cellular phone depends on how strong the signal FROM you local base station is. The power levels are adaptive, the idea being that only enough power as necessary is sent from the mobile device. This is to maximize battery life and also to minimize signal congestion at the mast. If you are close to a mast, the chances are the mobile will be told to reduce it's output power, making it harder to detect.
Brian.
I had the same experience with several different circuits I tested. No problem picking up WIFI signals or mobile base stations, but they either detected mobile phones only at very short distanes or not at all.
For instance, both power meters with the LTC5508 and the AD8313 were very insensitive to mobile phones, but picking up a jungle full of other signals.
But then again, I am not an RF expert, and experiment just for the sake of fun and curiousity.
The diode detector's sensitivity can be improved by the following:
1. Match the input impedance [1]
2. Change topology from doubler to single. This will improve the sensitivity for input power below -10 dBm [2].
3. use a zero bias Schottky [3]
Although some have commented on the 1N5711's C@0V < 2pF, but IMHO, it is NOT the most critical weakness of the leaded package because, replacing it with a plastic packaged SMD Schottky intended for microwave mixer/detector such as HSMS-282x will only marginally improve C@0V to < 1.5 pF. On the other hand, parasitic inductance is a greater difference as a leaded package has ~ 8nH [4], whereas a SOT-323 has ~1.1 nH [5]. Nevertheless, if impedance matching is used, the parasitics can be mitigated by absorbing them into the network.
References:
[1] Avago AN1089, "Designing Detectors for RF/ID Tags"
[2] Boaventura & et al, "Optimum behavior: Wireless power transmission system design through behavioral models and efficient synthesis techniques", IEEE Microwave Magazine 14(2):26-35, Mar. 2013
[3] Avago AN969, "The Zero Bias Schottky Detector Diode"
[4] R. W. Rhea, Oscillator design & computer simulation, section 1.10 "Component parasitics".
[5] Avago AN1124, "Linear Models for Diode Surface Mount Packages"