Gas and Oil Home Heating Furnaces - HEIL DC90 intermittently fails to heat

In January, our natural gas HEIL DC90 furnace (Model# NTGM100EHA2 Serial# L974181359) started to periodically not fully light. That is, the inducer motor would be on, the pilot would come on, and the blower would come on... but it would not be at full ignition.

Since it was rather cold (south-central Wisconsin), I was mostly concerned about getting heat to the house (and less about troubleshooting), so I tried some things to get it to restart:

- Turning the thermostat off and on fairly quickly didn't seem to do the trick (though later this has seemed to work... provided the thermostat is off for several minutes).

- Cycling power (rapid on/off) to the furnace generally (more on this below) seems to work to get it to start behaving.

After this started happening with some frequency (approximately once every couple of days to no longer than a week between incidents), I decided to do some investigation. After researching a bit on the internet, I shut the system down and opened up the combustion chamber. I was surprised to see quite a few intact leaves in there... not sure how long they have been in there or how they got in there given the apparent lack of access.

I looked at the flame sensor/HSI igniter assembly and lightly cleaned off the sensor with some 480 sandpaper.

The problem went away for about a week, then came back periodically.

Since about 3/14 (when it was getting warmer at nights here), it has been recurring in the morning (after having been off all night) and the afternoon (again... after having been off most of the day... especially if it was a sunny day). I should add that many of the occurrences prior to 3/14 were during these times as well. I would guess (though there may be a couple of outliers) that this failure-to-ignite happens only after the system has cooled for a few hours.

Anyway, since it is getting to be more repeatable, I am trying to be more vigilant... and ending up a bit more frantic.

After more research, I ensured that the condensate drain was clean and uncrimped, that the gas opening to the combustion chamber was unblocked, and that the pressure switch was functioning (went from infinte resistance to zero when the draft inducer was on).

I stumbled across the troubleshooting guide for the sv9501m 2728 Honeywell SmartValve (see p. 14 of http://www.transtaracsupply.com/pdf/sv95control.pdf) and got to the following decision point in the chart (this is after measuring 24V for the main connector):

MAIN VALVE OPENS AND
MAIN BURNER LIGHTS -> NO ->

• CHECK THAT PILOT FLAME MAKES GOOD
CONTACT WITH PILOT BURNER FLAME ROD
• CHECK FOR GOOD ELECTRICAL CONNECTION
THROUGH THE PILOT TUBING.
• IF BOTH OF THE ABOVE ARE GOOD, REPLACE
IGNITER/FLAME ROD ASSEMBLY

and I have run into a bit of trouble understanding the checks (and whether that is a complete diagnostic).

First, I am assuming that the pilot flame makes contact with the flame rod based on:

a) removing the combustion chamber coverplate and gauging where the rod is relative to the igniter (since I can't see the rod well through the small porthole)

and

b) because the system does go to full ignition many times

Are those fair assumptions?

Second, what do they mean by "GOOD ELECTRICAL CONNECTION THROUGH THE PILOT TUBING"? Continuity from where to where? Why the tubing and not the wire?

Third, why does this automatically mean the igniter/flame rod assembly? Couldn't it also be the connection to the SmartValve?

Also, is that chart complete up to that point, or should there be other tests prior to the one I'm stuck on?

As an aside, I tried calling Honeywell support, but the homeowner line quickly seemed overwhelmed with my questions and referred me to the retailer help line... who refused to answer any questions at all. Lovely customer support, that... :wall:

So, with those questions hanging in the air, I played around with a few more things in seeing what would get the furnace to fully light once it was in the pilot-only state:

- Switching the On/Off on the SmartValve does it (at least the two times I tried it);

- Unplugging/replugging the square plug from the SmartValve does it (usually);

- Unplugging/replugging the igniter/flame sensor plug from the SmartValve does it (usually);

- Disconnecting/reconnecting the vacuum tube from the pressure sensor does it (usually);

The "usually" thing also has this novice a bit baffled.... though I think that it doesn't happen due to my replugging/reconnecting later than the SmartValve would like. That is, if I am slow to reconnect then the igniter/sensor won't be ready soon enough after the SmartValve opens... so I get a pilot but not full ignition.

At the risk of further showing my naivete, it only occurred to me last night to actually go down and watch what happened when the thermostat made its morning call.... and it was like what happened when power cycling the furnace failed (on one or two occasions) to bring about full ignition:

- draft inducer motor came on in response to thermostat call
- click from the SmartValve
- igniter slowly started to glow
- pilot eventually lit
- furnace electronic filter and blower came on

Within 15 seconds or so of the draft inducer motor coming on I unplugged/replugged the square plug to the SmartValve, and then full ignition resulted... which I was surprised at, really... since that kind of blew my pilot-warming-the-igniter theory...

I was thinking of playing around with the pilot flame height, but - really - anything having to do with the actual natural gas has me feeling like calling in a service tech.

I'm just trying to figure out if it is a board or other non-gas part that I can replace, or if it is the SmartValve giving me this grief.

I have not disconnected the intake/exhaust ports outside to see if I could snake anything out, nor have I washed the electronic air filter components (since I typically only do that once a season and we keep a pretty dust-free house)... largely because I'm thinking that these wouldn't have any impact.

I apologize for the long post and while it appears that others have had similar problems I was not always clear on the final fix.

Look at where wires enter the enclosed burner compartment, and follow these back and see where they plug into, and read and list for us the abbreviation letters on the board, where they plug in, and tell us the number of wires involved.

There are three wires (2 blue and 1 black) coming from the combustion chamber that plug directly into the Honeywell SmartValve (sv9501m 2728). Are those the ones you mean?

As for the board, it is a Honeywell ST9120C 4040.

I should add that there are two (I think) red wires coming from the chamber, but they are behind the three mentioned above and I have not traced them yet.

Interestingly, power-cycling the unit this morning did not get it to go into full ignition... twice. I ended up unplugging/plugging the square plug (not the igniter/sensor).

Smart valves can go bad, but one wants to be certain it is not something else, as these things are pricey. Be clear on what wires go to smart valve and which go to the board and see if there are letters/abreviations where they connect up at each of these (more likely on the board).

I am not too familiar with smart valve furnaces, to envision how they are wired up. (I deal with more smart valve gas water heaters)

The two blue and one black wires go from the combustion chamber to the SmartValve and plug into the Igniter slot on the valve.

The other two red wires start with spade connections at the side of the combustion chamber (unlike the blues/black which go through the chamber wall and connect to the igniter/sensor) and are bundled in with the other control wires on their way to the board.

On their way, one of the red wires connects (via two spade terminals) to a small board screwed to the side of the furnace (under and separate from the combustion chamber... maybe I am not using that term correctly... the chamber I am referring to is where the pilot is housed and goes to full ignition).

The board has the following on it:

36T0183 42269 J97359
L180-20F 1320-547 MEXICO

Out of curiousity, I undid the two screws holding the board in place and found the other side to have two (approximately) 3" heavily insulated wires that end attached to a small silver disc about the same circumference as a large calculator battery and twice as thick.

