Is there a way I could tie this in with a simple circuit to activate a series of relays when it senses flame? I am building a small turret that will first seek brightness (in a dark room, so for example, a candle) and then when the flame sensor sensor is facing it, activate a series of relays, some ceasing motion of the 'turret' and one triggering a pump which will squirt a spray of water in the direction of the candle or other type of flame. The distance of the flame from the sensor is required because I want to be able to move the candle and still have the turret track a different location.
formatting link
If possible, I would also like to have the sensor indicate once the flame is extinguished, and reactivate the turret. This project is similar to the firefighting robots competition, only the turret is stationary.
But they may not be sensitive enough to detect a single candle flame in a dark room (CdS sensors like that are usually used inside of furnaces, pointed right at a bright yellow flame less than a foot away).
And CdS sensors are slooooooow.
So you would have to amplify their output, you'd have to use at least two (to sense the direction to the candle), and you'd have to _really_ slow your application down.
Why not use a camera and some simple microprocessor signal processor, instead of 1940's sensors driving 1940's electronics? You can get web cams these days for a pittance, USB has been around long enough that you should be able to get an old slow laptop for cheap, and if you don't like that solution you can get all the bits to build your own camera for fairly cheap.
On the other hand, dual CdS sensors could be used like that old Carl and Jerry episode from Popular Electronics back in the '60s to track a moving flashlight. (Boy, am I showing my age!) The idea was that the right sensor activated the left wheel of the robot, and the left sensor activated the right wheel. If the light was straight ahead, the robot charged straight at it; otherwise, it turned until both sensors were equally illuminated.
I always wanted to build one of these, but never got a round to it...
If you just want to move the turret and not the whole robot, you'd need some logic to stop the motors when both sensors were equally lit.
Best regards,
Bob Masta DAQARTA v4.51 Data AcQuisition And Real-Time Analysis
formatting link
Scope, Spectrum, Spectrogram, Sound Level Meter FREE Signal Generator Science with your sound card!
Oh well... I had thought that they sensed flame based on heat, not light. I had previously sonsidered using an IR sensor to detect the flame but was informed at how difficult it would be to accurately detect flame, and how expensive it would get. ): Do you have any alternative suggestions to detect the flame? I had planned on using light sensors I have salvaged to run two sets of motors. One controlling X axis movement and the other Y. So basically all I need is something that can detect a flame fairly close to directly ahead of it...but a range of more than a foot would be nice...thanks for all previous and future submissions!
If you have access to OLD broken? cd players, they have interesting photosensors. I have found one with ~8 mm (1/3 inch) four quadrant cell, and if you put a badly focushed picture of the flame on it, and put differential amplifiers between the horizontal pair and the vertical pair, you can sense the direction of the flame in two dimensions.
Use a piece if infrared filter from a broken remote control, to stop visible light.
One place to start would be browsing the archives of the comp.robotics.misc newsgroup. I seem to recall the use of UV sensors to distinguish "flame" from "light", but I don't remember the exact details.
If you'd like a rapid-acting non-contact thermal (IR) sensor, check out the Melexis MLX90614 from:
Melexis Microelectronic Systems
formatting link
If you think that the MLX90614 would be useful, you can purchase a serial-interface module from Parallax:
Parallax
formatting link
Hope this helps.
Frank McKenney
-- To invent, you need a good imagination and a pile of junk. -- Thomas Alva Edison
-- Frank McKenney, McKenney Associates Richmond, Virginia / (804) 320-4887 Munged E-mail: frank uscore mckenney ayut mined spring dawt cahm (y'all)
A combo of red and green acrylic / "plexiglas" sheets passes most IR that silicon photosensors sense, and blocks visible rather well.
Keep in mind that a yellow flame has ratio of IR to visible merely an order of magnitude (or somewhat less) above that of incandescent lighting and incandescent flashlights. Also consider that sunlight and daylight have a significant amount of such infrared, especially direct sunlight.
Do not plan on achieving something that reliably hoses candles and reliably does not hose incandescent flashlights with an engineering effort short of or possibly even comparable to "college/university engineering-major senior design project", which is something done by a group rather than by an individual.
There are flame detectors that sense 310 nanometer UV, which is a UVB wavelength close to borderline UVA, that blue flames produce weakly. Small flames usually have a blue region producing some of this wavelength.
