Fullscreen HTML5Med Res (??MB)HTML4

CIRCUIT NOTEBOOK Interesting circuit ideas which we have checked but not built and tested. Contributions will be paid for at standard rates. All submissions should include full name, address & phone number. Building a better mousetrap Here in rural SE Queensland, there has recently been a proliferation of mice. Apart from spreading nasty diseases like leptospirosis, they can also do a lot of damage. We had one in the house, and every night it would go through the pantry, ripping open packets and destroying the contents. So I purchased a cage trap from the local hardware store. This has a mechanical trigger, operating via a rod to release a hinged door. This door is spring-loaded with a bar which falls as it closes, preventing the door from being pushed back open. This looked like a good concept, but unfortunately, that was not the case. Most times, the bait would be taken, but the trap would not trigger. Who said mice were dumb? I concluded that a more sensitive trigger would be the solution, and decided to design a light-beam trigger for the trap. This subsequently proved successful. I removed all the existing trigger parts siliconchip.com.au and made an aluminium plate which clips onto the cage and can be easily removed to facilitate mouse removal. This plate was fitted with a solenoid release mechanism, a microswitch (to turn off the power once triggered), and also provides support for the electronics and the IR LED and receiver. A primary criterion was to minimise current consumption so that it could be powered by dry cells (eight AA cells giving around 12V). My final design draws only 1mA, so it should run for at least a year. I selected an IR receiver as used in all manner of remote-controlled devices. They cost less than $1 and are easily sourced from local suppliers or eBay. They have an amplifier with AGC, and a bandpass filter at 38kHz plus a data detector and output driver. The 38kHz BPF provides immunity from outside optical interference. They run from 5V. I initially tried sending a 38kHz square wave via an IR LED, but the receiver detects for only about 200ms then stops. I concluded that the IR LED needs to be modulated to simulate data so that the IR receiver will operate continuously; the data sheet is not clear about this. I subsequently pulsed the 38kHz IR LED at 15Hz, and this gave a continuous 15Hz square wave at the receiver output. IC1a-IC1d are schmitt trigger NAND gates. The 470nF capacitor is alternately charged and discharged as output pin 10 of IC1c toggles, producing a 15Hz square wave. This is fed to pin 1 of 38kHz oscillator IC1a, switching it on and off. The pin 3 output of IC1a feeds the 180pF/10kW RC high-pass filter, and on each falling edge, a short pulse toggles IC1c, reducing the output duty cycle to 1.7µs. This greatly reduces the current drawn by the IR LED, which is driven by 2N7000 N-channel Mosfet Q1. I found that the optical path had a range of a few metres, far more than required for my application, so in the 61 interest of minimising current drain I adjusted the dropping resistor for the IR LED to 3.3kW, giving a range of about 300mm. Too much LED current can also saturate the area with the optical signals, causing reflections to prevent the beam from breaking when it should. The receiver output is a steady 15Hz signal, and to detect that the beam is broken, this is fed to IC2a, a monostable wired as a missing-pulse detector. The R & C values set it to about 160ms, so the light beam has to be broken for at least that long to trigger the trap. This prevents false triggering. I used a solenoid purchased from eBay, rated at 28V AC, but found it works fine on 12V DC. The 5600µF capacitor delivers a hefty pulse and the solenoid closes very rapidly (I wouldn’t want to put my finger into it!). The output of IC2a feeds a pulse to another 2N7000 (Q2) via a 4.7µF capacitor, and Q2 switches on P-channel Mosfet Q3 which drives the solenoid. Q3 is seriously oversized for the job, but I had it in my junk box. The diode across the solenoid protects against any back-EMF. The reason for the RC network on the output of IC2a is to deliver a single drive pulse. Without this RC network, if the power were left on with the light beam broken, the solenoid would be permanently activated, flattening the battery. The red LED in series with the 5V regulator input shows that the power is on, and it flickers with the 15Hz modulation. I wired it in series as the regulator only draws about 1mA, and that gives a noticeable glow without adding to the overall current consumption. The IR detector I used was mounted on a small PCB, with a 10kW pull-up resistor. Checking the data sheet revealed that the device already has an internal pull-up, so I removed that external resistor to save current. I then placed this PCB in a small box to prevent the entry of unwanted external light. If you have an oscilloscope, you can monitor the optical receiver’s signal output to set the correct amount of transmitter LED current. Editor’s note: due to variation in schmitt trigger thresholds, it’s worth checking that the signal at pin 3 of IC1a is close to 38kHz, and if not, adjust the value of the 6.8kW resistor. Bruce Boardman VK4MQ, Highfields, Qld. ($125) 62 Silicon Chip Australia’s electronics magazine siliconchip.com.au