After the detour, the red wire joins its mate and they end up with four other wires in a 6-wire plug at the control board (ST9120C 4040).

In review of things said:

You say the pilot always comes on and stays on, right?

And you said that there is ignition, but not complete ignition? What exactly do you mean by that? Do some or all of the buners light up? Is the flame height on the burners insufficient? Please explain about the partial ignition you talked about in your first post.

If when the pilot comes on, and you disconnect the black wire on the smart valve at the ignition wires hookup, does the pilot flame go out?

Guessing off the top of my head that the red wire going to 2 spade connectors under the burner area is for flame roll out or some other limit switch wires. Whatever they are, you should have continuity through those spade connectors and also 24 VAC through the spades, when the inducer motor runs.

In review of things said:

You say the pilot always comes on and stays on, right?

Yes... the pilot always comes on and stays on... until, of course, I do something like unplug the control/igniter/etc.

And you said that there is ignition, but not complete ignition? What exactly do you mean by that? Do some or all of the buners light up? Is the flame height on the burners insufficient? Please explain about the partial ignition you talked about in your first post.

What I meant was it either goes (and stays) at the pilot stage, or it goes to full ignition with all burners lit. It has never done a partial, with only a few burners. Sorry to have been unclear on that.

As to flame height, I would assume it is good (when they are fully lit) since the house gets warm, but I can't completely verify that since I never really looked at them prior to the furnace acting up.

If when the pilot comes on, and you disconnect the black wire on the smart valve at the ignition wires hookup, does the pilot flame go out?.

The black wire is part of a plug with the 2 blue wires. When I unplug that from the SmartValve, the pilot goes out and the SmartValve makes a click-buzz. The same thing happens when I unplug the other plug (marked 'Control') on the SmartValve.

Guessing off the top of my head that the red wire going to 2 spade connectors under the burner area is for flame roll out or some other limit switch wires. Whatever they are, you should have continuity through those spade connectors and also 24 VAC through the spades, when the inducer motor runs.

Yes, resistance is at 0 ohms (at all times, if I remember correctly). I didn't check voltage, but I can confirm both and get back on that.

I get 24VAC from one spade to any other piece of metal, but 0 when touching the other spade. This reading is constant regardless of whether there is no pilot, pilot, or full ignition.

Please let me know if the attached schematic is (or not) an exact match for the wiring and components of your system. And if there are differences, where at (wire colors? labels of terminals? etc.)

http://img2.freeimagehosting.net/uploads/dc50667711.jpg (http://www.freeimagehosting.net/)

Yes, that is a slightly simplified version of the Connection Diagram pasted to the lower door of the furnace.

Missing from your diagram and present on mine are:
- LP-specific switch
- horizontal-installation-specific switch
- optional capacitor

Missing from mine and present on yours is the G coming off the C from the Furnace Control board.

Otherwise, it is identical.

Using the schematic I posted yesterday night (post #10) as a reference...look at the 6-wire plug on the ST9120C board.
Terminal #1 (top left) and terminal #6 (bottom right) have each a red wire. These two red wires are connected to the limit switch (this is the "silver disk" you describe in post #6).

Note: I'm not sure if you have rollouts here. Pictures would really help, as a defective rollout could very well be the reason for this malfunction.

moving on:
Without disconnecting the red wires attached, put a jumper between the two spade-type electrical connectors. In other words, I want those two red wires to stay engaged (connected) with each other. And IF there are rollouts (the pics will show), what I would really like is the ends of this jumper to be at the ends of each one of these two red wires coming off terminals 1&6 on the board (in other words, bypassing the limit "and" if there are any, the rollouts as well)

Call for heat.

Tell me what the furnace does (or does not)

p.s.: as a side note: the two blue and the black wires going to the pilot burner assembly are as follows:
The two blue ones feed the 24V HSI. The black attaches to the flame rod (this is the one, I suppose, you cleaned with the sand paper)

I was set to try this this morning and (like yesterday morning), it fired right up without a problem. It did fail yesterday afternoon and evening, so there is hope to try the experiment tonight or after.

However, there is no guarantee that it would be "correct" in that the furnace may light properly of its own accord.

So, is it safe to jumper the two red wires and leave it that way?

I am thinking that if the problem doesn't then reappear, I could then reasonably replace that limit switch... or could that limit be failing the first time around for a perfectly good reason that jumpering might aggravate?

And, yes, the black wire is connected to the flame rod that I cleaned (very, very lightly... not like sanding a dowel or something).

Thanks for the input and I'll report back when I know something.... or more likely when I don't. :)

So, is it safe to jumper the two red wires and leave it that way?

No, it is not safe to leave the jumper in place. This switch is in charge of shutting the main burners OFF if an excessive temperature build-up were to occur in the heat exchanger.
Not that this is your case (an over-temperature condition), but when it comes to safeties, one can never be too safe.

The jumper is to see if the limit is the source of the trouble. I am suspicious of it b/c you report that together with the inducer blower and the pilot burner flame, the main (house) blower runs albeit not the main burners on a call for heat. A situation such as that often is caused by a hanging limit switch.

If you were able to post some pics, that would be quite a treat.

Unfortunately, I don't have any way of getting pics up as much as I agree that that would help...

And I figured that that would be the answer WRT safety...

I take it that a failure of the Main Limit switch would have different symptoms (I ask only because that is also on the same line of red wires).

I take it that a failure of the Main Limit switch would have different symptoms (I ask only because that is also on the same line of red wires).

How many switches are there in between the red wires?
Any one of them failing (tripping into the open position) will cut power to the main burners and kick the main blower ON.

From your schematic (which isn't showing for me at the moment) and what I see, there are two:

- the one I described below (which interrupts one of the wires) which is the Rollover (?)

and

- the Main Limit (I presume) which both plug into via spade connectors at the combustion chamber (just behind where the igniter/flame rod wires breach the chamber wall)

I am presuming that those are the two (as the third listed on the schematic for those wires was listed as only being on some models... and I only see two possible places for a switch/sensor to be for those wires... i.e. between the combustion chamber and the control board there is only the 'interruption' at the chamber and the other one I described below).

When I removed the one switch/sensor to look at it, it would have been sitting (when installed) in a void between two vertical walls of metal grating.... I don't know if that helps with which switch that is.

Schematics are of a general nature. Heil's NTGM line of furnaces comes in four different sizes (50, 75, 100 & 125)...yours is the 100. The smaller sizes may have 1-2 rollouts, the larger 2-3 of them. The only way to know for sure is looking at the furnace itself...the schematic may not tell you how many there are.

The important thing is to put the jumper at the end of the red wires coming from terminals 1&6 on the board. Doing so bypasses all of the safeties there (limits and rollovers), whatever number of them there may be. If this gets the furnace back on its own two feet then we'll discuss individual troubleshooting to pinpoint the defective safety.