Keep in mind that this wavelength is only weakly produced by flames, and flame sensors based on this wavelength assume lack of this wavelength from normal light sources. However, many fluorescent lamps can produce enough of this wavelength (even when not to an extent problematic to humans) to confuse flame sensors based on this wavelength. A few incandescents, especially some halogens, can do likewise. Daylight also usually has enough of this wavelength to trigger flame sensors based on this wavelength. There are common filters to block this wavelength from "legitimate light sources" to allow such light sources to be used where flame sensors using this wavelength are used, including most polycarbonate plastic sheet materials such as "lexan".
There are also flame sensors based on a very deep UV wavelength of hydrogen, 121.6 nm - which is a "vacuum UV" wavelength that does not get through air well (nor most other tyranmsparent materials). It works as a detector of flames, preferably blue flame such as flame bases, only in very close proximity. This wavelength is only very weakly emitted by blue flames, and flame detectors based on this wavelength assume lack of such wavelength from other sources. Sensors for this wavelength may also detect radioactivity and X-rays.
Please keep in mind that many, probably most yellow flames are far short of "full thermal radiators".
I just tried my "Raytek" non-contact thermometer on a 6 inch diameter pan with a shallow layer of 91% isopropyl alcohol ignited by a He-Ne laser power supply. The fire was merely somewhat small and just a little on the dim side for a fireplace, and the flames had temperature readings almost entirely less than 150 degrees C, mostly less than 90 degrees C.
Hm. On re-reading, it seems that the OP's exact words were "candle or other type of flame". I got sidetracked by the reference to the Trinity Firefighting contests.
Ouch. Daylight and incandescents produce all sorts of wavelengths. I remember watching my carefully-crafted IR-interrupter wheel encoders become erratic and then fail completely the first time they hit direct sunlight.
I had to re-read this a few times before I realized what you were referring to: you can use such a sensor to detect flames, but only if you make sure the environment doesn't have any stray 310nm UV. And if people complain about stumbling around in the dark (silly creatures ), well, you have to make sure none of the squirtgun-triggering wavelengths accompany the room lighting.
Wouldn't work for Trinity, of course. Too uncontrolled.
S'funny. You get into the habit of thinking the human eye is the only wideband optical sensor, and it catches you by surprise when you run across another.
Your flames sounded like they might be a lot bluer than a candle's (I was still thinking Trinity), so ran my own tests using a MicroTech MT100 (pocket-sized, 8:1 dispersion):
Birthday candle at 1" 117-178degF Birthday candle at 2" 77-93degF Open mouth at 1" 89degF
So... more light than heat. And definitely not much help if you want to detect the flame before it melts your sensor.
Then I re-(re-)read your words abuot "full thermal radiators" and went back and ran another test. This time I held a table knife about 1/8" over the candle flame
Soot-covered table knife at 1" "Hi" Soot-covered table knife at 4" 103degF
In other words, the OP can use a noncontact thermometer as a flame detector as long as he first places a more effective radiator over the flame... which shouldn't be too hard, once he has located it.
Which leaves... what? You've described a UV sensor for controlled environments, but what characteristics of a flame could we use in a (relatively) uncontrolled space? Firefighting isn't the only application I can think of that could use such a sensor; what about (say) placing welding rod within an arc, or guiding a piece of glass into the center of a bunsen burner flame?
I could believe that a given flame had a unique IR-visible-UV spectrum, but that might be hard to obtain quickly, and you really want to classify "flame" vs. "non-flame", not scan a list of umpteen thousand spectra to see if you get a match.
How about variability? A candle flame certainly varies in color and intensity, and combustion doesn't take place at lightspeed, so maybe there are intensity or spectral changes that could be used.
Oh, well. We're pretty well eliminated the Melexis MLX90614 and friends; I guess that's anough accomplishment for a Monday morning.
Interesting bit of trivia: a "birthday candle" fits nicely between the tines of an ordinary table fork. This is very handy to know for those rare cases when no birthday cake is available.
Thanks again, Don.
Frank
-- ...the fact that some geniuses were laughed at does not imply that all who are laughed at are geniuses. They laughed at Columbus, they laughed at Fulton, they laughed at the Wright Brothers. But they also laughed at Bozo the Clown. -- Carl Sagan
-- Frank McKenney, McKenney Associates Richmond, Virginia / (804) 320-4887 Munged E-mail: frank uscore mckenney ayut mined spring dawt cahm (y'all)
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.