Fullscreen HTML5Med Res (??MB)HTML4

Is your microwave oven safe? By Nicholas Vinen Don’t get zapped!!! Test it with this . . . Microwave Leakage Detector Just because your microwave oven still looks shiny and new does not mean it is safe. It could be leaking lots of microwave energy, potentially putting you at risk of being zapped. Now you can easily test it with our tiny Microwave Leakage Tester. As a bonus, it will also test WiFi access point activity. D O YOU SLAM the door of your microwave oven after you have used it? Of course, you do! Everyone does! That repeated slamming can damage the integrity of the mesh inside the glass door so that as time goes by, the shielding becomes less effective, allowing some microwave radiation to leak out around the edges. In particularly bad cases, enough RF 34  Silicon Chip could leak out to cause injury. Since RF energy is invisible, the question becomes how do you know whether your microwave oven is still safe? Basically, you need to run our Microwave Leakage Detector around the edges of the door while the oven is operating, to check that it’s safe. The level of microwave leakage (field strength) is indicated with an 8-LED bargraph. As the microwave oven is operating (and the turntable is rotating), you will see a surprising variation in leakage as you run the microwave detector around the door edges. The detector is powered by a lith­ ium button cell with low drain and no standby current. This unit can also be used to check whether 2.4GHz transmitters, such siliconchip.com.au as those in WiFi routers, are active. If the Detector is held up to the base station antenna, its LEDs will flicker in response to network activity and the number of LEDs lit will indicate the transmission power level. siliconchip.com.au +3V + REG1 LM4041 DIM3–1.2 100nF 15k – 417mV Operation Essentially, a microwave leakage detector is a type of AM radio receiver tuned for signals around 2.5GHz, with an indicator of the RF field strength. Because we’re mainly concerned with indicating the presence of fields and whether they are above a certain hazard power threshold, we don’t need a very complex circuit. Issues such as distortion, bandwidth, linearity and so on are not important. The basic principle of this detector is based on an article published in the July 1979 issue of Electronics Today International (ETI). A dipole antenna is formed from two collinear tracks on the PCB, with each track’s length being one-quarter of the wavelength for 2.5GHz signals (12cm). A “hot-carrier” Schottky barrier detector diode is used to detect the signal and its output is filtered by a 220pF ceramic capacitor. Two low-value inductors, comprising zig-zag tracks on the PCB, connect the detector diode to the filter capacitor. These prevent the capacitor’s low impedance from excessively loading the dipole and also enhance the filtering operation of the capacitor at microwave frequencies. The original ETI project used a tiny moving-coil analog meter (as used in squillions of tape recorders in those days) to display the received signal strength. This had the advantage of making the device entirely passive, ie, the received RF energy operated the meter and thus no battery was required. However, while this was cheap and simple in those times, it would now result in a fairly bulky and moderately expensive unit, with the meter costing around $17.50 today. Our new design can be built for less than that in its entirety. Instead of an analog meter, we’re using an 8-LED bargraph driven by two quad comparators, powered by the CR2032 lithium button cell. Since you’re only going to use this tester for a few minutes now and then, the button cell should last for years. The SMD parts used are compact and relatively inexpensive. The LEDs re- S1 12k +1.2V 7 6 1k 3 IC1b 1 1k 10Ω LED9 K 1k 313mV ANTENNA 1a ANTENNA 1b 2 IC1d 13 261mV 9 8 IC1c 14 L2 λ S D LED7 K D A G Q1 BSS138 BSS138 Q2 G LED6 K λ A 209mV 7 6 3 IC2b 1 1k 1k LED5 K 157mV K 1k – K 52mV 1k BSS138 IC2a 2 11 10 1k A 5 4 104mV MMBD301 λ A A λ POWER LED1 K IC2: LM339N LM4041DIM3-1.2 IC2d 13 9 8 D SC A +3V 1k LEDS A 20 1 6 λ 100nF 220Ω G 1k 1k 220pF + 1k K 12 L1 CATHODE BAND 1k 10M LED8 S K A IC1a 11 10 1k D1 MMBD301 5 4 CELL1 3V A 22 µF IC1: LM339N 365mV λ IC2c 14 1k 1k 1k LED4 K λ A LED3 K λ A LED2 K λ A 12 S MICROWAVE LEAKAGE DETECTOR Fig.1: the circuit of the Microwave Leakage Detector. The dipole consisting of Antennas 1a & 1b picks up ~2.5GHz radiation and this is rectified by D1 and filtered by L1, L2 & the 220pF capacitor. The voltage developed across the 220Ω load resistor is indicated by a LED bargraph consisting of red LEDs2-9 which are driven by quad comparators IC1 & IC2. The whole unit is powered from a 3V lithium button cell and switched on (for a minute or two at a time) by pressing switch S1. spond very quickly so you can easily see if the field is steady or pulsed and they’re bright and easy to see, even at arm’s length. Having said that, if