If your unit is an upflow model, it has only one limit and likely no less than one rollover switch (which is also a limit but of the manual reset type). Downflow and horizontal models have two limits and likely one or more rollovers.

Of course there's always the possibility that the problem may be somewhere else, but first things first...we'll test the safeties.

There are only two (main limit and rollover) switches in-line with the red wires.

I jumpered around both on two separate occasions.

The first time I set up the jumpers, it stayed at pilot on the first attempt after I jumpered. So, I thought that I may not have done it well (the clips were a bit large and were a little difficult to place).

The second time I crafted some better jumpers using spade connectors. It behaved and heated the house... until this morning's attempt (which was - I think - a clear message that these two switches are not the issue, since the mornings had actually been trouble-free for the last 2 or 3 prior).

So, that rules out those two switches?

Yes, from the test you have performed, the limits are now ruled out as possible cause for this malfunction.

Before anything else, please check the blue wire that starts at terminal #3 on the board's 6-wire plug and ends at terminal #4 of the Smart Valve's square-shaped wire plug. Let me know how many controls are connected in series in this loop. If only one, that is the pressure switch; if two, the second is an Auxiliary Limit switch. For the explanations below, I will assume you only have one, the pressure switch.

Let's now check and see if the pressure switch is doing what's supposed to. From terminal #3 of the 6-wire plug on the board, there's a blue wire that goes to the Pressure Switch and from the switch another blue wire goes to terminal #4 on the Smart Valve's 4-wire square-shaped plug (note that the smart valve has two wire plugs, the second one being of rectangular shape).

You see, the way a pressure switch is supposed to operate in a furnace such as yours is as follows:
-Prior to the call for heat from the t-stat, the contacts of this switch are supposed to be open.
- t-stat calls for heat, inducer blower kicks ON, and right after, the pressure switch contacts are expected to close and remain closed for the remainder of the call for heat

--> If the contacts were closed prior to the call for heat from the t-stat, the burners won't ignite (the board prevents them)
--> If after the burners are up and firing the contacts of the pressure switch were to open for whatever reason (even for an instant), the board turns the burners OFF

Well...to perform this test you may have to wait until the furnace misbehaves again, but here is what I want you to do when it happens:
(a) switch the t-stat in the house to OFF
(b) unplug one of the two blue wires connected to the pressure switch
(c) put a jumper between terminals W-and-R on the board's t-stat terminal strip.....this simulates a call for heat and the inducer blower will start
(d) put a jumper between the terminals of the two blue wires of the pressure switch.....this will simulate good airflow and proper operation of the inducer blower

if the burners now start and run well, the problem is the pressure switch...or perhaps there's an onstruction on the combustion airways.

If the above 4-step procedure does not take care of the problem, without undoing anything from (a)-thru-(d) above, continue as follows:
(e) put a jumper between terminal "W" on the board's t-stat terminal board and the quick-connector of the blue wire that ends up at the smart valve (terminal-4)

Why we do "(e)" above?
...The t-stat's call for heat is a knock on the door of the board. A call for action.
...The board now acts upon it and sends a signal to the Smart Valve via the wire harness.
...If "e" above solves the problem, then there is a broken link in the communication signal; either the board is not generating it (bad board), or it is getting lost in the wires (loose/defective wires on the wire-harness), or is not being understood by the Smart Valve (defective Smart Valve)

We elimitated the limits as possible defective components. We're now trying to see if the Pressure Switch, wire-harness or board is/are the culprit(s). Later, if the problem is still not solved, we'll perform tests on the Smart Valve. It's a step-by-step approach, sorry!

There is only the one switch, a large silver metallic disc.

Well...to perform this test you may have to wait until the furnace misbehaves again...

Is this necessary, or can this diagnostic be performed any time the thermostat is off (since - typically - any action to "shut off"/interrupt the furnace results in success the next time around)?

Or - by getting the misbehavior - are we potentially preserving the pressure switch in a state of badness?

I do appreciate your laying this out for me... step-by-step is exactly the kind of thing I prefer and I always want to know what is going on in any black box.

There is only the one switch, a large silver metallic disc.
That is the pressure switch then. Your unit is likely an upflow, that is why it does not have a secondary limit

Is this necessary, or can this diagnostic be performed any time the thermostat is off (since - typically - any action to "shut off"/interrupt the furnace results in success the next time around)?
What makes your problem tough to fix is its random nature.
If you can predict with some certainty that the furnace won't fire (typically after a few hours of inactivity), have the t-stat turned to OFF for a few hours, after which go to the furnace and force a call for heat by jumping R & W. If this replicates the "main burners won't fire" situation (hopefully you'll get lucky here), proceed with the rest of the steps suggested in my previous posting.


Or - by getting the misbehavior - are we potentially preserving the pressure switch in a state of badness? The idea is being able to pinpoint if the pressure switch is the one to blame. If that is proven (with the test), then you can replace it.

I do appreciate your laying this out for me... step-by-step is exactly the kind of thing I prefer and I always want to know what is going on in any black box. You bet!

If the 5-step test suggested in post #20 passes with flying colors but the original problem persists, then the limits, pressure switch and ST9120 are likely in good condition. The test that follows check for proper power feed and functioning (or lack thereof) of the smart valve system.

1. turn OFF power to the t-stat and go to the furnace
...(do not turn OFF power to furnace at service switch)
2. remove "control" plug at smart valve
3. place tips of a voltmeter between terminals 1 & 3
...(term-1 is the one with the black wire, term-3 the one with the white one)
...you should read 24+ volts between both
4. keep the tips of the meter on 1 & 3 above, call for heat from the unit by placing a jumper
...between R & W at the board's thermostat's terminal strip. The induce shall turn ON
5. you should still read 24+V at the meter
6. disconnect R-W jumper (inducer turns OFF)
7. move lead from terminal 1 at smart valve control plug and relocate to terminal 4
...(this is the one with the blue wire). voltage should read "0"
8. put back R-W jumper. After inducer is at full speed, meter should read 24+ volts
9. remove R-W jumper
10. put/connect back control plug to smart valve
11. remove igniter plug from smart valve
12. set meter to read ohms at place leads on the terminals at end of the blue wires
...(these are terminals 1 &2 on this plug and are the igniter's ends).
...You should read less than 10 ohms with igniter at room temperature.
13. connect igniter plug back to smart valve
14. connect leads of voltmeter to terminals 2 & 4 of the plug at the ST9120 electronic board
...(NOT THE SMART VALVE'S CONTROL PLUG)
15. put back R-W jumper. Inducer will kick ON and igniter will glow. As the igniter is glowing,
...notice the voltage on the meter. It should read at least 19.5V
16. remove R-W jumper; remove meter

Please report the voltage and ohm readings on the various steps above. Depending on the results reported we will be able to ascertain one (or more) of the following:
a) the trasformer in the furnace is weak
b) a break in one or more of the wires in these wire harnesses
c) a broken/defective ignitor (the glow coil)
d) a defective smart valve
e) a defective flame rod on the igniter assembly

From what I've been reading (NTGM, SV9501 & ST9120 manuals plus your reports), I believe the Igniter/flame-rod assembly is your problem. Your report on the tests above will confirm that.

