Electrical conductivity of flames (OT?)

[George Herold]

Hi guys, A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work. The sensors are just a
metal rod that sit’s in the flame. They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive. Now here’s the weird part. The flame sensor shows
rectification and so only has to sense a DC current. I’m totally
clueless as to how you get rectification. If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you’ll see he talks about different areas being important. But no
other explanation. Anyone have an idea of what’s going on?

Thanks
George H.

 

[George Herold]

<swag> Similar to the effects of a heated cathode in a triode tube?
The metal rod is, I'd guess, much hotter than the nozzle so more free
electrons leading to a rectification effect. </swag>

If you download the pdf of the patent and look at fig 1. it shows the
direction of the diode... from the probe to the grounded nozzle. I
also get the impression that it works fairly fast (a second or less.)
which hardly seems like enough time for the probe to heat up. If one
assumes that all the current is carried by the electrons in the
plasma, then you see more current when the electrons are moving in the
same direction as the mass flow in the flame.... maybe that is
important?

I don't see how the relative areas matter though.

George H.

 

[tm]

Hi guys, A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work. The sensors are just a
metal rod that sit’s in the flame. They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive. Now here’s the weird part. The flame sensor shows
rectification and so only has to sense a DC current. I’m totally
clueless as to how you get rectification. If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you’ll see he talks about different areas being important. But no
other explanation. Anyone have an idea of what’s going on?

Thanks
George H.


If he lightly sands off the oxide coating on the rod, it will work again for
six to 12 months. The rod gets an insulating layer formed by the combustion
products in the flame. Think dust, etc.

I have the same issue on my gas water heater and that is how I get around
it.

tm

 

[George Herold]

Hi guys,  A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work.  The sensors are just a
metal rod that sit’s in the flame.  They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive.  Now here’s the weird part.  The flame sensor shows
rectification and so only has to sense a DC current.  I’m totally
clueless as to how you get rectification.   If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you’ll see he talks about different areas being important.  But no
other explanation.   Anyone have an idea of what’s going on?

Thanks
George H.

If he lightly sands off the oxide coating on the rod, it will work again for
six to 12 months. The rod gets an insulating layer formed by the combustion
products in the flame. Think dust, etc.

I have the same issue on my gas water heater and that is how I get around
it.

Yup, he cleaned the rod and it's working again. It looks like a
yearly cleaning of the rod is now recommended maintenance.

George H.

 

[Adrian Tuddenham]

Hi guys, A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work. The sensors are just a
metal rod that sit's in the flame. They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive. Now here's the weird part. The flame sensor shows
rectification and so only has to sense a DC current. I'm totally
clueless as to how you get rectification. If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you'll see he talks about different areas being important. But no
other explanation. Anyone have an idea of what's going on?

Thanks
George H.


If he lightly sands off the oxide coating on the rod, it will work again for
six to 12 months. The rod gets an insulating layer formed by the combustion
products in the flame. Think dust, etc.

I have the same issue on my gas water heater and that is how I get around
it.

I was recently called out to repair a gas boiler system, ostensibly with
a failed flame-conducting sensor. It turned out that the sensor worked
from live to earth and someone had reversed the live and neutral at the
supply meter tails a few days before. Every electrical appliance in the
house was live when it was switched off.

 

[Robert Baer]

Hi guys, A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work. The sensors are just a
metal rod that sit’s in the flame. They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive. Now here’s the weird part. The flame sensor shows
rectification and so only has to sense a DC current. I’m totally
clueless as to how you get rectification. If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you’ll see he talks about different areas being important. But no
other explanation. Anyone have an idea of what’s going on?

Thanks
George H.

Simple.
Think of a vacuum tube, where a heated element emits electrons that
are collected elsewhere.
Collection areas are different and so capability, control, etc vary -
and according to relative bias.

 

[Robert Baer]

Hi guys, A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work. The sensors are just a
metal rod that sit’s in the flame. They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive. Now here’s the weird part. The flame sensor shows
rectification and so only has to sense a DC current. I’m totally
clueless as to how you get rectification. If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you’ll see he talks about different areas being important. But no
other explanation. Anyone have an idea of what’s going on?

Thanks
George H.


If he lightly sands off the oxide coating on the rod, it will work again
for six to 12 months. The rod gets an insulating layer formed by the
combustion products in the flame. Think dust, etc.

I have the same issue on my gas water heater and that is how I get
around it.

tm

" sit’s "?????????????????????????????

 

[[email protected]]

If you download the pdf of the patent and look at fig 1. it shows the
direction of the diode... from the probe to the grounded nozzle.  I
also get the impression that it works fairly fast (a second or less.)
which hardly seems like enough time for the probe to heat up.  If one
assumes that all the current is carried by the electrons in the
plasma, then you see more current when the electrons are moving in the
same direction as the mass flow in the flame.... maybe that is
important?

