How can I detect the flame under the sunlight?

Thread Starter

ghostmanzero

Joined May 16, 2017
10
Hi everyone, I'm working on a flame sensor project. Ive seen many sensors that can detect flame even under the sunlight. Those were IR-UV sensors.what is the principle of these sensors. I tried both of those sensors but it didn't work for me. How can I detect the flame and disable the sunlight's effect. Those sensors are working by detecting the ir and uv waves which are coming from the flame. But as you know sun is the biggest ir and uv wave source. So all sensors that I've tried ended up with fail. I need your suggestions and advice. Can you help me please?
 

wayneh

Joined Sep 9, 2010
17,496
Hi everyone, I'm working on a flame sensor project. Ive seen many sensors that can detect flame even under the sunlight. Those were IR-UV sensors.what is the principle of these sensors. I tried both of those sensors but it didn't work for me. How can I detect the flame and disable the sunlight's effect. Those sensors are working by detecting the ir and uv waves which are coming from the flame. But as you know sun is the biggest ir and uv wave source. So all sensors that I've tried ended up with fail. I need your suggestions and advice. Can you help me please?
In what context are you looking for the flame? Is it possible to use an emitter and a detector with the flame in the beam in between? Or do you want to “see” a flame within the sensor’s field of view?
 

Janis59

Joined Aug 21, 2017
1,834
1.st way: to use the thermocouple in the zone of expected flame. This is used at almost all the kitchen gas-cookers.
2.st way: to detect the intense long wave IR spectra. Sunlight has ca 6000C but flame ca 1000 C. therefore the frequency will be shifted hardly, and surplus to that, Sun constant is 1,38 kW/m2 even in near Space, at Planet surface it is sooner betwen 300 and 500 W/m2 at hot sunny day, however at flame the IR flux is hundreds of times stronger. The IR narrow banwidth diodes is the best sensor type. Sorry to tell, them all cost around 100 USD or more. Other method is to use an narrow band interference filter and bolometer type sensor, what costs only hard tens of USD. Probably You may shift away even those zone filter and rely only on microbolometer signal alone.
 

Thread Starter

ghostmanzero

Joined May 16, 2017
10
In what context are you looking for the flame? Is it possible to use an emitter and a detector with the flame in the beam in between? Or do you want to “see” a flame within the sensor’s field of view?
It will like detecting the a lighter's flame in a room.but of course this room has windows. I used some cheap ir flame detectors and they worked very well if there is no sunlight. But they completely blind under the sunlight.
 

Reloadron

Joined Jan 15, 2015
7,501
If I were you I would likely look at the several types of flame detection currently in use, the advantages and disadvantages of each and then decide which method best suited my needs. Flame Conductivity VS Flame Rectification Systems is one method I have used on a few occasions over the years primarily with boiler and furnace systems. Less looking at your project in detail it is difficult to say what would work the best.

<EDIT> I saw your latter post so obviously Flame Conductivity is not an option. Infra red would be an option but distinguishing body heat from the flame of a cigarette lighter gets tricky. It can be done but not in a simple and easy process. </EDIT>

Ron
 
Last edited:

Thread Starter

ghostmanzero

Joined May 16, 2017
10
You also need to look for the flame flicker in the 1Hz - 20Hz range.
Thank you very much. I read some projects about this. But 1hz-20hz range is very slow.in this case I have to measure twice everything to eliminate the mistakes. The system must be very fast. (unfortunately reaction time must be less than 5-10 ms. :D)
 

Thread Starter

ghostmanzero

Joined May 16, 2017
10
1.st way: to use the thermocouple in the zone of expected flame. This is used at almost all the kitchen gas-cookers.
2.st way: to detect the intense long wave IR spectra. Sunlight has ca 6000C but flame ca 1000 C. therefore the frequency will be shifted hardly, and surplus to that, Sun constant is 1,38 kW/m2 even in near Space, at Planet surface it is sooner betwen 300 and 500 W/m2 at hot sunny day, however at flame the IR flux is hundreds of times stronger. The IR narrow banwidth diodes is the best sensor type. Sorry to tell, them all cost around 100 USD or more. Other method is to use an narrow band interference filter and bolometer type sensor, what costs only hard tens of USD. Probably You may shift away even those zone filter and rely only on microbolometer signal alone.
Thank you. One of these methods can be the solution I think. I will work on it.
 

wayneh

Joined Sep 9, 2010
17,496
It will like detecting the a lighter's flame in a room.but of course this room has windows. I used some cheap ir flame detectors and they worked very well if there is no sunlight. But they completely blind under the sunlight.
The glass of the windows should block long IR enough to give a good signal from the flame versus the noise of the sunlight. You "just" need to filter out what the glass lets through or use a detector that is less sensitive to the wavelengths coming through the glass. Not all glass windows are the same, though. Some high-efficiency windows are very good at blocking IR.
 
Top