The following tester from Honeywell would come in real handy:
http://customer.honeywell.com/Honeywell/CatalogNavigator.aspx?Definition=Search&Catalog=&ChannelID=%7B96CDA7E5-4A39-47AC-98EA-E78614E28BD2%7D&ReturnUrl=http://customer.honeywell.com/Honeywell/UI/Templates/CMSTemplates/Base4Column.aspx?NRMODE=Published%26NRNODEGUID=%7B40A4279D-9BFB-4F9E-AC70-76670E0646B2%7D%26NRORIGINALURL=/Business/Cultures/en-US/Default.htm%26NRCACHEHINT=Guest

From Post 20:

Well...to perform this test you may have to wait until the furnace misbehaves again, but here is what I want you to do when it happens:
(a) switch the t-stat in the house to OFF
(b) unplug one of the two blue wires connected to the pressure switch
(c) put a jumper between terminals W-and-R on the board's t-stat terminal strip.....this simulates a call for heat and the inducer blower will start
(d) put a jumper between the terminals of the two blue wires of the pressure switch.....this will simulate good airflow and proper operation of the inducer blower


Before I reveal the results, I need to confirm a couple of things. Will this be a valid diagnostic:

1) if the t-stat *is* making a call for heat?

and

2) if the pressure switch is already jumpered before that heat call comes (i.e. (d) is done before (c))?

The reason I ask is that it is difficult to know beforehand when the furnace will fail to reach full ignition, so I left the pressure switch jumpered and was waiting for the no-heat situation to appear... which I am assuming (correctly?) to mean that I could then proceed to:

If the above 4-step procedure does not take care of the problem, without undoing anything from (a)-thru-(d) above, continue as follows:
(e) put a jumper between terminal "W" on the board's t-stat terminal board and the quick-connector of the blue wire that ends up at the smart valve (terminal-4)

(As an aside, if the heat always came on when in this jumpered state, then I wasn't completely sure that that would pinpoint the pressure switch).

In any event, with the ordering of steps (a)-(d) slightly different, I had a no-heat condition this morning and did step (e)... and still got no heat.

I see you have posted the next clue in this treasure hunt ;) ... am I free to proceed to that at this point?

these electronic boards have an internal self-check feature that will prevent the furnace from firing if prior to a call for heat a switch is closed when it is not supposed to.

In this case, if you jumper the pressure switch prior to R-W, the furnace will not fire. Which I presume from reading your last post is exactly what has happened.

So no, you may not have that jumper in place prior to a call for heat

these electronic boards have an internal self-check feature that will prevent the furnace from firing if prior to a call for heat a switch is closed when it is not supposed to.

In this case, if you jumper the pressure switch prior to R-W, the furnace will not fire. Which I presume from reading your last post is exactly what has happened.

So no, you may not have that jumper in place prior to a call for heat

Now I am confused... well, more confused...:confused:

...because the furnace did fire with that jumper in place.

If it is not there, then I am not sure how I can have any confidence that diagnostic has worked since if the furnace fires, I cannot know if it would have anyhow (given the intermittent nature of the problem).

Is there a way to proceed so that we (temporarily) assume the pressure switch is good and can later prove that assumption true or false?

My bad!
This board is a lot less sophisticated than the average type out there. So yes, if the pressure switch is jumpered, then as soon as R-W are jumpered too, the furnace will attenmpt ignition.

As a matter of fact, with this board, all that the call for heat does is enable an internal switch that will in turn send 24V to the smart valve thru the pressure switch.

What I should have said is that IF one of the safeties is open (limit or rollout) then jumping the pressure switch will NOT fire-up the furnace

So no, you're not going crazy.

So no, you're not going crazier.

There, fixed that for ya! :D

So, am I able to proceed to those next instructions for the SmartValve... or is there a modified diagnostic for the pressure switch?

You may proceed...

BTW, assuming that the safeties are O.K. (limit & rollout), which is a pretty safe assumption since you tested them...
...If you place a jumper between R & W and another jumper between the two terminals of the pressure switch AND by doing so the problem goes away for good, then the board is doing its job and the problem is on the ignition side (either the igniter/electrode assembly OR the smart valve itself).

I am assuming that the flame off the pilot burner is nice and steady and is engulfing the flame rod AND the hood above it.

I'm at work now and our firewall here does not allow me to see/add/download pics. But when I get home I'll post FYI the internal schematic of your board plus a pic of what a healthy flame off the pilot burner should look like.

A bad flame (too small, weak, unsteady or wavery, etc.) often causes problems like the one you're reporting. A bad igniter/rod assembly does that as well.

On the pic below, the solid black lines are actual wires you can see. e.g.: blue wires connect to the pressure switch, red the limit & rollout. The broken (dotted) lines represent internal board connections...those you can't see, they are imbedded in the board. Notice the blue and red arrows, this is how power goes from the transformer, thru the board, safeties, R-W (on a call for heat) and finally thru the pressure switch prior to reaching the smart valve terminal.

the L-shaped rectangle represents the plug at the board
http://i261.photobucket.com/albums/ii64/pflor_bucket/ST9120internalschematic.jpg

Pilot flame pic shown below
http://i261.photobucket.com/albums/ii64/pflor_bucket/pilotflame.jpg

I started on the steps from Post 23 and could only (maybe, sort of) do steps 1-13:

The test that follows check for proper power feed and functioning (or lack thereof) of the smart valve system.

1. turn OFF power to the t-stat and go to the furnace
...(do not turn OFF power to furnace at service switch)
2. remove "control" plug at smart valve
3. place tips of a voltmeter between terminals 1 & 3
...(term-1 is the one with the black wire, term-3 the one with the white one)
...you should read 24+ volts between both
4. keep the tips of the meter on 1 & 3 above, call for heat from the unit by placing a jumper
...between R & W at the board's thermostat's terminal strip. The induce shall turn ON
5. you should still read 24+V at the meter
6. disconnect R-W jumper (inducer turns OFF)
7. move lead from terminal 1 at smart valve control plug and relocate to terminal 4
...(this is the one with the blue wire). voltage should read "0"
8. put back R-W jumper. After inducer is at full speed, meter should read 24+ volts
9. remove R-W jumper
10. put/connect back control plug to smart valve
11. remove igniter plug from smart valve
12. set meter to read ohms at place leads on the terminals at end of the blue wires
...(these are terminals 1 &2 on this plug and are the igniter's ends).
...You should read less than 10 ohms with igniter at room temperature.
13. connect igniter plug back to smart valve

I am less than certain on whether I completed them satisfactorily because a strange thing happened round about Steps 3-5.

Way back in Post 1, I mentioned (or at least implied by my going through the SmartValve troubleshooting flowchart) that I measured voltage on the Control plug.... I mention this only because I have done it before.