I don't see how the relative areas matter though.

George H.

this also mentions area http://www.robertshawtstats.com/spaw2/SiteContent/150-2163_RevB_1.pdf

something kinda related is ion sensing ignition, biasing the spark
plug gap to 80V
and looking at the current right after the spark things like misfires,
peak pressure position and
knocking can be detected

-Lasse

 

[Robert Macy]

Hi guys,  A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work.  The sensors are just a
metal rod that sit’s in the flame.  They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive.  Now here’s the weird part.  The flame sensor shows
rectification and so only has to sense a DC current.  I’m totally
clueless as to how you get rectification.   If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you’ll see he talks about different areas being important.  But no
other explanation.   Anyone have an idea of what’s going on?

Thanks
George H.

I ralize my answer here is off topic for your question but others may
be interested.

It is my understanding that years ago Japan had a major fire
throughout a city, so some engineers took the opportunity to use the
blocks and blocks of fire to make some measurmeents. They beamed
through the flames and smoke and obtained REAL data regarding the
conductivity over a spectrum of an actual fire.

Don't know where that data resides, but it does represent some
incredible information. Anybody know where it is?

 

[tm]

tm wrote:


Water heaters would almost certainly use a thermopile, and be powered
by the flame, no external power supply needed. Usually, when the
thermopiles fail, they have internal shorts or opens, and have to be
replaced.

Jon

This is an automatic control that needs to instantly detect ignition of the
gas. A thermocouple would take too much time to heat up. In that time, the
whole system gets filled with the perfect air-gas mix. Then you get a "Joerg
super-phuut".

tm

 

[rickman]

Hi guys, A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work. The sensors are just a
metal rod that sit’s in the flame. They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive. Now here’s the weird part. The flame sensor shows
rectification and so only has to sense a DC current. I’m totally
clueless as to how you get rectification. If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you’ll see he talks about different areas being important. But no
other explanation. Anyone have an idea of what’s going on?

---
If you have a setup like this: (View with a fixed pitch font)

where you have one spherical electrode (large area) and one pointy
one, (small area) then when the pointy end goes negative, the field
strength at the point will be high and it'll be easy for electrons to
jump the gap, as long as it's not too great.

When the spherical electrode goes negative, however, the field
strength will be much lower and it'll be hard for electrons to jump
the gap.

Voila, rectifier!

. +---O <--+
. | |
. +--[AC]--+

Why would the field strength be different for different applied
polarity? The field depends on the applied voltage and the electrode
geometry.

The field strength at the same electrode is the same, but the end that
would be emitting electrons changes. It is much harder to emit atoms
with a positive charge than electrons, so only the negative end matters.

 

[rickman]

Hi guys, A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work. The sensors are just a
metal rod that sit’s in the flame. They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive. Now here’s the weird part. The flame sensor shows
rectification and so only has to sense a DC current. I’m totally
clueless as to how you get rectification. If you scroll down to the
description (back ground of invention) here,

http://www.freepatentsonline.com/5472336.html

you’ll see he talks about different areas being important. But no
other explanation. Anyone have an idea of what’s going on?

Back when I was in high school I read about a flame loud speaker. I
built one using a propane torch and two pieces of carbon from dry cells.
It actually worked! Not very loud but it worked. I entered it in the
science fair. It didn't occur to me to use the torch for one of the
electrodes. I've still got the brackets for holding the torch somewhere.

 

[brent]

Brilliant!

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot comhttp://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation

I think he is confusing that with a rectum-fryer.

 

[George Herold]

Hi guys,  A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work.  The sensors are just a
metal rod that sit’s in the flame.  They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive.  Now here’s the weird part.  The flame sensor shows
rectification and so only has to sense a DC current.  I’m totally
clueless as to how you get rectification.   If you scroll down to the
description (back ground of invention) here,

you’ll see he talks about different areas being important.  But no
other explanation.   Anyone have an idea of what’s going on?

---
If you have a setup like this: (View with a fixed pitch font)

where you have one spherical electrode (large area) and one pointy
one, (small area) then when the pointy end goes negative, the field
strength at the point will be high and it'll be easy for electrons to
jump the gap, as long as it's not too great.

When the spherical electrode goes negative, however, the field
strength will be much lower and it'll be hard for electrons to jump
the gap.

Voila, rectifier!

.   +---O <--+
.   |        |
.   +--[AC]--+

As far as I know the area thing may be a red herring.

If I believe the patent the prefered electron flow direction is from
the large area towards the probe... go figure.

