Wi-Fi Light Sensor

Stubits

Member
I am in need of a Wi-Fi light sensor. Our facility is quite large and we have an existing Wi-Fi network in place, so Wi-Fi sensors make the most sense. There are not many such sensors commercially available and those that I have been able to find simply sense any presence of light... I need to detect excess levels of light, not just regular light.

First, any suggestions for a commercially available battery operated Wi-Fi light sensor that meets these specs?

Assuming one is not available, I'd appreciate any suggestions as to how I might be able to cobble one together. We have had great luck using the Monnitt brand Wi-Fi sensors for other applications and thought it might be possible to retrofit one of them to function as a light sensor. In particular, They have a
Dry Contact Sensor
-Open/Close Sensor
Analog Voltage Sensor

Unfortunately I cannot yet pits any links to these devices.

Might any of these be able to be retrofit with a photo resistor? If so, what sort of photo cell/photo resistor would I need to use?

Any thoughts, suggestions, or recommendations you can make would be tremendously helpful.
 
I think you could do this with either the Monnit Dry Contact sensor or the Analog Voltage sensor.   You would need a photocell, an op amp and a handful of other components to make the light detection circuit.   I would probably lean towards the dry contact sensor, since the analog voltage sensor looks like it may be easily damaged if the voltage from the photocell circuit exceeds its input range.
 
Here's a simple photocell circuit that might work for your application.  You would get rid of the night light bulb and AC power connection on the relay outputs and just connect the relay contacts to the dry contact inputs and you'd be done.
 
Really not sure exactly what you mean by 'need to detect excess levels of light'.
 
Did you try just hooking up a photoresistor/1.2 volt dc supply in series (measuring across the photoresistor)?  Does the analog Monnit unit provide an output voltage?  If not, how do you plan on powering the sensor?  Batteries since you want 'wireless'?  Do you have power available in the areas you want to measure?  How accurate does the light detection need to be?
 
http://www.radioshack.com/product/index.jsp?productId=2062590
 
Based on the above answers, I would think your sensor selection might be the bigger part of your measurement problem.
 
Here played with this stuff indirectly in a warehouse environment mostly though relating to switching on and off IR lighting and what amount of light or no light it took to switch between the two.
 
I do have some Optex outdoor PIR's which have a piece just for the lighting and adjustments for the NO/NC piece relating to just the lighting.  It is a PITA to adjust though outdoors; indoors would be easy.  These are wired sensors but I believe there are proprietary wireless sensors.  There might might be now some 801.XX wireless sensors out there; haven't looked though.
 
I do see a few DIY's out there googling.
 
As you write above you could "cobble" together a sensitive light circuit (adjustable) and connect it to a sensitive (adjustable) debounce circuit then to the wireless 802.XX stuff and put it on a battery.  Last warehouse I was in thinking the side walls were a bit more than 30 feet; which is a bit high relating to mounting battery operated sensors; then though their was a mention of putting in HV double duplex boxes up high for a similar purpose such that you could have wireless using a wired power source. (POE would work cuz its only one wire for network and power).   Thinking we ran cabling for the warehouse endeavor mentioned in the first paragraph of this post.
 
BraveSirRobbin said:
Did you try just hooking up a photoresistor/1.2 volt dc supply in series (measuring across the photoresistor)?  Does the analog Monnit unit provide an output voltage?  If not, how do you plan on powering the sensor?  Batteries since you want 'wireless'?  Do you have power available in the areas you want to measure?  How accurate does the light detection need to be?
 
The analog sensor does not supply any voltage.  It only measures the voltage present on the input, and has a pretty narrow range. 
 
Your idea of using a battery and a photocell might work with the analog sensor, but one problem I see with that approach is that it provides no hysteresis.  Unless the purpose is only to sense a large change in light that happens quickly, I think having some hysteresis in the circuit is necessary.
 
Also, depending on the normal light level and the resistance of the photocell, it could discharge the battery in a relatively short time.
 
Ral;
 
The analog sensor does not supply any voltage.  It only measures the voltage present on the input, and has a pretty narrow range.
I was thinking this 'may' be the case, which is why I asked if a power source was available.  Some analog measurement devices DO provide a DC supply, some you can even excite the supply, take a reading, then turn off the excitement.  Yes, the narrow range is there but you are assuming the OP want's just 'one' light level threshold to detect, then close/open contacts based on that (I did not get that reading the post).
 
Your idea of using a battery and a photocell might work with the analog sensor, but one problem I see with that approach is that it provides no hysteresis
This would have to be accomplished in the code.  Again, there may be other levels that need to be measured ;)
 
Wow, there's lots of great info here. I am open to any feasible solutions, doesn't have to be Monnitt. I am partial to using Wifi because we already have the network in place.

What we are looking to do is detect the strobe light associated with a fire alarm. The strobe is obviously much brighter than any ambient light in the space, but the space will have regular office lighting as a norm. Of course it would be better to tap directly into the alarm system, but that's just not an option here... We are located in a massive government building, so just trying to get creative.