So, imagine my surprise when I get 0 volts for a reading between any 2 pins on the plug... and actually between any pin and anything else (i.e. any random piece of metal on the furnace).

As I was puzzling about this, a noise started coming from the inducer which sound sort of like a cross between sparking (I saw no flashes) and a piece of paper getting stuck in a slow-moving ceiling fan.

So, I disconnected the W-R jumper and cut the inducer... then I started getting 24V on 1&3.

I reconnected the jumper and got 24 for 1&3 and 1&4... so maybe I had the switch on the meter not completely set to measuring volts? :confused:

As an aside, the inducer noise seems to be either from a warped case or other, as pushing on the circle that is the end of the motor housing (with the end of the spinning shaft as its center) got the noise to go away... as did having the furnace go to full ignition.

So, the other measurement that I was able to make was at 12... and I got 0 Ohms.

Then I ran into trouble with this step:

14. connect leads of voltmeter to terminals 2 & 4 of the plug at the ST9120 electronic board
...(NOT THE SMART VALVE'S CONTROL PLUG)

I was unclear if these were wires 2 and 4 as labelled on the plug that goes into the ST9120 board (which - to be honest - I did not see labelling for)... or the wires that come from 2 and 4 of the Control plug for the SmartValve...

Also, with regard to the pilot, you said:

I am assuming that the flame off the pilot burner is nice and steady and is engulfing the flame rod AND the hood above it.

The little porthole for viewing this doesn't give me the best view. I see the pilot and the top of the igniter/flame rod assembly... and the top seems nicely engulfed with a largely blue flame.... tho' I can't for sure say that I actually see the top of the flame rod there.

Way back in Post 1, I mentioned (or at least implied by my going through the SmartValve troubleshooting flowchart) that I measured voltage on the Control plug.... I mention this only because I have done it before.

So, imagine my surprise when I get 0 volts for a reading between any 2 pins on the plug... and actually between any pin and anything else (i.e. any random piece of metal on the furnace).

As I was puzzling about this, a noise started coming from the inducer which sound sort of like a cross between sparking (I saw no flashes) and a piece of paper getting stuck in a slow-moving ceiling fan.

So, I disconnected the W-R jumper and cut the inducer... then I started getting 24V on 1&3.

I reconnected the jumper and got 24 for 1&3 and 1&4... so maybe I had the switch on the meter not completely set to measuring volts?

As an aside, the inducer noise seems to be either from a warped case or other, as pushing on the circle that is the end of the motor housing (with the end of the spinning shaft as its center) got the noise to go away... as did having the furnace go to full ignition.

When you were reading "0V", the thing that was stuck inside your inducer fan housing was probably preventing the motor to reach full speed. lower speed means lower airflow, which in turn could cause the pressure switch to remain with its contacts open. In other words, if there is an obstruction to the airflow, the contacts of the pressure switch may not close, and then you have the reason for the "0V". after removing the jumper in R-W and reconnecting, the obstruction probably cleared and you had full flow once again, so the pressure switch contacts closed and hence the proper voltage readings

So, the other measurement that I was able to make was at 12... and I got 0 Ohms.
This reading is likely a MISREADING. "0-ohms" would mean a shorted igniter. If anything I would have expected a resistance greater than 10-ohms here.


Then I ran into trouble with this step:

Quote:
14. connect leads of voltmeter to terminals 2 & 4 of the plug at the ST9120 electronic board
...(NOT THE SMART VALVE'S CONTROL PLUG)

I was unclear if these were wires 2 and 4 as labelled on the plug that goes into the ST9120 board (which - to be honest - I did not see labelling for)... or the wires that come from 2 and 4 of the Control plug for the SmartValve...


The labels are NOT on the unit (board or elsewhere) but they ARE on the schematic I attached to post #10


And how about the voltage on step-15?

At any rate, from your last post it is clear that you should consider cleaning-up the passageways (pipes) of the combustion air. You may even have an inducer blower that is on its way out (from the noise you report it makes)

I'm becoming more and more convinced that your main problem here with the burners erratic behavior is due to a "sometimes" inadequate electrical signal going from the electrode in front of the pilot burner to the metal hood on top of it (the ground electrode). This electrical current is very tiny, in the order of the micro-amperes. A weak or wavery flame or a sooted/rusted flame-rod or ground hood could cause that the electrical current be below the minimum value required for the smart-valve to authorize the main burners to fire (increase gas pressure to pilot burner). A rusted, or loose, or too long a pilot tubing, could cause that too. As could as well a dirty (plugged) pilot burner orifice.


You should have an electrical signal on the flame-rod circuit NO LESSER than 1.3 micro-amps. Anything less than that will cause the problem you're having. But if you do not have a meter capable of taking micro-amp readings, you won't be able to know for sure. About 4 micro-amps is that is needed here for the smart valve to do its job consistently.

I'm wondering if I tossed you a whole bunch of red herrings... or at least described them as such.

The 0V readings were for any of the pins. Wouldn't I have gotten 24V on pin 1&3 (so long as the system was powered on)? I'm honestly thinking that (at 6 a.m. after a not-so-great sleep) the meter was low on batteries or mis-set... or maybe it was just me.

I'm not sure what was happen with the inducer fan. I've never heard it sound that way before (or tonight) even tho' it has been on for longer times with the system going to full ignition. I'll keep an ear out for it, but I'm wondering if it was some flukey thing...

The 0 Ohms reading wasn't so much as misreading as a mis-estimating. My lowly meter only has one Ohms setting (and it has x1000 by it)... so it was likely something greater than 0 but I just couldn't tell what reliably.

I am (hopefully tomorrow) going to be borrowing a meter that does micro-amps and Ohms (without being x1000) so I should be able to get much better readings.

Which brings me to the question of:

How do I measure the microamps when there is only one wire coming from the flame rod? :confused:

As far as measuing the voltage at the board... I guess I better get those glasses I think I'm needing. I had a copy of your schematic in front of me and still didn't see the wire numbers this morning... so no reading for Step 15 yet. Thank you for re-pointing that out to me!

At any rate, from your last post it is clear that you should consider cleaning-up the passageways (pipes) of the combustion air.

Can you elaborate on this... both why you think so and how I would do it (or is it something for an expert)?

I'm wondering if I tossed you a whole bunch of red herrings... or at least described them as such. I got a chuckle. Not having been raised around here but rather way, way south, I don't have a clue what this means. :p

The 0V readings were for any of the pins. Wouldn't I have gotten 24V on pin 1&3 (so long as the system was powered on)? I'm honestly thinking that (at 6 a.m. after a not-so-great sleep) the meter was low on batteries or mis-set... or maybe it was just me. maybe the meter or a bad connection. Especially not being a digital one (or so it seems from what you describe).

