Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

RF Detector not detecting cellular phones

Status
Not open for further replies.

M84AB3

Newbie level 5
Joined
Feb 2, 2019
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
110
Hi all,

I was hoping somebody can tell me why the below rf detector is not picking up any cellular phone signals? I am using through hole components and have soldered everything onto a CNC made PCB boards, keeping all tracks as short as possible (see image). I have no problem detecting 27Mhz, 40Mhz (wireless home phone), 466Mhz (walkie-talkie) but I am getting nothing from 2.4Ghz WiFi or 3G/4G frequency(s) even when holding the phone literally a few millimeters next to the antenna. I have ordered a high frequency, wideband LNA monolithic amplifier to try boosting the signal before rectification eventhough I think a cell phone transmitting at at least 1 watt when attempting to make a connection immediately next to the pickup antenna would not need a LNA pre-amp stage. Curious to hear some of you more experienced RF guys input!!!


rf_detector.jpg

rf_detector_1e.jpg
 

I don't think that diode is appropriate for 2GHz.

After a closer look, maybe it's ok. But I would look carefully at your layout.
 
Last edited:

I guess the diode may work up to 1GHz but a signal amplifier ( even a simple one with a single transistor) will help to detect the signal strength.
The diode is not appropriate at all for 2.4GHz..
 

Thanks for the response guys!

Why would the diode be inappropriate for 2Ghz? It has 2pF junction capacitance. Plus, others seem to be using it successfully for fine tuning their 2.4Ghz RC transmitters. See link.

https://www.rcgroups.com/forums/showthread.php?1557232-low-cost-rf-field-strength-meter-for-rc

4G carrier frequencies in my country are anywhere between 850Mhz and 2.6Ghz so I am thinking I should be picking up something. I think I should mention that I have also tried a pair of BAT62 diodes but on the breadboard circuit, not PCB and that also had no issues detecting 40Mhz and 466Mhz but not the cell phone.
 
Last edited:

"It has 2pF junction capacitance."

You do realize that 2 pf is a HUGE amount of capacitance at 2.4 GHz, right?
Get a diode with 0.2 pF.

Also, does your PC board have a ground plane on the backside?

I would also change it to a single diode.
 

On the net I found another schematic, also using 1N5711 diode, but followed by an amplifier.
http://www.techlib.com/electronics/cellhelp.htm
This schematic was designed by Charles Wenzel, one of the biggest guru in RF.

Using the spice model offered by ST for 1N5711 and using LTspice, I see that 1N5711 works fine at 2.4GHz AS a detector.

Remove the 50 ohms resistor at the input of your detector, and probably your circuit will work just fine.
 

Attachments

  • LTspice.jpg
    LTspice.jpg
    66.9 KB · Views: 155

"You do realize that 2 pf is a HUGE amount of capacitance at 2.4 GHz, right?"

Obviously not big enough to stop a 2.4Ghz FlySky FS-HF030 RC transmitter. See video below. I do mention in the same post that I have tried a pair of BAT62 diodes that are advertised, in the datasheet, as "Low barrier diode for detectors up to GHz frequencies" and have capacitance of 0.35pF (typical) but that too produced no results when trying to detect my cell phone (iphone 6s on 4G network). Neither did it work in detecting the 2.4Ghz WiFi signal from my router.

I had forgotten that I had a 2.4Ghz RC transmitter and tried that and sure enough the rf gets detected. It is based on the A7105 chip and uses FSK/GFSK modulation. This to me at least proves that the circuit works at least up to the 2-3Ghz band, with the highest currently operational 4G frequency being 2.6Ghz, only 200Mhz above 2.4Ghz. I thus cant help but wonder if is something to do with the short cell phone bursts? All three devices I demonstrate in the video below have a continuously emitted RF signal but I am not sure if cell phones transmit the same way.


- - - Updated - - -

"Also, does your PC board have a ground plane on the backside??"

I would say so, see attached image.

WP_20140522_007a.jpg
 

All three devices I demonstrate in the video below have a continuously emitted RF signal but I am not sure if cell phones transmit the same way.

As I mentioned, seems that the detector works at 2.4GHz. I see that you didn't remove the 50 ohms at the input. That will improve a bit the sensitivity of the detector. You can replace that resistor with a 0.1uH RF choke (the same as in the 2nd schematic in link that I posted above).

Cell phones in 2G mode transmit bursts with duty cycles varying between 1/8 and 4/8. In 3G and 4G modes there is a continuous transmission during a call.
In cell phones, their output power vary from -50dBm to about +24dBm in 3G and 4G and up to +33dBm in 2G. So the power is not constant during a call as in your cordless phone or RC transmitter.
 

As I mentioned, seems that the detector works at 2.4GHz.

But this layout is far from ideal. By shrinking the layout and using proper RF layout technique, the detector could be much more sensitive.
 

Totally agree, a good PCB would help. And is not impossible because seems that have access to a PCB milling machine, which makes the job easier.
But even the best PCB design would not help if a relative high impedance detector input is shortened to the ground with a 50 ohms resistor.
 

As I mentioned, seems that the detector works at 2.4GHz. I see that you didn't remove the 50 ohms at the input. That will improve a bit the sensitivity of the detector. You can replace that resistor with a 0.1uH RF choke (the same as in the 2nd schematic in link that I posted above).

Cell phones in 2G mode transmit bursts with duty cycles varying between 1/8 and 4/8. In 3G and 4G modes there is a continuous transmission during a call.
In cell phones, their output power vary from -50dBm to about +24dBm in 3G and 4G and up to +33dBm in 2G. So the power is not constant during a call as in your cordless phone or RC transmitter.