Thoughts?
 
I'm not aware of any off-the-shelf WiFi sensors other than the Monnit units.   One thing to think about if you were to use the analog sensor is that it appears to have an adjustable sample time (what they call "assessments per heartbeat") as well as a sample interval (the heartbeat time).  It's not clear from the data sheet what the minimum settings are for these (the example shows a heartbeat in the minutes range).   So one question is whether you could sample quickly enough to capture the very fast flash of a strobe light.   Even if it could make several samples per second, strobe light flashes typically last less than a thousandth of a second.  So all the samples could still miss the flash.
 
Also, each heartbeat results in a packet being sent on the network, which uses power and ultimately reduces the Monnit battery life.
 
So this would make me more inclined to use the dry contact sensor with a separate photocell circuit that could capture the flash of the strobe and hold the relay closed long enough to allow the Monnit unit to detect the incident.
 
So I'm wondering... why can't the fire alarm control panel (or the company that monitors it) give you the notification you need rather than you designing your own method to detect that the alarm is going off?
 
Yup; wondering same thing here.  There typically is a sound device that also triggers with the strobe; eh?
 
Is this just a personal "quicker than the eye" project?
 
Stubits said:
What we are looking to do is detect the strobe light associated with a fire alarm. The strobe is obviously much brighter than any ambient light in the space, but the space will have regular office lighting as a norm. Of course it would be better to tap directly into the alarm system, but that's just not an option here... We are located in a massive government building, so just trying to get creative.

Thoughts?
This is an enormous can of worms you just opened!
 
I'm not trying to sound like a kill joy but why do you need this fire system strobe monitored?  What actions will be done?  Is this life saving?
 
BTW Subits,
 
Welcome to the Cocoontech Forum.  Lots of friendly folks here.
 
Well a strobe is pretty quick. (and a bit of a tangent here).
 
But yeah you can have the issue of: (don't look at the strobe light while testing)
 
Photosensitive epilepsy (PSE).  It is a form of eplepsy in which seizures are triggered by visual stimuli that form patterns in time or space.
 
That said many folks can have epilepsy and not even know it until it shows up in seizure activity.  (like a little baby Jacksonian seizure)
 
Typically it is said that the "roots" of said stuff (epilepsy) sometimes comes from a foci to an old brain injury causing a willy nilly short circuit in the brain. (geez could even be a mild stroke/anueurism that caused some brain tissue injury)
 
Still wondering though about the why of your endeavor?
 
We are a relatively small tenant in a very large building. Our space includes a six floor open atrium area. Due to the size of the building we utilized a zoned evacuation approach wherein only three floors are evacuated at a time. The problem we have is that because of the open atrium when the alarms go off, they are heard loud and clear throughout all six of our floors, not just on the three that need to evacuate. In order for our emergency team to effect a calm and orderly evacuation, we need to know which 3 out of the six floors are being evacuated. Presently, the only way to do this is for a team member to run up 6 flights of stairs to check to strobes.

We are the only tenant in the building with such a atrium space, as such, the building management see no need to go through the hassle of allowing up access to or tying into their system.

This is not meant to be life saving, just informative and time saving.

As such, would appreciate any thoughts/suggestions.

Thanks!
 
Presently, the only way to do this is for a team member to run up 6 flights of stairs to check to strobes.
 
Kind of went through these stuff in the moving into a new building work environment.  I just watched and listened relating to the documented approach.
 
There were configured sectional or quadrant emergency team leads (well volunteers).  These emergency team leads in a building "auto" communicated with the other team leads during an evacuation procedure for an orderly evacuation.  There were building seating / office maps relating to VOIP telephones and where they were in the building (desk / conference room / office stuff) concurrently relating to the other stuff mentioned. 
 
Building management provided the means and requested the tenants to follow a process based on what they provided.  There was a follow up checks and balances to said procedures that came from the building management itself.  There were no choices or whom or what in the building other than everybody.  IE: the non placement say of a strobe where there were people would be an issue probably related to the original construction (well code). 
 
Sometimes that happens after new office space is put in where there was no office space or people and it is a detrimental life safety thing oversight.
 
All tenants of said building whatever (company or subcontracted) were aware and participated in said endeavors.
 
It shouldn't be anything that a particular group / subgroup of individuals in a building should have to worry about unless there happens to be a concern where evacuation takes place and something bad happens.
 
Ok folks, first of all any thanks for the interest in the topic.

Second, I agree that this is a bit of an unorthodox project, and maybe it won't work or help solve the problem, but I would really love the help of the group to at least give it a shot.

So, a few specific questions...

1) do you suggest that I use the Analog Voltage Sensor or the Dry Contact Sensor?

2) how specifically would I wire up the photocell to the sensor, I assume it is MUCH me complicated than that.

3) I know the photocell circuit will need power, is there any chance we can "steal" power from the sensor itself, even if it reduces battery life. The sensor monitors battery life and alerts when running low, so would be good to have them tied together.

Any help on these issues would be great!
 
Back
Top