Advertisement
Smart Smoke Alarm Circuit Update 1

Smoke Alarm Circuit

The original smoke alarm circuit electroschematics.com/5450/smart-smoke-sensor-alarm/ has some problems and has left some hobbyists and students disappointed. Rather than to attempt to solve its many problems, I have created an update that uses the venerable LM741 op amp. It is simpler, more conventional, easier to understand and has much better battery life.

Smoke Alarm Circuit Schematic

Smart Smoke Alarm Circuit Update 1

Smart Smoke Alarm Circuit Update 1

Photo-interrupter module

The photo-interrupter is simply an optical coupler with the elements separated with a slot. Anything that enters this slot reduces the current transfer ratio (output current /input current). I used the H21B1 Darlington device since I did not have the H21A1 on hand. I am glad I did, because I learned that it is a better device, requiring far less LED current. Since the H21A1 /H21B1 series is no longer available from DigiKey, I selected a currently available, more inexpensive device—the Sharp GP series—see schematic(s) for details. Sharp appears to be the leader in this product.
I did not test the circuit with these devices.

Datasheets

Sharp GP1L52VJ000F Darlington photo-interrupter:
http://www.sharpsma.com/download/GP1L52VJ000Fpdf-0

Sharp GP1S093HCZ0F photo-interruptor:
http://www.sharpsma.com/webfm_send/1555

Schematic, Alternative Photo Interrupter

Alternate Photo Interrupter

Although, I recommend the Darlington device, do not despair if you have the non-Darlington version. See the smoke alarm schematic for alternative devices /circuits. These were not tested, but I see no reason why they are not viable.

Initial evaluation

Oscillograph Smoke Detection

I had my doubts as to the effectiveness of the technique, so I tested a photo-interrupter with smoke. The oscillograph demonstrates what happens to the photo-transistor collector voltage when the gap is obscured with smoke from my soldering iron as it melted rosin core solder. Cool! It does work.

The oscillograph also has a trace showing what happens when a transparent poly film is inserted into the gap. It has a lower signature. I figured if it could see this effectively, it is an easier test, so the remaining testing is with film rather than smoke.

Ambient light also affects the output voltage. In my case, it was only about ±0.1V, but the sensor may need to be shielded from outside light sources if it is a problem.

Op amp selection

The old LM741 is a good choice, but I also found a low power substitute (TI TL061) that reduces battery current—see schematic for details. As it is, battery load is only 1.1mA—not bad!

Smoke Alarm Circuit Operation

The LM741 is applied as a comparator /Schmitt trigger. The voltage on the inverting input is biased at 4.2V (half of Vcc). The collector of the photo-transistor is adjusted via R2 for 4.0V. Any reduction of the current transfer ratio causes the collector voltage to increase beyond 4.2V. At this point, the output voltage of the op amp goes positive and turns on the MOSFET. The drain of the MOSFET drops to zero volts and C1 couples a positive feedback signal to the inverting input via R8. The R*C of R8 & C1 causes the LED & siren to stay on for at least 1sec after the smoke has cleared. C1 has to be either film or ceramic because the voltage polarity reverses.

Adjusting R1?

R1 must be selected to provide approx 4V at the collector of the photo-transistor with R2 centered. Its resistance is a function of photo-interrupter transfer ratio and may vary as much as 40:1. When this is selected, R2 will be within range of making operational adjustments. The question marks on the schematics are for this purpose—actual resistance must be determined. In my smoke alarm circuit, one device required 10K while the other required 22K.

Protoboard photo

smart smoke alarm update protoboard

Note that I did not have a siren module on hand—only the LED.

7 Comments

Join the conversation!

Error! Please fill all fields.
  • thenotsoknowingperson

    Tried this update circuit. Seems it got a problem too. Where when I connected the battry, LED and buzzer will on without detecting smoke. Seems like it directing the connection thru the transistor BS170. I tried pulling out the jumper on the Source pin that is connected to the ground of the bs170, buzzer and LED off. That is the direct connection I detected.

    • Jim Keith

      In trouble shooting this circuit, measure the voltage at the output of the phototransistor –it should be adjustable to Vcc/2 –if not, find out why –wiring error, insufficient LED current, excessive LED current?

      To troubleshoot the op amp circuit, replace the phototransistor with a pot connected between the 2 power supply rails, and observe where the threshold lies –should be close to Vcc/2.

  • Radio Fanatic

    So, the older version of the smoke alarm is not working or what?

    • muhd nik

      Tried this update circuit. Seems it got a problem too. Where when I connected the battry, LED and buzzer will on without detecting smoke. Seems like it directing the connection thru the transistor BS170. I tried pulling out the jumper on the Source pin that is connected to the ground of the bs170, buzzer and LED off. That is the direct connection I detected.

    • Jim Keith

      Too many have had problems with this circuit –probably worked OK for the creator, but there are many variables involved –not recommended.

      Another issue –the opto interrupter module appears to be aging –I have had it working for a few days and now the potentiometer adjustment has changed –most likely the LED is not putting out quite as much light –this would never have been noticed except it is applied here as a linear device –interrupters are normally applied as switches.

  • Alexei

    Dear Jim, what type of siren module can we use?

Looking for the latest from TI?