Consistent failure after one cycle... troubleshooting a furnace

Discussion in 'General Electronics Chat' started by s_mack, Nov 27, 2012.

  1. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    I tried on HVAC forums, but I think the issue is electronic so I'm trying here.

    I have a furnace where all mechanical problems have been ruled out. The interesting thing is the consistency. The system ALWAYS runs normally after a hard reset (power removed for > x minutes... not sooner [still attempting to determine x]) for EXACTLY one cycle (that is, for as long as the thermostat tells it to run) and NEVER runs again after that one cycle, until another hard reset.

    Does that ring any bells as to where I should start looking? I mean, does that sound like a bad capacitor somewhere? a uC problem? Some stuck relay?

    I realize this is a bit of a bogus question with WAY too many variables and I'm not looking for a concrete answer... just perhaps an idea from a different point of view.

    I can replace the main circuit board, but its very expensive, hard to get, and not guaranteed to solve the problem.

    Thanks. Its cold.
     
  2. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    Show that circuit board.
     
  3. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    If it's a gas furnace, I'd look for a delay transfering flame from the pilot to the main burners. This is caused by accumulation of rust particles in the burner tubes. Take them out and tap them on the floor to drain the rust powder. That's assuming you don't have a complete pilot failure.

    Of course, telling us what the fuel is might help.
     
  4. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    Yeah, its gas. The gas fitter (at least says he) ruled out all mechanical causes. That is, he checked the gas pressure, physical characteristics of the ignition system (positioning, flame performance, etc), stuff like that. I would suspect that also means he checked for rust in the tubes. Its a newer furnace so there's no "pilot" per se.

    Regardless, the odd thing (to me) is how it always works after a power cycle. If it were rust or something physical then that wouldn't be the case. 100% of the time after shutting down the power for > 5 minutes... it works. But it works exactly one time. From that successful ignition to the moment the thermostat indicates it should shut off... all is golden. But the very next call for heat fails. That pure consistency is what has me suspecting something electronic. I was just hoping that kind of behaviour might trigger an "aha" moment in someone's head.

    Not that I expect anyone to go through it, but here's the service manual: http://icpindexing.toddsit.com/documents/086477/44108201001.pdf

    What I'm seeing is pretty much exactly as the sequence described under the heading "Heating Request with Gas Supply Line Shut Off".

    - Steven
     
  5. mbohuntr

    Active Member

    Apr 6, 2009
    413
    32
    Had a new utica gas boiler about 8 years ago, would not fire in the fall (the first heating cycle). Had the tech come out and troubleshoot all the mechanical interfaces without finding anything. He suspected a voltage surge hit the board. He also directed me to the distributor for the board, it was very easy to replace, a cable header and a couple of screws. The tech said they charge $300 plus for a controller replacement. :eek: Thanks buddy!!! :D I'd buy him a case if he was still around... (passed away shortly after he spoke w/ me. ) Good luck! Got me a whole house surge protector now!!!
     
  6. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    I have a Bryant gas heat/air central unit that does just that. It's a control board (on mine anyways) that's doing it. I've had it out twice to repair broken traces from vibration(?). Anyways it's acting up now and I'm going to replace it, sick of messing with it. Plus wife is real unhappy, since it usually happens when she's home and I'm not.
     
  7. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    Sorry I can't be of more help. I live in Florida and have met exactly one balky gas furnace in 40 years. This site has members from all over the world, but how many of them do residential appliance repair?
     
  8. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    Well, to be fair... I never really asked anything about a residential appliance. I described the situation and asked for it to be thought of as a circuit.... any circuit... and what might come to mind as a cause for the behaviour. From a general electronics point of view, not appliance repair. It shouldn't matter if its a furnace, a car, a rocket ship or a child's talking toy. The point is that it works on a hard reset, exactly once, then fails. That sort of consistency, I would have thought, might give clues to people who work with electronics on a daily basis.