you wanted to, you could simply fit the detector diode, filter capacitor and loading resistor and measure the voltage developed across it with a multimeter (available for under $5). It’s up to you; this is the cheapest and simplest option but of course, will be somewhat more awkward to use. Circuit description The complete circuit is shown in Fig.1. The dipole antenna is shown at left, connected to either end of the MMBD301 UHF diode. This then feeds the 220pF filter capacitor via lowvalue inductors L1 & L2. One end of the filter capacitor is grounded, while April 2016  35 D1 (underside): MMBD301/352 04103161 RevB 22 µF LED2 A 10M K 10Ω BUTTON CELL HOLDER 1k 1k SILICON CHIP 1k LED1 K A 1k + Microwave Leakage Detector LED9 8x Hi Red 1k 1k 1k 1k 1k 100nF 11k 1k 1k 220Ω BAT1 3V 220pF 1k REG1 1k 1k 15k 1k IC1 LM339 Q2 IC2 LM339 Q1 100nF 1k 1 S1 12k Fig.2: follow this PCB layout diagram and photo to build the Microwave Leakage Detector. All parts except for detector diode D1 go on the top of the board and most are SMDs, the exceptions being BAT1 and S1. The dipole antenna is on the bottom layer and is visible along the top of the board, as are the two zig-zag tracks that form the inductors below it. During construction, watch the orientation of IC1, IC2 and LEDs1-9. (Note: photo shows prototype board). WARNING DO NOT PUT THIS DETECTOR INSIDE A MICROWAVE OVEN AND TURN IT ON. IT WILL BE DESTROYED IMMEDIATELY! You may think that this is a silly warning but we understand that Dick Smith Electronics had a number of similar kits returned in a rather melted condition because people had done just that! the other is loaded with a 220Ω resistor. The voltage developed across this resistor depends on the microwave field strength. This voltage is fed to the inverting inputs of eight comparator stages, based on two LM339 quad comparators which are cheap and will run from a 3V supply. The non-inverting inputs are connected to a resistor ladder which provides a series of linearly increasing voltages to each subsequent comparator stage. These are derived from a 1.2V reference voltage from REG1, which is reduced to around 417mV by a 15kΩ resistor, in combination with the 8kΩ resistance of the ladder. If the voltage across the 220Ω load resistor is above 52mV, the output of IC2c will go low, pulling current through LED2 (the left-most red LED) and its 1kΩ current-limiting resistor. This resistor sets the LED current to around 1mA, sufficient for a highbrightness SMD LED to be quite visible. Similarly, if the voltage goes above 104mV, LED3 also lights, and so on. Above 417mV, all eight red LEDs (LED2-LED9) will be lit. LED1 is on while ever the circuit is powered and similarly draws around 36  Silicon Chip 1mA. IC1 & IC2 together draw around 1mA, for a total quiescent current of around 2.4mA and a maximum current draw of just over 10mA, with all LEDs lit. REG1 is a shunt regulator (like a zener diode) and is fed from the 3V battery via a 12kΩ resistor, which sets the nominal current level to (3V - 1.2V) ÷ 12kΩ = 150µA. The current through the ladder is 1.2V ÷ (15kΩ + 8kΩ) = 52µA. That leaves around 100µA of bias current for REG1; the minimum specified for proper operation is 60µA. This means the circuit should work OK even if the cell voltage has dropped to 2.55V (which would make it quite flat). The remaining components protect against a reversed battery and provide the power switch-on and auto-off timer. Mosfets Q1 & Q2 are connected back-to-back (ie, in inverse series) so that they will block current flow from the battery regardless of its polarity. With correct battery polarity, when switch S1 is pressed, the 22µF capacitor charges to a positive voltage via the 10Ω resistor and this brings the gates of Q1 and Q2 high, switching them on and powering the circuit. The 22µF capacitor is slowly discharged by its 10MΩ parallel resistor and once its voltage falls below the on-threshold of Q1 & Q2 (around 1.25V), the circuit shuts down. LED1 dims and eventually goes out. If the battery is inserted backwards, pressing S1 simply pulls the gates of Q1 and Q2 negative with regards to their source terminals, which only serves to switch them off harder, so nothing should be damaged; the circuit simply won’t operate. To calibrate the circuit, we simply adjusted the value of the 220Ω load resistor until a full scale reading was reached with fields just strong enough to set off the alarm on a commercial microwave leakage detector we purchased. Construction The Microwave Leakage Detector is built on a double-sided PCB coded 04103161 and measuring 64 x 32mm. Most parts are surface-mount and all but one are fitted on the top side of the board. The exceptions are the battery holder and power switch (both through-hole parts) and the RF diode (D1) which is soldered on the underside. Refer to the PCB overlay diagram, Fig.2, during assembly. Start by fitting the SMDs on the top side, beginning with the two ICs. Note that these are orientated