Here's my furnace repair story:

My furnace, after a long summer of no use (and a little in the spring when
I had bought the house), worked, but started smelling a bit funny in the
utility room.  A couple days later it started getting unreliable!

Luckily, it wasn't too cold out yet and it was also Sunday.  So I started
looking into the innards.  I figured that I was qualified to work on the
electrical part, and as long as I didn't take a wrench into there I didn't
have to worry about a gas leak.

The funny smell I had figured was something burning clean, as furnaces tend
to do in the autumn.  Anyway, I played around, check voltages, etc.  My
furnace has a hi-tech control box, so I opened that up and looked for scorched
components.  The problem seemed to be that the signal never got to the gas
valve, or else the gas valve was ignoring the signal.  I finally ended up
getting the control box to start everything else, then I jumpered the 24VAC
signal right to the valve.  It worked!  So I needed a new control box.

$95 later, I installed the control box, and everything worked.  Except for
the smell!  And, a safetly thermal switch had gone off once.  Watching the
thing run, I saw flames "leaking" a bit from under the hood-thing that is
supposed to direct all the flame up thru the heat exchanger.  In a flash of
inspiration, I used an old fire extinguisher, filled with air, to blow a bunch
of soot out of the heat exchanger!  Then it all worked!  No more funny smell!
And it wasn't *that* much soot, but enough.

So the story turned out to be:
1.  Clogged heat exchanger, leading to hot gases getting into the control
areas.
2.  Old control box subjected to extra heat, leading to failure of gas valve
relay.
3.  Extra heat occasionally sets of safety switch!  Clue!
4.  Funny smell from burning gas (just like a huge gas stove, so not
incredibly dangerous).

42 responses total.

How a gas furnace typically works:
A gas furnace works by burning gas (natural, propane, etc) in a compartmented
chamber called the "heat exchanger".  Air is blown thru separate compartments
and then into the rest of the house.  

The gas comes in a pressurized pipe, and just needs to be directed and lit
on fire.  Most furnaces use either a "pilot light", which is a very small
amount of gas burning all the time, or else (like mine) "hot surface", which
means a little tiny electric element like a stove burner.  Either way, the
"thermostat", mounted in some central location in the house, sends a signal
whenever heat is needed.   Some kind of control system, almost always
electrical and usually based on 24 V AC, sees that signal and turns on the
gas.  If needed, the "hot surface" ignitor is turned on.  Most furnaces (and
ovens, too) have a safety feature:  They turn on a small amount of gas, then
wait for an exposed part to heat up before turning on the main gas.  This way,
a dead pilot light won't result in the full gas feed going into the house
instead of being burned.

Finally , a second thermostat mounted inside the furnace turns the big air
blower on and off.  The blower is turned on only when the furnace has "warmed
up", and off after it has cooled down.  If the furnace blower was run directly
off the main house thermostat, it would prevent the furnace from warming up
properly (efficiency suffers) and then cut off while there was still heat
left in the heat exchanger after the gas cut off.

We had a "high efficiency" furnace installed a couple of years ago. Last
winter it started to have a problem with the inducer fan going on and off
many times before the furnace started. This fall I had the installer out
to check out the system and also to see what the problem was. This cost
$$, but he couldn't get the cycling to repeat. Finally, I turned on the
furnace for this Fall - and cycling started. This time, however, having
had the furnace checked out, I *rebooted the computer controlled
thermostat* (which consists of pushing the reset button - though I also
cleaned the backup battery terminals). Voila - cycling has stopped. I
don't know why that didn't occur to me last winter (especially observing
Grex being rebooted frequently for its accumulated gremlins).

I installed my own high efficiency furnace about 12 years ago.  It's
not the quietest thing on the planet but it hasn't given me much
trouble either.  I keep it *very* clean and well oiled.  there is a
lot more to go wrong in modern, more efficient furnaces so it pays
to keep them up.  The first day I had it it refused to come on.  I
found that the "BLOWER DOOR OPEN!!" safety switch had failed open.
Since I don't mess with the blower with the power switch ON, I 
replaced the swicth with a piece of wire. (I also use power mowers
that keep running when you let go of the handle.)  It also refused
to ignite one time and I found that there was a low spot in a little
tube between the inducer fan and a pressure switch.  Condensation 
would plug this tube so that the pressure switch could no longer "see"
that the inducer fan was doing its job and not allow the burners to'
fire.  I replumbed this little tube so no condensation could collect
and it's been fine ever since.  On second thought, that was not a
pressure switch but a vacuum switch.

Must have been an air-flow switch, since either a static pressure or static
vacuum switch wouldn't be much affected by water.  The water would just be
pushed/pulled one way or the other.  Unless you had a *bunch* of water?