George H.

 

[George Herold]

Brilliant!

"OK genius, what's your answer"
(in the voice of Kayla from Firefly. :^)

There's also a big thermal gradient from the nozzle to the probe.

Seems like the electrons would like to flow along the thermal
gradient, from hot to cold.

George H.

 

[Artemus]

Back when I was in high school I read about a flame loud speaker. I built one
using a propane torch and two pieces of carbon from dry cells. It actually worked!
Not very loud but it worked. I entered it in the science fair. It didn't occur to
me to use the torch for one of the electrodes. I've still got the brackets for
holding the torch somewhere.

I remember that. I think I was in HS too. As I recall there were 2
electrodes and HV was required but I don't recall how high. Was
there a DC bias too? The low end response was limited by the length
of the flame between the electrodes but the high end was way up there.
It was also supposed to work as a microphone too.
Art

 

[Bill Sloman]

where you have one spherical electrode (large area) and one pointy
one, (small area) then when the pointy end goes negative, the field
strength at the point will be high and it'll be easy for electrons to
jump the gap, as long as it's not too great.

When the spherical electrode goes negative, however, the field
strength will be much lower and it'll be hard for electrons to jump
the gap.

Voila, rectifier!

.   +---O <--+
.   |        |
.   +--[AC]--+

As far as I know the area thing may be a red herring.

If I believe the patent the prefered electron flow direction is from
the large area towards the probe... go figure.

The interesting part of electrical conductivity through gases usually
turns out to be what's happening at the surfaces of the electrodes.

At the positive electrode it's usually electron capture. At the
negative electrode you've got to have something going on that ejects
electrons from the electrode surface.

Mostly this is positive ion bombardment. If you get up into the arc
regime, the surface gets hot enough to deform and in an electric field
it deforms into a bed of spikes, with the tips of the spikes sharp
enough to give you thermally assisted field-emission. Keeping a glow
discharge going with positive ion bombardment needs a lot more voltage
drop up against the anode than does an arc.

The pointy electrode is going to get to arc discharge conditions a lot
faster than a smooth spherical electrode. That might do your
rectification.

 

[miso]

The therrmopile is certainly standard for many older water heaters. But
the application is the pilot is already on, and you are just detecting
when the pilot light goes out. If the pilot goes out, the thermopile can
no longer hold open the relay for the gas.

Obviously over the years the safety features have become more
complicated. Look at a Takagi water heater if you want to see a
complicate control system. Not only does it detect if there is a flame,
but it can check the quality of the flame to insure the heater is
burning the fuel correctly.

Takagi makes good stuff if you are looking for a tankless heater. They
don't spend a lot of money on mass marketing.

Back to flame detection, it seems counterintuitive, but some flame
detection schemes look for UV light.

 

[Uwe Hercksen]

Hi guys, A colleague had his flame sensor in his gas furnace fail and
this led to a discussion about how they work. The sensors are just a
metal rod that sit’s in the flame. They apply an AC voltage to the
rod and measure the current going from the rod to the flame nozzle.
The flame is a plasma and conducts ~micro amp currents with ~ 100
volts of drive. Now here’s the weird part. The flame sensor shows
rectification and so only has to sense a DC current.

Hello,

here is another application of currents through a flame:
http://en.wikipedia.org/wiki/MHD_generator
A generator without moving parts. It is possible also using
thermocouples, but the efficency is poor.

Bye

 

[[email protected]]

The interesting part of electrical conductivity through gases usually

turns out to be what's happening at the surfaces of the electrodes.



At the positive electrode it's usually electron capture. At the

negative electrode you've got to have something going on that ejects

electrons from the electrode surface.



Mostly this is positive ion bombardment. If you get up into the arc

regime, the surface gets hot enough to deform and in an electric field

it deforms into a bed of spikes, with the tips of the spikes sharp

enough to give you thermally assisted field-emission. Keeping a glow

discharge going with positive ion bombardment needs a lot more voltage

drop up against the anode than does an arc.



The pointy electrode is going to get to arc discharge conditions a lot

faster than a smooth spherical electrode. That might do your

rectification.

I don't think this has anything to do with thermionic emission, it is an ionized plasma phenomenon. The charged particles arrange themselves to maintain some kind of space charge neutrality by developing a spatial potential, and this potential has to be overcome before the plasma conducts. Since thepotential has a fixed orientation, the plasma ends up being a rectifier. The patent requirement about relative areas of the probes probably has to dowith establishing the geometry and orientation of the plasma potential. The electrical properties of flames and their resulting plasmas have been studied extensively since the early 1900s but I don't find anything on the rectification effect until 1966 when computers became available to numericallysolve the multitude of equations representing all the physics.