I did remove the 50 ohm resistor by cutting the bottom lead with wire cutters. It is not obvious in the video. If the 4G transmission is continuous then why am I not able to pick it up with the antennas literally touching one another? If it works for 2.4Ghz RC transmitter, I don't see why it wouldn't work for a cell phone.

But this layout is far from ideal. By shrinking the layout and using proper RF layout technique, the detector could be much more sensitive.
As I do have a CNC machine I will go all out SMD on this one and see what results I get. I have ordered a few MGA-86563 MMICs to use as the front-end LNA.
 
Last edited:

Hi,

As I do have a CNC machine I will go all out SMD on this one and see what results I get.
I see the "ground plane" problem much more important than the SMD/THM problem.

Klaus
 

Hi,


I see the "ground plane" problem much more important than the SMD/THM problem.

Klaus

Can you please explain in a little more detail? That statement alone doesn't tell me much.

By ground plane I assume you mean something as described here under the "Microstrip" section.
 

Hi,

as stated in post#8:
It doesn´t help to generate a large area of "unconnected" copper. You need to connect it to system_GND and need to use it as GND_Plane.

Klaus
 

Alright, I have redone the PCB with all of the GND points connecting to the large piece of copper. I have also used BAT62 diodes instead of 1n5711 as they are officially rated for GHz and have even lower capacitance than 1n5711 diodes. Unfortunately, the circuit is now even LESS sensitive, actually SIGNIFICANTLY less sensitive than the original one I built (depicted in my first post), without a proper ground plane. How or why would this be? There are no issues (shorts or wrong connections) with the PCB, I have checked like 10 times. It is worthwhile mentioning that the original PCB actually started picking up 4G (only data) transmissions, and quite well I must say, once I placed a piece of copper/metal sheet underneath it with a piece of paper acting as a dielectric. It would however not respond to WiFi or Voice calls.

rf_detector_1f.jpgrf_detector_1g.jpg
 
Last edited:

From an RF point of view, that is a complete disaster. You really need to understand the concept of wavelengths then closely look at the tracks around the RF front end. The track carrying signal from the antenna is almost as long as the antenna itself. Why do you have five capacitors in parallel?

Also consider the antenna is hardly wide bandwidth and you are asking for a detector covering at least 27MHz to 2,400MHz.

A full schematic of your board would help with diagnosis.

Brian.
 

From an RF point of view, that is a complete disaster. You really need to understand the concept of wavelengths then closely look at the tracks around the RF front end. The track carrying signal from the antenna is almost as long as the antenna itself.

There is no need to be condescending. I am here because I am not an "expert" in RF design and am trying to learn. Calling something a disaster without providing constructive feedback is hardly helping. What makes you think I do not understand the concept of wavelengths? Do you think my CNC machine was a christmas present and here I am slapping together "some circuit"?

I understand that this is by no means an efficient design, and it should really be all SMD but I am not designing some ultra sensitive 5G receiver and I would have thought that for close range (close to the transmitter) it would pick up something across the major frequencies.


Also consider the antenna is hardly wide bandwidth and you are asking for a detector covering at least 27MHz to 2,400MHz.
This Antenna came off of a GPRS Shield v1.0 running a SIM900 module and works on the 850/950/1800/1900 bands. 850Mhz to 1900Mhz seems pretty "wide" to me. Additionally, it works "exceptionally well" at 40Mhz and 466Mhz. Similar to how an AM wavelength is in the range of 300 meters but you can still hear it on your radio without a 150m or a 75m antenna attached to your house I would have thought I would pick up at least something THAT close to the transmitting source.

A full schematic of your board would help with diagnosis.

Brian.

The schematic is exactly as depicted in my first post, minus the 50ohm resistor and the 1n5711 diodes have been replaced with BAT62 diodes. In addition to the 10pF and 1000pF capacitors I have added 100pF, 10nF and 100nF to try and see if it makes any difference with different carrier frequencies. The rest of the schematics is less important as it is just an op-amp (CA3130) with 100 gain that picks up the signal once it has already been rectified, in essence its just a DC voltage amplifier.
 
Last edited:

My apologies, it wasn't meant to be condescending.

The point I was making is that at high frequencies, because the wavelength is shorter, the effect of wiring length becomes critical. Taking your design, if the antenna is orientated as in your photographs, you have almost a quarter wave of 'backwards' track picking up almost the same signal as the antenna. The effect could be signal cancellation but the zig-zag wiring would make it hard to predict exactly how much cancellation might occur. At best you would get a very variable coupling response, it could be almost nothing at some frequencies up to total at others. SMD has little to do with it but it does help to make wiring lengths predictable. I have units here that do not use SMD but operate at 2.4GHz quite happily.

An antenna, especially one that probably has a coiled element inside it, has a complex 'gain' response and directivity pattern across different frequencies. It may work well at some frequencies but reject others. Because it is rated at 850MHz and at 1900MHz doesn't mean it works equally well between them. In all probability, it has 'traps' or gaps in the element so it peaks at specific frequencies. It should work against a ground plane anyway so for best results it should be at 90 degrees to the ground plane on the PCB. (facing the same way as the components)

You only need one of the parallel capacitors and the value isn't critical because in theory there is a DC voltage there. A single 1nF capacitor should be sufficient and adding more has no significant effect on frequency response.

I would still like to see your schematic. I wonder if you are seeing an effect that has nothing to do with the antenna or detector diodes at all. If I follow the component layout and tracks from your photographs, the biasing is wrong on the CA3130 and it can't possibly amplify. What you might be doing is measuring RF pickup on the amplifier pins rather than through the antenna. RF leaking into the IC might shift it's internal voltages and give the impression it is working, although not correctly.

Brian.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top