with pin 1 towards the bottom of the board. Pin 1 is normally indicated with a divot or dot in the corner of the part but if there is no such marking, then you will instead need to identify the side of the package with the bevelled edge. Pin 1 is on that side. Melt a little solder onto one of the IC pads, then slide the IC into place while heating that solder. Check its orientation and pad alignment. If both are good, solder the diagonally opposite pin. Otherwise, re-heat the initial joint and nudge the part into place. Finally, solder all the remaining pins and don’t forget to add a little solder or flux to refresh the initial joint. If any of the pins are bridged with solder, clean them up with some solder wick. A small dab of flux paste will help this process. Next, solder REG1, Q1 and Q2 in place. These are in more or less identical packages (SOT-23) so don’t get them mixed up. Use a similar technique as for IC1 & IC2. Then fit the four ceramic capacitors. These are siliconchip.com.au The dipole antenna etched into the PCB works well but you can improve the sensitivity by soldering four 30mm lengths of wire to the pads on either side of D1, as shown here. Keep the wires straight; our got a little bent during photography. in 2 x 1.2mm (2012/imperial 0805) packages with no markings. The same basic technique as described above will work for these too. Follow with the resistors, which are similar in size to the capacitors but have their value printed on top in tiny text. You will need a magnifying glass to read it. The nine LEDs are next; eight red and one green (LED1). These are in larger packages at around 3.2 x 1.6mm. Use a DMM set on diode test mode to determine which end is the cathode – when the LED lights up, the black probe is connected to the cathode. Solder the LEDs with this end towards the “K” on the PCB. Note that LED1’s cathode faces towards the top of the PCB while LED2-LED9 are soldered with their cathodes facing the bottom. Now you can flip the PCB over and fit D1 before fitting the final two components, BAT1 and S1, on the top side. Testing Insert the CR2032 cell into the holder, with the positive side up. Press S1 and check that green LED1 lights up. It should stay lit for a minute or so, then dim and eventually go out. Red LEDs LED2-LED9 should remain off. If they switch on, either there is something wrong with the circuit or you are in a rather strong microwave field and should probably move! Most constructors will have access to a WiFi router of some sort and this is the easiest way to test the device, especially if you have the type with one or more external stub or whip antennas. With the unit switched on, hold it up alongside one of the router’s antennas with its on-board dipole aligned with the antenna. Assuming there is some network activity (and there usually will be, if only because the router is broadcasting its SSID), you should see some of LED2-LED9 light up and flicker as the router transmits bursts of data. Desiliconchip.com.au pending on how close you’re holding the device, some bursts may be strong enough to light up all eight LEDs while others may result in just a few LEDs lighting. Bursts that light up all LEDs aren’t necessarily hazardous as they will be quite brief, so the total radiated energy should be low. Rotate the unit and note how quickly its sensitivity drops if it is not aligned with the radiated field. This is why, when checking a microwave oven, you will need to rotate the device as you move it around the oven. Improving the antenna While the dipole etched into the PCB works, we found that by soldering four 30mm lengths of thin, stiff insulated wire to the pads on either side of D1, the detector can be made less sensitive to antenna orientation. Basically, two of the pieces of wire are soldered directly in parallel to the PCB tracks while the other two are perpendicular, sticking out the front and back of the board (see photo). We used Kynar but you could also use “bell wire”, which is a light-duty solid copper core insulated wire that was historically used for telephones. Make sure that it can’t short to anything – you may need to insulate the ends with some thin heatshrink or a dob of silicone sealant. It’s still a good idea to hold the board so that one dipole or the other is in the assumed field direction. However, even if it’s not quite perfectly aligned, you’re more likely to get a reading with this arrangement. This does make fitting heatshrink tubing over the PCB somewhat more tricky but it can still be done. You would need to solder the two parallel antenna wires, fit the tubing, shrink it down, then cut a couple of small holes and solder the perpendicular wires in place. It may seem odd soldering antennas in parallel with the PCB tracks that act Parts List 1 double-sided PCB, code 04103161, 64 x 32mm 1 20mm button cell holder, through-hole (BAT1) (Jaycar PH9238, Altronics S5056) 1 CR2032 cell (BAT1) 1 micro SPST tactile pushbutton