(:cellpadding=5 cellspacing=0 width=1000:)

A5366CA : Photoelectric-Type Smoke Detector with Interconnect, Timer, Temporal Horn Pattern (aka certain beeping pattern)

Manufacturer : Allegro MicroSystem Inc.

Reported by Younji Choi, March 29, 2007


A5366CA. pdf



A5366CA can be applied in kitchen ideally, especially in apartments complex. The A5366CA is supplied in a low-cost 16-pin dual in-line plastic package. The lead (Pb) free version (suffix T), has 100% matte-tin leadframe plating. It is rated for continuous operation over the temperature range of -25C to +75C.


The A5366CA is a low-current BiCMOS circuit providing all of the required features for a photoelectric type smoke detector. This device can be used in conjunction with an infrared photoelectric chamber to sense scattered light from smoke particles. Special features are incorporated in the design to facilitate calibration and testing of the finished detector. The device is designed to comply with Underwriters Laboratories Specification UL217.

Ionization technology is generally for detecting small particles, which tend to be produced in greater amounts by flaming fires, which consume combustible materials rapidly and spread quickly. Sources of these fires may include paper burning in a wastebasket or a grease fire in the kitchen.

Photoelectric technology used in the sensor I'm reporting is generally for detecting large particles, which tend to be produced in greater amounts by smoldering fires, which may smolder for hours before bursting into flame. Sources of these fires may include cigarettes burning on couches or bedding.

All in all, I think this device is designed for detecting large particles of flames by using Photoelectric, however, I don't see big difference using two technology in specific ways in user's perspective. Because those two characteristics of fires each technology applies could happen at the house randomly, so my guess is that you can use any types of detector at home.


    * Interconnect Up to 50 Detectors
    * Piezoelectric Horn Driver
    * All Internal Low-Battery Detection
    * Power-ON Reset
    * Internal Timer and Control for Reduced Sensitivity
    * Built-In Circuits to Reduce False Triggering
    * 6 V to 12 V Operating Voltage Range
    * ESD-Protection Circuitry on All Pins
    * Temporal Horn Pattern


You may want to go to Allegro Micro directly, and order some free sample. You might have to sign in for further procedure.

price by manufacturer: not specified


A variable-gain photo amplifier can be directly interfaced to an infrared emitter/detector pair. The amplifier gain levels are determined by two external capacitors that are then internally selected depending on the operating mode. Low gain is selected during standby and timer modes. During a local alarm this low gain is increased (internally) by ~10% to reduce false triggering. High gain is used during the push-button test and during standby to periodically monitor the chamber sensitivity.

The internal oscillator and timing circuitry keeps standby power to a minimum by sensing for smoke every 10 seconds in a 10 ms window. A special three-stage speedup sensing scheme is incorporated to minimize the time to an audible alarm and also to reduce false triggering. Also, two consecutive cycles of degraded chamber sensitivity are required for a warning signal to occur.(from datasheet)


Datasheet explains fairly straight forward.


This data sheet has helpfully added the necessary components such as, capacitors and resistors to each output pin, so you don't have to figure out by yourself.

(image from the data sheet of A5366)

  1. C1 - output1, when pin15 is low
  2. C2 - output2, when pin15 is high
  3. DETECT - analog input to the photo amplifier
  4. STROBE - analog output strobed and regulated voltage of Vdd
  5. Vdd - the most positive supply Voltage(-0.5V to +15V)
  6. I red - output (to minimize noise, when the horn and LED output are active)
  7. I/O - disabled as an output. pin 1,2 are output as the
  8. HORN1
  9. HORN2
  10. FEEDBACK - conjunction with external passive components
  11. LED - output, low battery indicator.
  12. OSC. CAP. - forms terminal oscillator and sets the internal clock low time
  13. TIMING RES.- sets terminal oscillator and internal clock high time.
  14. Vss - Ground, the most negative supply potential
  15. HUSH - input for operation
  16. TEST


LED blinks under this condition below.

  1. Standby - pulses every 4.3 sec
  2. Local Smoke - pulses every 0.67 sec
  3. Remote Alarm - no pulse
  4. Test Mode - Pulses every 0.67 sec
  5. Timer Mode - Pulses 10 sec

Beeping loudness - 85 Decibel

Since this sensor works only with smoke, the only way to test it is to make smoke. My plan was to test with various sources of smoke. However, I wasn't easy as I thought because when I make smoke burning papers outside, a building instructor bared me from my test. So embarrassing.. =( I couldn't really get this test done as much as I wanted, so this information is pretty much relied on basic test and info in the instruction paper of smoke detector. But, I still looking for the chance to test in right way as I planed.

click to see the test video1

click to see the test video2

description: During an alarm, it beeps loudly 3 times and pause, 3 beeps pause. The LED blinks rapidly. Since this sensor is a photoelectric sensor, it's supposed to respond to cigarette fame, but didn't. I believe that if I burn a couch with cigarette smoldering, then it will work.


I thought that output pins keep sending values through this sensor to talk with other pins, and assuring there's no smoke particles in the air. So, I connected Arduino to pin 2 and 1 which is amplifier output getting signal from pin 15(HUSH) that is main part of this sensor to work this smoke detector. I expected to get stable values, however, it gave me very jumping value though I used 20uf capacitor. so I ended up to make metro high up to 150 to get average value in Max. I made Arduino talk to Max/MSP to 'LISTEN' the value, and visualizing as well. Because hearing is the primary part of this sensor. But no harm to have test viewer as well.

I used processing for viewer, and Max/MSP for sound generator.

click to see the value test video

Here's the sample Max patch.

click to downlaod


It actually didn't work with cigarette smoke, also with paper burning. I want to measure how much amount of smoke it needs to get it work. Plus, the value from pins were not different each other when it's high or low, unlike I expected it would be.