The vacuum created by the induction blower is very weak when everything is
working correctly.  A fraction of an inch of water I suspect.  The slug of
water in the 1/4" tube in question was only about 1/2" to 3/4" inch long.
I "fixed" it while the blower was running but the burners refused to light.
I disconnected the hose, draind the little water, reconnected it and the 
burners fired about a minute later, as they should.

My furnace (gas hot air) has a primary blower for the flame and was making
a lot of noise.  My repair person said that those burners need to be oiled
at least once a year, or even more often (not true usually of the big blower
driving the air through the duct work).  Mine has been quiet since it has
been oiled on a regular basis.

My induction blower has no provisions for oiling, so I haven't.  Perhaps
I'll get in there with a syringe oiler and try to get the shaft bearing
that way.  Either way, it has shown no sign of needing oil and the label
on it says something about being oiled for life.  The big blower motor
gets oiled every other season though.

Oiled for life....of the inducer? The motors on my furnace also have no oiling
point, including the big blower. I think I'll see if I can get some into the
bearings anyway. Are there any cases where a little oil is *detrimental*? Too
much oil in a clock is very bad, but this furnace is no clock.

Well, I gess if you put a lot in the furnace blower motor, it could run out
and get slung onto the hot heat exchanger and blue smoke could billow from
your heating ducts...

Somebody more knowing check me on this, but I have never had any success
oiling a sealed bearing.  If it goes, then replace it I believe is
the common solution.

I've done so but it usually takes disassembly to do it right.

I notice that so far this item is all about gas/forced-air.  

In this house we have gas/hot-water (actually wood-or-gas/hot-water, 
but that's another story) which this week was waterlogged as well as
needing a new thermocouple for the pilot light connection.  We've 
learned that we can replace the thermocouple ourselves, but the
"universal, fits most furnaces" kind from the corner store doesn't
fit, it has to be "universal, fits all furnaces".  (The old one
seemed to work OK after corrosion was sanded off, so we're keeping 
it for a backup)
 
Does anyone know of a reliable book about this kind of furnace?

Waterlogged hot water system?  What does that mean?  (I want to get the
terminology right since we may well be switching to hot water / radiant
floor heating in the next few years.)  When I think of something being
waterlogged, I think of something like wood of styrofoam that sinks in
water.  Is it that the float for the makeup water supply was waterlogged?
 
So far as your thermal couple, it may have been working all along.  They
generate a very low voltage but significant current.  If the thermal
couple connection to the gas solenoid is poor, there may not be enough
current to keep the gas valve open.  It may also have an intermittent that
you "fixed" by flexing it.  Cleaning the corrosion off the end shouldn't
have fixed it.  I'd suggest a visit to some of the heating / cooling
suppliers on Industrial Hwy. (Or where ever.), thermal couple in hand.
I also don't think any one book will be able to explain you furnace to
you.  Search the WWW, your library, book stores, etc. for snippits of
info and look at your system to see how well they corollate to what you
have.  You'll find that the basic principal is the same for most systems
but that there are differences in the details.
 
Back on the subject of thermal couples, I've noticed that they will often
fail slowly such that they work for a while but eventually let the gas
solenoid close because their current output has degraded over the years.

What drew our attention to the furnace, this year, was that the pilot
light went out more than once, notably when the main burners were just 
shutting down.  (Though once it died when we were looking at it with 
the protective covers off)

"Waterlogged" may not be the technical term, but the system needs to have
space for the hot water to expand into; ours has two tanks up next to the
ceiling in the furnace room that should be full of *air* when the water is
cold, and have some air in them at all times.  There are relief valves that 
operate to let water out before the water temperature or pressure gets 
dangerously high, one of which would put the fire out if we were burning 
wood at the time, but if the pressure is regularly that high then the
system has too much water in it for its own good.

Got it.  Much like the blader tank used in most well sourced water supply
systems.  I think waterlogged is the correct term.
 
Interesting that your pilot light goes out when the main valve shuts off.
Is the end of the thermal couple poised such that it is in the pilot light's
flame?  I'm think that the draft created by the main burner, or by your
having the covers off, could allow the couple to cool to a point where
it releases the pilot valve.

Well, the Furnace Guy was here this evening.  

My furnace had had a couple hiccups in the last couple days, essentially
needing a reboot (AC power off and then on).  

