Wi-Fi Light Sensor

[BraveSirRobbin]

I guess my first attempt would be to use a circuit like this that I built to detect an appliance light going on.
 
Instead of using a DS10a, I would have the transistor trip the trigger input on something like an Elk-960 one shot relay, and monitor its contact closures, since you are detecting a fast transient (that 960 can be adjusted to close for a few seconds or so when it's triggered).
 
Some mentionables:
 
The CdS photoresistor might not be sensitive enough to discern between the normal light background and the strobe.  It 'shoud' be fast enough as I believe they can detect in the many 10's or 20's milliseconds light pulse.
 
Power could be a problem, but in theory, there should be little 'stand by' draw.  Best to measure the stand by current after the circuit is finalized.  Hopefully something like a small 12 volt or so battery would suffice.
 
Some may suggest using a photo-transistor in lieu of a CdS sensor.  Problem is, they are more sensitive towards the red/IR and I think the strobe is probably more centered around the mid 400's or 500's nanometer wavelength (again not sure).
 
This is something that will require a bit of R&D with some trial and error testing.

 

[Stubits]

Just another thought, would it make more sense to try and detect the alarm sound, or eve the vibrations of the alarm speaker? Would either be easier than light?

 

[Stubits]

BraveSirRobbin- that diagram is excellent. Exactly what I needed.

Would you suggest I use the dry contact sensor or the analog voltage sensor?

 

[BraveSirRobbin]

Sound would be easier.

 

[BraveSirRobbin]

BraveSirRobbin- that diagram is excellent. Exactly what I needed.

Would you suggest I use the dry contact sensor or the analog voltage sensor?

As stated in the post, you would have the Elk-960 provide the contact closure via its relay.
 
I would really investigate the sound detection though as it should be a lot easier.
 
How much vibration is there on your alarms?  I have this tiny vibration detector that provides a contact closure that would make this a really simple project if the vibrations were sufficient to trigger it.

 

[BraveSirRobbin]

Check out this vibration sensor that I suggested for a dog to trigger (if it were inside a tennis ball) with his nose.
 
http://board.homeseer.com/showpost.php?p=1114173&postcount=12

 

[Stubits]

Thanks for all the help, lots of options!!

In my mind, the light seems to make more sense. I am not sure I fully understand the sound detector. For that to work, it would need to be able to distinguish between an alarm sounding close to it and an alarm sounding on another floor... Cant figure out how to do that. Would it need a separate relay as well, like the elk?

Regarding the vibration, it's the same as any speaker. My though would be to mount it right on top of it. Might that work?

 

[RAL]

I agree with BSR.   Sound or vibration will be much easier to detect.  
 
I've been digging into the light detection a bit more, and the problem I come up against is that the pulse of the strobe light is so very short, and a CdS photocell won't be fast enough to pick it up.   You would need to at least use a phototransistor instead, and even that might be difficult to get to work.

 

[RAL]

Thanks for all the help, lots of options!!

In my mind, the light seems to make more sense. I am not sure I fully understand the sound detector. For that to work, it would need to be able to distinguish between an alarm sounding close to it and an alarm sounding on another floor... Cant figure out how to do that. Would it need a separate relay as well, like the elk?

Regarding the vibration, it's the same as any speaker. My though would be to mount it right on top of it. Might that work?

 
The sound detector doesn't need to be very sophisticated.  It doesn't need to care about the sound frequency, just the volume level. The fire alarm horns are damn loud, so all you need to do it place the detector near the horn and set the trigger threshold high enough so that it only triggers from the near-by horn and not the others.
 
For the vibration sensor, yes - mount it right on the horn.   The main question will be whether the sound frequency makes it vibrate sufficiently to trigger.

 

[Stubits]

So the Elk is just to account for how rapidly the strobe flashes? Would it prevent the system from sending alert after alert after alert as the strobe cycles? Would it be worth trying it without the Elk first, to see if it works more simply?

Does it really need 12volt battery, your solution of tying into the sensors batteries was great, I'd like to do that if I can.

Should I use the dry contact sensor? Would it be better to use an open/close sensor or a magnetic sensor more like what you used?

You mention adjusting the sensitivity from the circuit, how does that work?

Thanks!

 

[Stubits]

Sound is great, if it will work better. Will it work directly with the dry contact sensor or will I need a relay?

 

[RAL]

If you use the Sparkfun sound sensor, you will need to connect the Gate signal from the circuit board to a relay (or the ELK960) that in turn triggers the WiFi dry contact sensor.   The Sparkfun board can only source or sink a small amount of current, (20-40mA and 10-20mA, respectively).  This relay might do the job.
 
If this is a standard fire alarm, it will be sounding in a temporal pattern, which will give you multiple pulses from the sound detector and might result in the Monnit sending multiple alerts.  For this reason, the ELK960 is attractive, since you can control the output contact closure time.  You can drive the input of the 960 directly from the sound board Gate signal.
 
Since you are trying to detect only loud sounds, you will also need to add a resistor or trim pot to R17 on the sound detector to adjust the sensitivity.  You'll have to experiment a bit to determine the value that works best.

 

[Stubits]

Good news is that the Monnitt software has a feature to only send the alert once , so not sure we need to handle that through the relay. With that said, do you think the Arduino relay is the best option? What should I look for in a relay?

What is a "trim pot"?

The sensor is powered by 2 AA batteries, do you think I will be able to power the sound detector circuitry using the sensors batteries the same way BraveSirRobin did with his light sensor?

This actually does seem pretty simple to pull together. Thanks!!

 

[RAL]

The sound board needs a minimum of 3V to operate, so if you try to run it from the batteries in the Monnit, it would be at the very lower limit of where it can operate.  You would also need a 3V relay, like this one, rather than a 5V relay.  My concern would be that as the batteries age even slightly, their voltage would drop below 3V and things might not work.  In my opinion, you'd be better off running it with a 5V power supply.   On the other hand, the parts are inexpensive enough that you could give it a try and see what happens.
 
A trim pot it just a variable, adjustable resistor, like these.  Since it's hard to say what value you will need, getting an assortment is a good idea.   You'll need to solder this to the sound detector board.

 

[Stubits]

Thanks! I was hoping I would be able to power the sound detector using the same batteries as the sensor. These will be located in some out of the way locals in our facility and the sensor has the ability to alert when battery is running low. I am afraid with separate batteries for the sound detector we could end up with a situation where the sensor is live but not the sound detector and the setup will fail.

Any suggestions, possible solutions? Are there any lower voltage sound detectors out there?

Thanks!