switch (S1) (Jaycar SP0611) 1 80mm length of 30mm diameter clear heatshrink tubing (optional) 4 30mm lengths thin, stiff insulated wire (optional) (antennas for improved pickup) Semiconductors 2 LM339, LM239, LM2901 or LM3302 quad comparators, 3.9mm wide SOIC-14 (IC1,IC2) 1 LM4041DYM3-1.2 micropower 1.2V shunt regulator, SOT-23 (REG1) 2 BSS138 logic-level N-channel Mosfets, SOT-23 (Q1,Q2) 1 green high-brightness LED, SMD 3216/1206 (LED1) 8 red high-brightness LEDs, SMD 3216/1206 (LED2LED9) 1 MMBD301 single or MMBD352 dual Schottky hot-carrier diode (D1) Capacitors (all SMD 2012/0805) 1 22µF 6.3V X5R 2 100nF 50V X7R 1 220pF 50V C0G/NP0 Resistors (all 1% SMD 2012/0805) 1 10MΩ 17 1kΩ 1 15kΩ 1 220Ω 1 12kΩ 1 10Ω as antennas, but it’s important because the impedance of the PCB tracks is much higher than the thin, circular cross-section wire. So the wire antennas will dominate the response unless they are fitted in pairs as described. Using it Pressing S1 switches the unit on for 1-2 minutes. You can hold down S1 or press it regularly to keep the unit on while you are using it. It will then switch off by itself. The LED bargraph indicates the voltage generated across a 220Ω load resistor, in roughly 50mV steps. Thus April 2016  37 detector on and move it around the outside of the door – both the front face and around the sides. Also check around the edges of the window. It’s generally best to hold it such that the dipole is facing across the edge of the door. In other words, when holding it at the front of the oven, point the dipole at the centre of the door and when holding it at the sides, align the dipole so that it is pointing to the back of the oven. Note that because the contents of the oven are normally on a rotating tray, the leakage field will change over time, as the contents will interact with the field. That means you will need to move the detector slowly and pause if you get a reading to see what it will peak at, at that location. Doing multiple sweeps is also a good idea. The microwave oven is checked while it is in operation by moving the Microwave Leak Detector around the edge of the door and around the edge of the viewing window. If all eight LEDs in the bargraph light, then there is excessive leakage and the oven can be considered hazardous. Note: this is closer than you would normally hold it. the segments correspond roughly to received power levels of 11µW, 45µW, 100µW, 180µW, 284µW, 410µW, 556µW and 727µW. The danger level is generally considered to be 5mW/ cm2 however we need to determine how effectively our unit picks up the radiation in order to calibrate the bar graph response. We compared the response of the bargraph against a commercial microwave leakage detector and found that, with the 220Ω load resistor, a full scale reading (ie, all eight red LEDs lit) corresponded pretty closely to 5mW/ cm2 (the legal limit, above which it is considered hazardous) on the commercial detector. This assumes the dipole is in alignment with the field, which we determined by rotating the detector for maximum response. Operate the oven The oven needs to be operating in order to check for leaks but it’s a bad idea to operate a microwave with nothing in it. Unless you happen to have something you want to heat anyway, the simplest solution is to fill a bowl or large mug with cold water and microwave this for a few minutes while testing, then tip the water out. Take care as it may be very hot; it’s best to put in enough water to avoid it boiling during the test period. So, if you want to check that your microwave is safe, start heating some water at full power, then switch the Antenna distance Generally, you should keep the antenna around 50mm from the oven as you make the sweep. But while this is the specified distance for the legal limit, the relatively long wavelength of microwave radiation (~12cm) means it’s possible that the field strength could actually be higher further away from the oven, due to constructive and destructive interference. So a second sweep at a somewhat greater distance would not hurt. If you want to use the detector to sense 2.4GHz radio signals, it’s simply a matter of holding it as close to the radiating antenna as you can and rotating it until you get a response. Note that while it’s quite effective at picking up WiFi router transmissions, at the low power levels generated by battery-powered WiFi devices, mobile phones and other 3G/4G devices, you may have difficulty picking up enough SC energy to light the LEDs. Your new source for value Raspberry Pi gear! New Raspberry Pi 3 Model B now in stock • Fastest ever - 1.2 Ghz Quad-core CPU • Onboard WiFi and Bluetooth LE • Incredibly versatile for work, research and play • Compatible with existing accessories • All boards and accessories in stock $69 inc GST Local stock! • $5 delivery • Visit tronixlabs.com.au/sc support<at>tronixlabs.com • Telephone 0488 TRONIX • PO Box 5435 Clayton 3168 38  Silicon Chip siliconchip.com.au