The furnace guy had a couple interesting insights:
1.  The most likey cause was the hot surface ignitor.  Seems that Ruud had
designed the control box to use some electrical aspect (I'm guessing
resistance) of the ignitor to determine when the flame was lit.  Most furnaces
use a separate doohicky for this (called a "rectifier").  My furnace has a
kit that puts in a separate rectifier tube and patches it into the ignitor
circuit.  He replaced the ignitor to fix it short term.
2.  The fan is partially clogged with accumulated gunk.  Nothing actually
blocked, but enough to reduce the efficiency.  This explains why the burner
tends to cycle on and off - reduced airflow, furnace hits temperature limit
and cycles off for a few minutes. 

So, no way would I have guessed this stuff myself.   I need to obtain and
install this little kit for long term reliability, and take the fan innards
to the car wash for the same reason.

You can take the guts outside and blow the dust out with an air compressor.
I have also used a canister vacuum for this purpose.  I find that the fan
blades in the blower like to collect dust for some reason.  It really gets
caked on too!
 
My furnace uses eletronic ignition.  I don't know how it knows to turn off
the spark when the burner lights.  I'll have to investigate!
 
A rectifier?  Humm.  Somehow I think the guy is just throwing terms at you.
Rectifiers are used to turn AC into pulsating DC.  I can't imagine that two
electronic components, serving totally unrelated functions, can have the same
name ;-)

The original concept of the "gas rectifier" was not electronic... in a lot
of furnaces it just senses (by expansion of a fluid) the heat of burning and
uses that to tell the gas valve to turn on full flow to the burners.  

Some of the definitions of "rectify" in my dictionary look appropriate to this
application.

I don't know scott.  Neither my Websters or electronic dictionary
mention anything about expanding fluids.  Only rectification as it 
pertains to the purification of alcohol.  The device you mention sounds
more like a thermometer with a remote sensing bulb.  The pressure of
the liquid increases with temperature and trips a switch when the 
pressure reaches a predetermined point.  It would work but would be
sensitive to ambient temperature and slow to respond.  I would guess
that an IR detector would be more appropriate.  (Several years ago I
did an off-the-cuff experiment involving measuring the electrical 
resistance of a flame.  At the time I found it a viable method of
flame detection.  I have not since repeated the experiment nor do I
recall what the fuel source was.  I do recall wondering if the fuel
type would change my resistance readings, making it possible to 
determine what fuel was being burnt.)

You have just invented the Flame Ionization Detector, used in gas
chromatography. 

As an on/off sensor, drift is probably not much of a problem (this thing
actually sits *in* the flame of the starter burner).

Random House College Dictionary:
1.  to make, put, or set right; rememdy; correct
2.  to put right by adjustment or calculation

The dongle in question is designed to turn off the gas if no flame develops.

I suppose one could use rectifier in that sense.  (Klaus thinks about
calling anti-lock brakes rectifiers  ;-)
 
Are you sure the device your refering to isn't called a thermocouple?
*very* common in gas fired appliances.

"That part is used to rectify the gas" is the context.  Doesn't matter exactly
what the thing is, it rectifies.  

It's possible to use some pretty crude devices for this.  It isn't like a
thermometer device, it just needs to sense two different states:  cold, and
really really hot (it sits *in* the flames!).

Ok, perhaps skid rectifier would be more in line than ;-)
 
Yes, it is possible to use some pretty crude devices for this because of
the large temperture (Humm, need to activate my spelling rectifier..)
differential between flame and no flame.

Quick update:  

I bought and installed the "remote flame sensor" kit, and it seems to have
cured the flame cycling problem that the fan cleaning didn't fix.

I spent some time working on my furnace this weekend after I noticed the
induction blower was making a tiny rattle while running.  It has no oil
holes and I had never oiled it in the 12 years I've had it.  (Sticker
on it said "Oiled For Life")  I figured its life might be coming to an
end so I disassembled the blower and finally got the motor pulled apart.
Sure enough, the shaft end bearing and shaft were showing signs of wear.
I stuffed some cotton around the oil impregnated bushing as the original
oil-holding material was minimal at best.  I then saturated it with a
good quality 20 weight motor oil.  (Stay away from 3-in-1 Oil!!!  Bad
stuff!  I have had it gumm up motors in the past, basically destroying
the motors!)  After putting everything back together, all is fine and
I don't expect to have to fool with it for another ten years or more.
 
While fooling with the furnace, I also investigated its electronic,
spark, ignition system.  It seems to use flame ionization to detect
a lit pilot.  There is no other sensor in or around the pilot or main
flame and the ignitor (Spark plug type thing.) tip is immersed by flame
once the pilot is lit.

