A naked flame is a complex soup of ionised gases, that possesses an unexpected property. As you might expect with that much ionisation there is some level of electrical conductivity, but the unusual property comes in that a flame can be made to conduct in only one direction. In other words, it can become a diode of sorts, in a manner reminiscent of a vacuum tube diode.
[Paul Stoffregen] has made use of this phenomenon in a flame detector that he’s built to be installed on a Burning Man flame-based art installation. It forms part of a response to a problem with traditional pilot lights: when the wind blows a pilot light out, a cloud of unignited gas can accumulate. The sensor allows the pilot light to be automatically re-ignited if the flame is no longer present.
The circuit is a surprisingly simple one, with a PNP transistor being turned on by the flame diode being placed in its base circuit. This allows the intensity of the flame to be measured as well as whether or not it is present, and all at the expense of a microscopic current consumption. A capacitor is charged by the transistor, and the charge time is measured by a Teensy that uses it to estimate flame intensity and trigger the pilot light if necessary. Interestingly it comes from a patent that expired in 2013, it’s always worth including that particular line of research in your investigations.
All the construction details are in the page linked above, and you can see the system under test in the video below the break.
We’ve looked at this property of a flame before, a project in which someone made a functioning flame triode.
Is the polarity to do with the moving particles of burning fuel? So the movement of the particles carries, say, positive ions, so whichever electrode is furthest along the flame is the anode?
Would be fairly easy to test, you’d just need a flame and 2 movable electrodes. Might try it out some time. Distance might have something to do with the voltage drop, or resistance, or both.
Just a thought: Would the moving ions in a flame be able to produce a current? If you connect a wire to the nozzle and have the flame pass through a grating connected to a different wire, would there be a potential between them? It should be possible to make the gas flow at very high speed, supersonic, to maximize that effect. That would make the simplest combustion engine/generator possible!
I think probably not. This is more about letting voltage travel downhill, than pushing it uphill against it’s nature. Sort of thing.
Maybe if you stuck a magnetic field in there somewhere. I’m thinking about ion engines and the right-hand rule.
I thought about doing something like this, but with a starter built in so I could more automated propane home brewing.
If you’re smart, you can use the same contacts to send a high-voltage spark down, to re-light the flame.
I had this idea too, but it has a drawback: The sense circuit will be exposed to the high voltage ignition pulse. So you need some high value resistors with pulse rating in your sense line. And you can not connect the electrode directly the secondary of an ignition transformer, you need a switch or spark gap.
How about this hack?: Use this–older than dirt–driving an MOS device–
[from a building-supply ‘warehouse’]–
“Honeywell CQ100A10 Replacement Thermocouple for Gas Furnaces, Boilers, Water Heaters and Room Heaters
by Honeywell
Price
$9.99”
…or is this not complicated enough…?
I would come up with a detailed reply as to why you’re comment is out of place here, but you do have a not-unreasonable suggestion with that part; however, that device does not demonstrate the principals of a flame diode, which is part of what makes this post interesting. People do this sort of thing because they enjoy it, so no need to crap on other people’s fun. Get out (or stay in) and enjoy your life instead of trololloling internet forums. Unless that’s how you get your kicks, in which case, you should seek some kind of therapy. That is all.
Why would one say, “I would come up with a detailed reply…”, and then not “…come up with a detailed reply “? Certainly NOT because you have nothing to offer, as opposed to my insufferable ‘trololloling’?
You cannot know how much
Iwe appreciate your pointing out the totally non-obvious (to me) fact that a thermocouple is not a flame diode. We are all the much better for that. Thank you.Don’t be shy; don’t hold back…this is your opportunity to shine: let’s have your DETAILED reply to my comment. Then let’s have your hack . We’re all waiting…
There was a BUT in his first part. As for the fairly huge difference between a diode and a thermocouple, do you know anything at all about diodes or thermocouples? I think just about everyone here knows what a diode does. And you certainly post here enough. It’s the very basics of electronics.
There’s also the bit about ions being carried in the article, which you MUST have read. How could that make you think of thermocouples?
Your patronising and insincere “thank you” is long past when you should’ve stopped digging (making a hole in the ground using a spade). You posted one slightly snotty, dismissive criticism, and then got caught out not knowing what the flip you were talking about. Humble apologies (sincere ones) are probably the best way to go from there, Jawn.
Then your finale, getting pissy, wanting a “detailed reply” (why? there’s reams about diodes on the Internet, they’ve been writing about them since something like 1910) as if this is some obscure subject that nobody REALLY knows about, and dude is just bluffing to make you look stupid.
Finally, a hack of your own isn’t a necessary appendage to every post here. As indeed you know, from all your own posts. If you want to start a trend, I’d probably see if the local upper-school has any places going in physics lessons, cos they’ll teach you about diodes, and probably a few useful other things. Then your electronic hacks might be successful. You can’t be missing such a basic subject just by fluke. Diodes are a fundamental part of electronics knowledge. If they didn’t teach them, anyone who got any distance in electronics study would ask about them spontaneously.
@jawnhenry – Okay, you win the “I coulda dunnit better” naysayer prize.
“How about this hack?: Use this–older than dirt–driving an MOS device–”
Actually, given the inherent physical nature of a thermocouple, a BJT
might beis probably a better choice. Whatever you choose, be careful of offending anyone…Most ignition control modules for furnaces, boilers and commercial ovens actually use this very method it is called flame rectification. Most controllers use about 9-12 volts and pass about 5-8 microamps when properly adjusted. The advantage of this method is that it is reliability and response time. Thermocouples and thermopiles take a bit of time before providing feedback.
Oil burners often have an optical flame detection with a special photocell or LDR. It’s said that it is mostly sensitive to UV, which is produced by the flame.
(im sure your aware but others may not be)but thats no use here as its to be used outdoors. When experimenting with waste oil burners out of there housing this proved to be pretty annoying as it took very little ambient light to trip the sensor.
This idea is widely used in most popular detectors for gas chromatography, namely FID (Flame ionization detector). https://en.wikipedia.org/wiki/Flame_ionization_detector. I always wanted to actually see the flame in flame detector but burning pure hydrogen is invisible. Normally you use bench-top hydrogen generator for you GC set-up but I remember having in our lab big pressurised cylinder full of H2. That made everyone around super nervous ;)
So…
Can you combine two flames to make a valve-like device? Can you build a 555 from flame-based technology?
you can definitely make a an oscillator…
https://www.youtube.com/watch?v=-aAwyUoawkc
I was considering a flame based bridge rectifier for tesla coils.
A flame diode should be able to support quite large voltages.
I like the idea of a two “transistor” flame diode square wave or relaxation oscillator. It could make for a good driver. I wonder how high I could push the frequency.
Maybe use a flame tube with multiple tapped outputs to convert frequency to analogue or make a sonic scanning shift register. By incorporating the two into one device.
Propane and candle powered 8bit Flame CPUs are coming.
About the same amount of technology needed to light a flame only on demand as to monitor a pilot light for continual flame.
Here is one already built. https://store.qkits.com/flame-detector-module-mxa042.html
Why do you need pilot flame if you have means to electricaly re-ignite it anyway? In such case you can use electric ignition directly instead of pilot flame…