Wednesday, December 17, 2008

a very simple 2.4GHz meter

I built a very simple meter that reads the strength of radio energy in its vicinity, without amplification or anything fancy. It picks up radio waves roughly around 1 to 3 gigahertz. It does very well in detecting cell phones, microwave ovens, and wireless access points.

On a simple level, the antenna converts radio waves into an electrical AC voltage, which is then converted by the germanium diode into a pulsing DC voltage. A capacitor stores the pulses and smooths out the pulsing and leaves a very small DC voltage, which is measured directly at a tiny multimeter I picked up at American Science and Surplus. I set the meter to DC voltage, 200mV scale.

The biquad antenna is sensitive to vertically polarized waves and slightly directional as well. A lot of designs also stick a ground plane behind it to increase the directionality, but I was looking for more of a field strength probe, rather than having an antenna that got me the most gain. It's a trade-off since the detector is so simple and without amplification, but that's what you get.


What can we measure with this meter? My standard test suite for gigahertz-ish radio frequencies is delinquent*, so all I can think of is cell phones, microwave ovens, and wireless networking. I found that the office microwave oven puts much more energy out at the hinge side of the door and a fan vent on the side than it does the meshed window. It can saturate the meter at the 200mV scale (when right next to the fan vent). Cell phones periodically check in with their towers (you can also tell this with a set of computer speakers anywhere near a GSM phone). They also do put out a bit of gigahertz radiation; I can detect them sending text messages from about six feet away and more when talking--they can also reach 200mV. Wireless 802.11b and g networks are actually pretty low-power in the scheme of things--they hard to see except close-up until they are transmitting data; then I can detect them six or so feet away. During idle they emit a "beacon" 10 times a second. The wifi antennas are also a good source for checking the polarization of the biquad--I get nearly nothing from them if I rotate the receiving antenna 90 degrees.


Outside the signal level varies greatly. There is a pervasive field which is presumably cellular networks and the addition of all the 802.11b/g/n networks. On the University of Chicago campus near the Regenstein Library the average strength varies from 0.2mV to 0.6mV, with a couple of spikes to 1.0mV. There are also areas of much stronger than ambient. For instance, outside of the Medici on 57th street the average field strength is 2-3mV and peaks at moments at 8mV. There are cellular tower antennas on a school across the street; so it seems likely the area is getting a particular sector of the tower.

I used this Field Strength Meter for 2.4 Ghz Wireless LAN as the excellent template for the project. For my version I used a standard 1N34A germanium diode--this is a more sensitive diode, and turns on at 0.3V instead of 0.6V like a standard silicon diode. I painted the diode black, as the diode proved to be photo-sensitive (all PN junctions are light-sensitive, and ones in transparent glass tubes even more so). Also, I didn't tune the capacitance at all. I then ran two wires to the inside plugs and drilled a hole in the case and hot-glued the antenna to the front of the meter. It's nice and compact, although the GP23A 12V battery in it doesn't last very long.

I could see easy modifications of this system, putting a simple FET amplifier or such to increase the meter response. If I use one of the inexpensive multimeters from Harbor Freight I might have enough room for a prebuilt circuit and a more directional antenna. I'd also like to see if I can pick up both aviation and weather radars with such a simple system. Maybe I should make a Sardine Can antenna?

*delinquent is also a synonym for nonexistent.


No Rad said...

Nice !
You need to calibrate it according to a real RF meter.

Unknown said...

Can someone post details of the circuit please if they have it..

The site article links to is long gone- Thanks !

Dean W. Armstrong said...