    It was a wing and a prayer. I don't "expect" a solution here.
     
  9. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    In the case of my heater, the ignition fails to light the gas , a trouble led blinks 8 times indicating that. The only way to clear the fault is to cycle the power. In my case the unit usually fires after reset (on a good day) and heats the house. When thermostat calls for heat again, it usually fails again.
    And the vicious cycle continues.
    Does this sorta describe what is happening to you?
     
  10. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    I'll try to get some pictures when I'm back there.

    It is a hot surface igniter. I replaced that (and the flame sensor) to no improvement.

    Re: overheat limit sensor... if its completely buggered then I'd expect it would trip always, even after a hard reset. Or it would shut it down in short order after a hard reset. If its botched to the point that it lets it get "sorta hot" then it wouldn't cycle for infinity afterward... it would "cool down" and let the fire happen at some point, but it doesn't. OK maybe the control board prevents any fun from happening until a hard reset because it acknowledged the sensor... but then (if logic prevailed at all) it wouldn't keep trying every 5 minutes.
     
  11. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    What I meant was that people that fix gas furnaces for their day job would be more likely to know what this symptom points to. One of them might even look at the brand and model and plunk the answer right in your lap. Gerty might have already done that. Personally, I use my electronics knowledge as my "secret weapon" so I can treat anything, cars, furnaces, talking toys, as a bunch of circuits. It works almost all the time, but there is nothing like experience to speed up the diagnosis.

    Basically, I make my money by looking into a mess of colored wires and figuring out what they are intended to do, knowing all the time that one of the circuits isn't doing what it is supposed to be doing. It takes time to figure out each circuit and eliminate it as the problem by proving it still works. Eventually, I find one that doesn't do what it is supposd to do, and that is where I find what the machine needs.

    On machines that I am familiar with, I can eliminate most of the circuits by simply thinking. On machines I am not familiar with, I have to learn how this one was designed, and then eliminate circuits from the list of possible causes. Like most of the people on this site, I have a basic block diagram of a gas furnace in my head. It's the details of your exact model that are needed to narrow the search down. Standing flame pilot? Hot plate ignitor? Spark ignitor? Does it use a photocell to see the flame or a thermocouple straight to the gas valve? Is the the stack switch wired as a latch-out safety? These are the details one doesn't know until he meets the machine in person. The more you tell, the better we can logic this out.
     
  12. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    Obviously I started there! No, my first stop wasn't "all about circuits" to troubleshoot an HVAC issue. My first call was to a pro... who promptly didn't fix it and left me several hundred dollars poorer. I have another one coming on Friday. In the meantime, I've posted on dozens of HVAC and DIY forums and Googled the crap out of the interweb. Conventional help didn't work... so I came here.

    The "obvious" problem (according to the "pro" as well as most forums, etc) is a dirty flame sensor. He tried cleaning it but broke the ignitor in the process so I ended up buying a brand new one. It didn't solve the problem. OK, so the next "obvious" problem is the control board. In the past few hours though, I've read that on this model that does little more than control the fan and communicate with the thermostat (which was previously replaced/ruled out).

    What it seems to be is the "smart" valve that this unit appears to be plagued with. That's several hundred more, and the part can't be here until next week.

    In the meantime, we continue to freeze.
     
  13. crutschow

    Expert

    Mar 14, 2008
    13,000
    3,229
    As a temporary workaround you might buy a timer with multiple on/off cycles and just have it periodically turn the power to the furnace off and on for a short period to reset. It might occasionally turn the furnace off when it is running, but that won't hurt anything.
     
  14. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    Hey, now that's an idea! Thanks.
     
  15. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    OK, so here's my plan: scour through the furnace manual again and find out what has voltage (or lack thereof) when, and only when, the problem condition exists. Once I have that, voltage divide that down to something an AVR can handle on an input. Use that to start a timer and trigger a relay to shut down the system for a predetermined amount of time and then turn on again.

