I am new to this forum, but I will do my best to answer any questions I can (I am a general contractor with a background in engineering and residential construction and machine design)

so, here is my very first question:

how many times should a regular residential gas furnace (not a high efficiency furnace) attempt to light the burners? mine is cycling through the start process (pre-light, ignitor energize, combustion, and burn for a few moments but then go out) four or five times before its achieving stable constant gas burn. the ignitor works fine (new last year), the temperature probe is clean and must be working (as it can actually achieve steady burn), and I am not getting any error lights or codes on the control board. interior of the gas chamber and everything in general is clean and in good repair. and yes, the power is on and the filter is clean.

its never been a problem, I always have heat, but it seems like it should light and maintain on the first cycle. otherwise, you are decreasing lifetime of all the ignition parts.

anybody have any ideas why its doing this or whether or not its normal?

thanks in advance

  • 1
    Once the furnace achieves combustion, it should not shut down until the thermostat is satisfied. If it's shutting down early, there's a problem with the system. However, without knowing more about the system (make and model), and a detailed description of the whole cycle (including the state of all the limits throughout the cycle), it's difficult to provide an accurate answer as to what the problem might be.
    – Tester101
    Nov 23, 2015 at 16:35
  • What's the make and model of the system?
    – Tester101
    Nov 23, 2015 at 16:37
  • thanks for the response. its a goodman gmp100-3 revision b. basically, to rephrase what I said previously, it runs fine once lit, but the start cycle repeats for some reason. when the thermostat is set to fire, a) the exhaust fan relay kicks on, and b) the exhaust fan starts. then c) the glow type ignitor energizes. d) gas valve solenoid trips, gas flows through the venturis and ignites. so this is all normal operation. however, then the gas stops, and there is a few second delay and then everything from c) to d) repeats. it seems to do this 4 times, and then stable constant burn... Nov 24, 2015 at 20:11
  • ...no lights other than the little single red status light on the board, which blinks just before each activity occurs (seems to be single digital pulse). I was concerned about a cracked heat exchanger, but I don't see any of the telltale backflow or flame licking into the venture chamber that you typically see with that. it all seems normal, except for the "fourth times the charm" sequence. Nov 24, 2015 at 20:13
  • just one more thing. just watched it again, and its actually three times that it tries to light, not four. which Is even stranger as I just read in the manual that it would lock itself out for one hour if it goes beyond three times attempting to light. Nov 24, 2015 at 20:17

1 Answer 1


I consult to most all warm air furnace OEMs. The best answer is it lights the first time, every time. There are of course things that can prevent that...low gas pressure switches, low air pressure switches, limits not being made (door switch, air switch, fan speed proving, condensate switches, etc.). That being said...EVERYTHING good...GAS supply...AIR supply...SUFFICIENT spark/ignition source (all three parts of the fire triangle), begin to look at the flame sensing. Very early systems were single rod...hot surface ignitor glows & lights flame...then switches to sensing. This system works well..for about a year and degrades and will not sense well. Junk them. The two rod system has an HSI to ignite, and a flamerod to sense. These are great. HSI must get fully to temp to light. If it actually "lights" the flame everytime..it is good. The flamerod sends the signal to the flamesafeguard...which powers the gas valve and monitors all the limit switches. Good signal..keeps valve open...heats the house. There are specific meters to check flamesignal (DC microAmps)...but they may be difficult to read if the signal is manipulated by the flamesafeguard manufacturer (also must be read in series). Easy fix, take a piece of very fine emory paper and clean the oxidation from the flamerod. It will give a higher signal. A flamerod light two or three times will help warm it up and boost the signal...hence what you have. Oxidation will kill the signal. There is some science to flamerods & grounding area ratios. Typically 1:4..but your OEM figures that out for you. Make sure your green ground screw is actually grounded...this will boost your signal. There is one somewhere near the actual burner...and one on the cabinet. If the clean flamerod (properly located & installed) fixes your problem...get a spare. Cleaning them causes micro grooves...more surface area to oxidize. Nobody wants to clean a $7 widget everytime the house gets cold. Best Wishes.

  • thanks for the response. my original problem was solved months ago by replacing the flame sensor. so your solution is pretty much accurate for the problem i had. thanks again Feb 8, 2016 at 18:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.