The 0 Ohms reading wasn't so much as misreading as a mis-estimating. My lowly meter only has one Ohms setting (and it has x1000 by it)... so it was likely something greater than 0 but I just couldn't tell what reliably.A precise instrument is a must nowadays with the new equipment. There's no other way, unfortunately

How do I measure the microamps when there is only one wire coming from the flame rod? :confused: Connecting the leads of the meter in series with the flame-rod...but you will need to cut the black wire...a "non-invasive way of doing it is by getting an adaptor which Honeywell makes for this purpose
http://img2.freeimagehosting.net/uploads/8d47f20e46.jpg (http://www.freeimagehosting.net/)

If I were is your shoes, I would buy a new pilot-burner/flame-rod assembly. They should be cheap, and will save you the trouble of all this testing, which I'm sure must be wearing you out :rolleyes:
I foresee three most likely reasons for this malfunction: (1) a defective pilot-burner/flame-rod assembly (replace), (2) a lousy pilot burner flame (clean pilot burner orifice and/or increase gas pressure on pilot line), or (3) a defective smart valve.

But you may want to keep your ears wide open for funny noises off the inducer motor. You may have some work there that needs to be taken care of in the near future.

"Red herring" would be something that sets one off on the wrong path. I follow them all of the time! :wall:

Having come this far, I am going to press on. My father-in-law has a good meter, so I will complete those tests and see where we stand at that point.

... and I don't mind interrupting the black wire from the flame rod (I've got a number of other splices in place from diagnostics and repairs in some other kitchen appliances).

Again, I appreciate your time and I'll let you know.

12. set meter to read ohms at place leads on the terminals at end of the blue wires
...(these are terminals 1 &2 on this plug and are the igniter's ends).
...You should read less than 10 ohms with igniter at room temperature.
13. connect igniter plug back to smart valve
14. connect leads of voltmeter to terminals 2 & 4 of the plug at the ST9120 electronic board
...(NOT THE SMART VALVE'S CONTROL PLUG)
15. put back R-W jumper. Inducer will kick ON and igniter will glow. As the igniter is glowing,
...notice the voltage on the meter. It should read at least 19.5V
16. remove R-W jumper; remove meter

So, backing up a little bit, I re-took the resistance measurement for Step 12 and got 3.7 Ohms.

I again got to Step 14 and had some problems knowing exactly how this measurement should be made. If I unplugged the plug at the board, then the blower kicked on.

If I left the plug in, I got 24V without jumpering W-R.

I measured the micro-amperage for the flame rod in the meantime, and (of course) had a bit of trouble (as the probes allowed the current to flow and things went to full ignition... so I had to be quick to see what the reading was during pilot-only).

After several tries (and it seemed to lock-out and provide no reading and the igniter remained unpowered... not sure if that was the meter or the smartvalve), it *seems* like the reading was in the neighborhood of 2.8 microamps or so. It was hard to see at just the right time as it always went to full ignition fairly quickly after the pilot came on.

So, backing up a little bit, I re-took the resistance measurement for Step 12 and got 3.7 Ohms. 3.7 Ohms --> ignitor is good

I again got to Step 14 and had some problems knowing exactly how this measurement should be made. If I unplugged the plug at the board, then the blower kicked on.

If I left the plug in, I got 24V without jumpering W-R. good! that's the expected reading, means that the smart valve is getting the power it needs to perform its duties.

I measured the micro-amperage for the flame rod in the meantime, and (of course) had a bit of trouble (as the probes allowed the current to flow and things went to full ignition... so I had to be quick to see what the reading was during pilot-only).

After several tries (and it seemed to lock-out and provide no reading and the igniter remained unpowered... not sure if that was the meter or the smartvalve), it *seems* like the reading was in the neighborhood of 2.8 microamps or so. It was hard to see at just the right time as it always went to full ignition fairly quickly after the pilot came on.
Shortly after the pilot flame is lit, the micro-amp signal will start and should continue for the length of the heating cycle. 2.8 micro-amps is a good reading...the real test is to see if the signal remains steady while the burners are on full flame. If the pilot flame is unsteady, the micro-amp signal will fluctuate, and a drop below 1.3 would trigger the problem you have been reporting.

From the readings you've reported so far, the board, wiring, smart valve and ignitor/rod assembly all seem to be doing their job the way it is expected of them. You should leave the micro-amp meter connected next time you run this test again to see how steady is the micro-amp signal remains (if at all) and if it drops below 1.3 at some point. If something blows the pilot flame out or makes it waver so that it no longer engulfes BOTH the flame-rod and hood above (which is the ground rod) AT THE SAME TIME, then the micro-amps drop and the smart valve shuts the main burners OFF.

From the above, it looks like your problem here is likely a less than ideal pilot flame (small size?, flickery? weak?)...which would point to either low gas pressure in the pilot line, a dirty orifice (do you know what is it?), a wavery flame resulting from drafts, or a combination of all three.

If the flame is weak, the recommended thing to do is to clean the pilot burner orifice, and to do that you first have to shut the gas OFF The pic below shows you the inside of a pilot burner...the orifice looks like a very tiny cone (like an ice-cream cone), but metallic and with a tiny orifice on it. If it gets dirty, the flame becomes small, wavery or both.
http://i261.photobucket.com/albums/ii64/pflor_bucket/th_pilotsandorifices-PM.jpg (http://i261.photobucket.com/albums/ii64/pflor_bucket/pilotsandorifices-PM.jpg)

But even after you clean it up, you may need to increase the pilot burner gas pressure. to do so, find in the picture below where it says "pilot adjustment under cap-screw". Locate it in your gas valve...remove cap screw, get a small flat-head screwdriver, insert, turn it counterclockwise (I think) to increase the pilot gas pressure and flame size. Needless to say, the unit should be "in full bloom" (firing) for you to be able to perform this adjustment.
http://img2.freeimagehosting.net/uploads/c7abc2e788.jpg (http://www.freeimagehosting.net/)

Shortly after the pilot flame is lit, the micro-amp signal will start and should continue for the length of the heating cycle. 2.8 micro-amps is a good reading...the real test is to see if the signal remains steady while the burners are on full flame. If the pilot flame is unsteady, the micro-amp signal will fluctuate, and a drop below 1.3 would trigger the problem you have been reporting.

From the readings you've reported so far, the board, wiring, smart valve and ignitor/rod assembly all seem to be doing their job the way it is expected of them. You should leave the micro-amp meter connected next time you run this test again to see how steady is the micro-amp signal remains (if at all) and if it drops below 1.3 at some point. If something blows the pilot flame out or makes it waver so that it no longer engulfes BOTH the flame-rod and hood above (which is the ground rod) AT THE SAME TIME, then the micro-amps drop and the smart valve shuts the main burners OFF.

I actually did keep watch of the readings once full ignition started. They seemed to vary in the 6.8-7.4 range, and didn't ever dip down.... which reflects just the eyeball observation of the pilot seeming fairly consistent in size and color once it has lit.

From the above, it looks like your problem here is likely a less than ideal pilot flame (small size?, flickery? weak?)...which would point to either low gas pressure in the pilot line, a dirty orifice (do you know what is it?), a wavery flame resulting from drafts, or a combination of all three.

With regard to drafts... I did notice that the puddy that surrounds the igniter/flame rod wiring as it leaves the combustion chamber is a bit loose. Not gaping, but just seems like it could stick better and be more snug.

I did notice the screw adjustment for the pilot flame on the smartvalve, but it was very tight and - at the time - I didn't want to force it and break something.

Am I correct in that the coupling that gives me access to the orifice doesn't require any special teflon tape or other sealant when it gets reassembled?

6.8-7.4 micro-amps is a very good value of current.

I'm not sure I know what "puddy" is. Do you perhaps mean "putty"? are you maybe referring to the ceramic cover that surrounds the flame rod an isolates it from its holding bracket? Even a minute crack there could cause this tiny electrical current to leak to ground instead of going to the ground electrode (the hood) through the flame...something that would happen in an unpredictable manner (like the failure of your furnace to ignite, which seems not to have a definite pattern)

I'd love to know what happens with the micro-amp reading when the main burners fail to ignite. I think that's the ultimate clue needed to solve this mystery. But you'll need to leave that meter in place until the problem appears...and your in-law may not be to thrilled to leave his toy for too long with you I suppose :D

You won't break anything by putting a bit more force than usual in loosening the cap-screw that would allow you access to the inner screw for pilot-gas pressure adjustment

No sealants of any kind on the threads of the pilot burner and couplings when accessing the orifice. The connections there are strictly "pressure" fit...a small adjustable wrench is all that is needed there.

I'm not sure I know what "puddy" is. Do you perhaps mean "putty"?

Yes... I did mean putty. Sorry, I've only been speaking English for 38 years or so. ;)

are you maybe referring to the ceramic cover that surrounds the flame rod an isolates it from its holding bracket?

The wires that come from the bracket go through the wall of the combustion chamber and on to the smartvalve.

Where the wires penetrate the wall, it is "sealed" with what looks to be plumber's putty... or at least a fair approximation of it. It is still pliable, but I wouldn't be surprised if it wasn't a perfect seal.

I'd love to know what happens with the micro-amp reading when the main burners fail to ignite. I think that's the ultimate clue needed to solve this mystery.

I suppose I could try setting that up tomorrow morning/afternoon. That probably would be instructive.

You won't break anything by putting a bit more force than usual in loosening the cap-screw that would allow you access to the inner screw for pilot-gas pressure adjustment

Actually, I got the cap-screw off (and I did think I had broken it). It was the inner screw that seemed too tight to move.... so, I didn't at the time.

No sealants of any kind on the threads of the pilot burner and couplings when accessing the orifice. The connections there are strictly "pressure" fit...a small adjustable wrench is all that is needed there.

OK... I'll give that a look. I have a leaking water softener (what next?) that may end up taking priority, but we shall see.

Random thought...

Could it be the thermostat in some way? Whenever I jumper W-R, the furnace goes to full ignition.

Very unlikely that the t-stat is the problem.

The t-stat is just a "nice looking" switch, it does the same job as the jumper you've been placing between R & W at the furnace's terminal board, except that it does so automatically, based on the temperature it senses in the surroundings.

If the t-stat was the problem then on a call for heat, the inducer blower won't kick ON, nor the ignitor would glow, nor the pilot light-up.

t-stat puts the jumper in place and that is the end of its share of work. The ST9120C notices and takes action...it's time to work. The board checks that things are safe (safeties not tripped), turns inducer ON and pressure switch instructs the smart valve to get going. Smart valve goes into ignition duty and once the presence of the pilot flame is confirmed (the micro-amp signal), smart valve allows gas to main burners and heating process starts.

And it is in this last step that you're having the problem. The pilot is lit but the main burners won't (randomnly)

so... there's something fishy with the pilot flame, or the integrity of the flame-rod, or the ground electrode pathway, or the smart valve itself (erratic performance). A pic would help. The path to ground is via the pilot-burner gas tube...so it'd better be tight, not loose. And I wonder if your board has a ground connection that I do not know of but may be loose or rusted (you had indicated no ground connection exists at the board).

I didn't say that all of my random thoughts were good thoughts (you should have seen what I originally wrote then wisely edited a few minutes later) ;)

The meter I am using (Fluke 189 series) has a logging feature, so I have had it hooked up several times over the weekend to see if we can catch the reading at pilot-only times.

Of course, the furnace has gone to full ignition every time. :wall:

... and I haven't had time to clean the pilot orifice yet. I hope to do that today.

The one interesting thing that I noticed while trying to take micro-amp readings on the flame rod (both with leads I held there and with clips during the unattended logging) was that the meter would display "Leads" on the display briefly and intermittently. The manual indicates that it might be a fuse issue, but isn't very helpful beyond that.... nor can I find anything on-line. So, I am just wondering if that is indicating some sort of continuity with the flame sensor after all.

Which brings me to another thought... how do we know that the smart valve is correctly interpreting the reading from the flame rod?

How does one ensure the pilot-burner gas tube path to ground?

I am assuming a grounding connection from the board would/should be a (green) wire off to somewhere?

The meter I am using is a Fluke 189 series Good meter!

The one interesting thing that I noticed while trying to take micro-amp readings on the flame rod (both with leads I held there and with clips during the unattended logging) was that the meter would display "Leads" on the display briefly and intermittently. The manual indicates that it might be a fuse issue, but isn't very helpful beyond that.... nor can I find anything on-line. So, I am just wondering if that is indicating some sort of continuity with the flame sensor after all. Here I wouldn't know what to tell you even if my life depended on it

Which brings me to another thought... how do we know that the smart valve is correctly interpreting the reading from the flame rod? If the micro-amp readings never fall below the 1.3 microamps (in other words, if they stay normal at all times), then the smart valve is "randomnly malfunctioning and should be replaced. But if it came to that, and I was in your shoes, I would not only replace the smart valve but the pilot/igniter/rod assembly as well (they may even come altogether in a box)

How does one ensure the pilot-burner gas tube path to ground? by making sure that (a) the pilot tubing connections are firm (and rust free) from the pilot burner to the gas valve, and (b) by making sure the connection to ground from the board is sound as well...but here is where a pic could help, since you report not seeing such connection, and indeed, some of these boards don't have one other than the screws that mount them on their bases. [/quote]

I am assuming a grounding connection from the board would/should be a (green) wire off to somewhere? very good assumption, though not always the case.

"Leads" on screen display? I'll take an uneducated guess; it's telling you something wrong with the leads. Either bad connection with them, or you have them reversed (for like DC testing you have to have the polarity right), or you have chosen the wrong 2 of the 3 holes in the meter to plug your red and black lead wires into for that particular test? Just taking a stab at it.

"Leads" on screen display? I'll take an uneducated guess; it's telling you something wrong with the leads. Either bad connection with them, or you have them reversed (for like DC testing you have to have the polarity right), or you have chosen the wrong 2 of the 3 holes in the meter to plug your red and black lead wires into for that particular test? Just taking a stab at it.

The "bad" connection is what I am thinking as 99.9% of the time the meter is reading (what appears to be) correctly... and I have the manual in front of me (with pictures even :D) to confirm lead wire placement.

Tho' to be absolutely clear, I am supposed to be measuring microamps DC, correct?

I did the meter self-diagnostic on the fuses, and those check out.

So, I am just wondering if this *could* indicate that the flame rod is intermittently failing... sometimes longer than others.

As I mentioned, the furnace hasn't failed in the last couple of days. Other than having the meter hooked up, it has also been warmer so it hasn't gone on as much either (in fact, it didn't fire at all last night). We are supposed to return to cooler days/evenings by mid-week, so I'm hoping that I'll be able to catch the pilot-only problem with the meter.

I prefer to not get TOOOO involved with this thread, as pflor has worked so hard on this, and it is very time consuming, and I am jumping all over this board on so many threads that I can only handle so much. There are regular furnace guys here who make their living at this (I'm a jack of all trades guy), that should be able to help out, also.) But I wanted to share a couple things that may help steer the thread to a possible solve after quickly reviewing some of the stuff said.

I prefer to not get TOOOO involved with this thread, as pflor has worked so hard on this Ecman: the more, the merrier... I sure do not claim ownership of a thread. just here to help when possible :thumbup:

Tho' to be absolutely clear, I am supposed to be measuring microamps DC, correct?
Correct! and the exact term to describe the type of electrical signal that is being generated, is: rectified DC current

Well, the orifice appeared clean... very shiny, no carbon or metal flakes and a nice hole down the middle.

I was, however, surprised to see a dull layer of carbon on the flame rod as I'd cleaned that a month or six weeks back and thought I would have noticed the dullness. So, a bit of Scotch-Brite and a minute later and it was all back to being shiny and smooth.

Could it have gotten that carbon layer again so quickly, or did I just do a poor job the first time around?

I also measured the resistance in the pilot tube (from the locking nut at the pilot assembly to the locking nut at the SmartValve) and got something in the neighborhood of 1 Ohm.

The only other thing of note was a faint whiff of gas when I turned the gas supply back on after re-assembling everything. It was clearly coming from the SmartValve. I'd smelled a slight trace before, so I assume that this is normal?

I think you may just have found out the reason for all the grief your furnace has been giving you, namely "the soot build-up" on your flame-rod.

Yes, they do soot-up rather quickly, 6 months to a year. And if the pilot flame is a bit on the weak side, it is also sootier...perhaps your pilot burner gas pressure needs some strengthening after all. The stronger the flame, the more air it will induce and the more bluish (and less sootier/yellowier) it will be.

And the trace smell of gas you noticed is quite normal.

resistance measurement is also O.K.

The good news (you see, there's a silver lining to everything in life) is that after all this work, by now there is nothing about your furnace that you do not already know.

Now for the interesting observation...

After going to bed, I realized (OK, it was strongly suggested to me ;) ) that I hadn't actually confirmed that the pilot would come on and stay on.

So, I turned the thermostat up and went to the basement, looked in the porthole, and... it was only pilot.

And...

The pilot was mostly orange now. :eek:

I powered the system down, and restarted it.

This time the pilot was mostly blue - but there was orange in there that I don't recall being in there before - and the furnace went to full ignition.

It also fired up nicely this morning, tho' again the pilot is looking a little more orange than before.

Since we had not had the furnace service for a couple of years (at least), I'm still wondering if something else is going on....

Probably nothing unusual there. Just make sure that the threaded connections at the valve and pilot are nice and tight.

Orange flame? you may need to increase the pilot burner line pressure by a tad. A slightly higher pressure will do it.

The furnace again failed to go from pilot to full ignition this afternoon. The thermostat was turned off and left that way for a few hours, so I hooked up the meter to the flame sensor when I got home and turned on the thermostat.

And it stayed at pilot!

The readings bounced around a lot (between 3.2 and 4.1) with the following characteristics:

- for periods of a couple of seconds it would "settle" in the 3.55 to 3.8 range... these stable periods would be approximately 20% of the time

- the rest of the time, it bounced around several times a second

- it only went over 4.0 maybe one out of every twenty readings (so once every 4-5 seconds and then only for a few tenths of a second at best).

I tried to adjust the pilot (which seems back to its blue self this evening) but found a couple of things confusing (big surprise, I know ;) ).

The documentation says:

1) Controls are provided with adjustment screws
set at the maximum pilot flow position.

and

2) Turn the inner adjustment screw clockwise to
decrease or counterclockwise to increase the pilot
flame.

So, I find that I cannot turn it clockwise at all.

I *can* turn it counterclockwise, but this does not yield a larger pilot that I can tell.

Also, if I go too far counterclockwise, I start smelling more gas and the adjustment screw gets very loose.

So, I put it back the way I found it (i.e. so it can't go clockwise any farther).

The one thing I did notice with the pilot that - while it is now pretty much completely blue - it is not symmetrical. That is, it "leans" away from the flame sensor (this is when it is by itself or surrounded by fully-ignited flame jets).

Could there be some sort of draft or misalignment of gas supply relative to the flame rod/igniter?

Am I pretty much at the stage of getting a new flame rod/igniter assembly?

Fluctuating micro-amps is no reason for concern, so long the reading does not go below the minimum Honeywell indicates (1.3 I believe).

When it comes to tests, all that could possibly be tried has been done by you with regard to your furnace. In your place I would buy a new smart valve & ignition set (igniter+electrode set) and replace first the ignition set to see if it fixes the malfunction...then proceed with the gas valve replacement if it does not. Smart valves do go bad.

The unplugging and re-plugging of wire harnesses, if not done gently, may also have a negative effect on the soundness of the electrical continuitiy of them (though I don't think that's a reason for concern in your case). Hopefully a new smart valve comes with new wire harnesses. If so, consider replacing them too.

Fluctuating micro-amps is no reason for concern, so long the reading does not go below the minimum Honeywell indicates (1.3 I believe).

Nope, never went below 3.2 when I was watching it... just never got above 4 for more than a tenth of a second or so.

I'm leaning toward getting (if available) the whole pilot assembly (flame rod, igniter, ground electrod, bracket, pilot outlet) as the pilot outlet (not sure what the proper name is... the bent metal where the pilot flame originates from) is very corroded anyhow.

I appreciate your taking the time to get me through this. While I was leaning toward the replacements you've suggested initially, it is nice to have done the work to confirm that they really are good candidates for the problem.

After an unfortuante stretch of warm weather :D in which we didn't use the furnace, I finally got around to installing a new igniter/flame sensor.

And...

The furnace failed to start the first time.

However, since then, it has actually started about 80% of the time... which is better than it was doing in April.

So, I'm thinking that that was at least part of the problem.

I expect to have a new valve installed by mid-week, so I'll post back on what happens afterward.

After a couple of days with a new SmartValve installed, the furnace has not missed an ignition.

So, once again, thanks pflor for your painstaking effort to arrive at the proper solution. I hope your customers appreciate your approach!