    I didn't think a simple timer would do it because I just don't think that would allow enough "working time" to get the house actually heated. I should be able to rig this up this afternoon. Hopefully that'll at least keep the home warm until the HVAC guy gets here on Friday with a new valve (I'm gonna try swapping that $300 puppy and crossing my fingers)
     
  16. crutschow

    Expert

    Mar 14, 2008
    13,000
    3,229
    I'm not sure I understand your concern. Just adjust for an OFF time of a few minutes to reset the furnace and an ON time of the maximum minutes the furnace runs when it is on. I would think that would allow enough ON time to keep your house warm.
     
  17. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    The issue is that they won't match up. Its not like the furnace runs 24/7 or at any predictable times. And it won't run for very long before it craps out, so I think I'd have to be pretty efficient on the timings to keep it on enough to heat the home.

    Let's say I have it shut off every hour for 10 minutes. There's two problems with this. a) I get (say) 15 minutes of good run time now I'm wasting up to 45+10 minutes where, potentially, it should be on but isn't before the next reset. House gets cold. b) I'm 2 minutes into the good run time and it shuts off for 10. That interruption means house gets cold.

    considering the furnace could get its heat calls at any time really, there's no way to set up a simple timer that's not going to interfere.

    Lol... lord I hope there's not really a problem. Forcing it on now sounds scary.
     
  18. crutschow

    Expert

    Mar 14, 2008
    13,000
    3,229
    This work work if you have a timer with a repetitive sequence of the desired period (which may be difficult to find). For example shut it off for the minimum reset time perhaps every 15 minutes or so, whatever seems a good sequence. I would think that would give you enough ON time to maintain the house temperature.

    Another approach would be to connect a room thermostat to a relay controlling the power to the furnace. It would have to have a separate power supply for the thermostat since they are usually powered from the furnace. A simple non-electronic (mechanical) type thermostat would likely be easiest to interface to. I would assume the down time between the thermostat going from OFF to ON would be sufficient to reset the furnace.
     
  19. s_mack

    Thread Starter Member

    Dec 17, 2011
    187
    5
    I tried my plan... it seems to be working well enough, although I uncovered (or rather "realized") another problem. I was mistaken before that it would turn on once and run until the thermostat told it to stop. It was just coincidence that it would shut off around the time the thermostat would have told it to stop, so I didn't realize that its actually shutting off on its own while the thermostat is still calling for heat. So I think when it gets really cold, even with my efficient switching, it may not be able to keep up.

    Anyway, I feel clever for what I did :) I used an Attiny1634 that I had leftover from a recent project. I couldn't find a wire that was charged (or not) *only* when the problem was occuring... so I tapped into the LED that flashes the error message. It flashes regularly when everything is operating normally. That LED line goes to an input on my AVR and the AVR "listens" to the light. When there is a delay over 600ms (which is only when there's a problem), the AVR triggers a servo (that I stole from my r/c airplane) and it hooked up to a servo tester (simplified the control of the servo since I'm just hacking this together until Friday) and the servo physically flips the emergency shutoff switch. Lol... I did it that way because I couldn't figure out how to turn off the power considering its a/c and like 40 amps so I figured this was the safest and simplest way. "on" moves the servo one way and "off" moves it the other.

    Its fairly precise in that it turns the power off within 2 minutes of the problem occuring (that's how long it takes the furnace to indicate a problem), waits 10 minutes (ironically, that was the hardest part of all this!) and then turns it back on. I had to play with the placement of the servo arm quite a bit... but now it seems to work 100% of the time.

    However, given the limited run time... I'm not sure it will keep the house comfortable especially at night.
     
  20. GetDeviceInfo

    Senior Member

    Jun 7, 2009
    1,571
    230
    if the system is performing a softlock, it would indicate the fault it sensed. What is the indication. You may have to watch a shortened cycle as indication may not be continuous.
     
Loading...