I have had a local plumbing and heating contractor look over the blue-
prints to the house I'm building so they could design the in-floor
hydronic heating system.  The designer came up with a basic plan for
tubing layout with 5 options for heatin the water that goes through
it.  Since I had told him that I wanted to go with propane, he did
not include geo-thermal heat pump systems.  Geo-thermal heat pump
systems are considered the most efficient these days.  However, I
find them expensive to install, mechanically quite complex and open
loop systems (Use well water as a heat source and dump the "expended"
well water into a lake, pond or drainage ditch at a rate of 8 to 12
gallons/minute.) use too much water.
 
The system I settled on use only one 65,000 BTU hot water heater
for heating the house and heating domestic hot water!  Bradford
White wakes a hot water heater with an integral heat exchanger.
A warm water storage tank, located next to the hot water heater,
will be maintained at whatever temp. is called for by the floor
heating system via this heat exchanger.  (80 to 110 F)  The floor
will then be heated off this storage tank.
 
Not only that, but this system won't require a heating contractor to
install since a hot water heater is not considered a boiler and one
doesn't need a license to install a hot water heater.
 
This cuts the costs of the system drastically.  I will also use a 
Reset Controller.  This is a little computer that senses outside temp,
inside temp and "boiler" temp.  It then uses this information to
determine the best temp for the supply water to the floor heating
system so that room temperature doesn't change quickly or short-cycle
when it's warmish outside.

the described system sounds good.  Make sure you insulate well under the floor
heating units.  I will be interested in hearing how it works over a couple
of years since I would eventually like to do something like that in my house.

I will have R-10 under the peremiter of the basement slab. R-5 under the
remainder.  The site is all gravel and I have yet to see water collect
in the basement, despite being uncovered and being through several torrential
downpours.  Dry conditions reduce the heat siphoned away.
 
The the first floor will not have insulation under the floor since the
basement will be heated too and loss there will be acceptable.  Besides, it
can always be added in the future if needed.

I finally got around to checking out why one of my rooms never got much
airflom from the furnace.  In my house (one story, no basement, no crawlspace)
the air to the bedrooms is run thru flexible ducts in the attic (yeah, I know,
massive heat loss, but I don't have much other choice).  This room had the
longest run, leading to theory 1:  Run too long for diameter of duct.  But
also there was theory 2:  Flexible duct has flat spot where somebody stepped
on it, restricting flow.

To today I went up into the attic (I turned the fan on before going up so I'd
notice any leaks) with the express purpose of finding the problem.  No 
squashed sections, killing theory 2.  Hard to really check theory 1 without
replacement, so I went on to theory 3: Obstruction (unlikely).  I removed some
duct tape and got the furnace end loose, finding that the duct had been
sloppily attached during insulation, allowing interior layers to fold inward
and block the flow (flexible round ducting; plastic, spiral wire for shape,
fiberglass insulation, and and inner layer of plastic).  I corrected that as
best I could, taped it back together, and now I have decent airflow!  Now that
room is warm, finally.

Side note:  My furnace also seems a bit large for the house, in that the fan
has trouble pushing enough air thru the ducting in order to keep the furnace
from overheating.  If the filter got sufficiently dirty the furnace would
start cycling as the heat exchanger got too hot.  Unclogging the duct should
allow more flow, allowing the fan to push enough air to keep everything happy.

I have also reduced the output of a furnace buy 1) removing one of the gas
jets on the burner assembly and replacing it with a 1/8" NPT plug. 2)
reducing the gas pressure from 3" H2O to 1-1/2" H2O (There is usually a
1/8" NPT plug in the gas manifold that can be temporarily removed and
replaced with a brab so that you can connect a piece of tubing with water
in it and measure the pressure.)  3) reduce the orifice sizes on your burner
so that you get fewer BTU output.  (As I recall, orifices are rated in
BTU's @ 3" water.)  steps #2 and #3 may require that you fiddle with the
air mixture to get a nice blue flame with as little yellow as possible.

    I was quoted US$ 9,000 to have our furnace, a/c and duct
work replaced.  I could pay extra to have a variable-speed
blower which is supposedly more efficient.  I don't have a
source of money for this kind of project so we've just been
paying the HVAC people to come out and bang our existing
furnace with a hammer when it fails. One of them even hacked
an air filter into it.

I'm glad to see, though, that no-one had any problems with their furnaces for
the past 12 years (since 1999....), until now. 

Grex, the furnace guardian![tm]

One advantage to renting is that I don't have to come up with $9,000 if 
the furnace breaks down.  I hope you get yours taken care of ok, Andy.

    I suppose we should get the furnace checked out (and
improvised filter replaced) before winter sets in.

Filter replacement is a good thing.
Maintaining the blower bears with silicon grease is a great thing.

Would bear grease work better?

bearings..whoops

- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss