1
\$\begingroup\$

11/17 results: Manually turned on heat on the thermostat (you will know what I mean) - lit right up.

Thank you all yet again. I have heat...which I will be turning off at night and keeping a fire detector for safety reasons until I figure out the short fully, but nonetheless.

enter image description here

enter image description here


Background:

  1. Turned on furnace from the 60s or 70s (guess, don't have model).
  2. Realized it's not heating.
  3. Went to panel. Every 5 seconds there was a spark on what I presume to be a gas venting pipe (yes, I realize the danger). The pipe was very hot, but did not shock me. The spark gap was bad enough to essentially cut into the pipe - there are metal shavings under it.
  4. I inserted a toothpick between the pipe and the metal bracket (attached to furnace body) holding it to insulate.
  5. Spark went away, pipe stopped heating/cutting electrically. Other random components started heating up.

Assumptions:

  1. A wire is stripped or poorly attached somewhere, leaking high voltage onto furnace body.
  2. Going through the wires one by one, checking insulation and proper contact on splices will fix the problem. There is no "fast" / methodical way to troubleshoot this.
  3. A bad ground is likely NOT the problem, as ideally, there shouldn't be voltage going to the furnace body itself.

Questions:

  1. Are my assumptions correct? (Assume 101 understanding of electronics - I understand voltage, current, resistance, DC, relays, fuses etc, but I don't REALLY understand things like voltage converters or why a car battery or a furnace is grounded, for example, unless it's purely a safety mechanism that doesn't kick in unless there is a failure).
  2. Any overall advice on trouble shooting this?
  3. What safety precautions should I take? Other than the obvious - cut power, cut gas, wear insulated gloves, etc.

P.S. Obviously I will do this after the furnace is off, which it currently is, with the circuit breaker to it turned off as well.

P.S. 2 - an electrician and an HVAC professional looked at it, charged me $80 each and said it's not worth paying them to find what's wrong on such an old unit. This is the same story I get with old cars when there is a $2 30-minute fix.

Not my furnace, but looks like the same model. enter image description here


Wiring diagram from my furnace:

enter image description here

Transformer (bare hands because breaker is turned off and input wire is completely disconnected):

enter image description here

enter image description here

Removing transformer:

enter image description here

Transformer input wires are spliced with a bunch of others. I had false expectations of single point connection to main.

enter image description here

11/14 Updates: I have to admit that I had expectations of magic connection terminals that connected via a metal loop and a bolt, rather than spliced wires. Nothing meaningful to report back yet, still working on it. So much for my 2 mins comment.

Disconnected:

enter image description here

Updates on 11/16:

  • Did as John instructed (all tests with gas cut off, burner not on) - disconnected the two input wires to the transformer entirely, capped and isolated them, put everything else back together as it were.
  • Turned on breaker, tried to turn on heat - nothing, no voltage to any of the components that were heating up before.
  • Reconnected trans input wires - turned power back on, turned heat back on - nothing, no voltage anywhere.
  • Edit: 20 minutes in - no voltage, no spark, transformer body is starting to heat on top. Barely warm so far.
  • 50 minutes in - no voltage, transformer is getting warmer. All other parts of system cold.

My only guess is that I should have waited longer. I am now suspicious that the short starts appearing when the components heat up. Letting it run with no gas to see if problem comes back. If it doesn't come back in an hour, I will ventilate, check for gas, and turn the burner back on.

Note that I don't think that I magically "fixed" anything - all the connections looked really solid and properly isolated before I touched them. Also, sorry for the slow replies, I have been swamped at work.

P.S. - why does the transformer have two input wires? Also, if I can buy a replacement transformer for less than $50 and there is 40%+ chance that it's the problem, I think I want to try that next.

11/17: It ran for several hours, transformer heated up, but not anywhere near being hot enough to burn to the touch. No sparking. Turned on gas and lit it. Not seeing flames to heat exchanger. Next guess is whatever is responsible for gas not opening it.

\$\endgroup\$
14
  • 2
    \$\begingroup\$ The problem isn't the "$2 fix." It's the hours long search for the cause of the problem, then being held responsible when the ancient pile of junk blows up next year because something else wore out. \$\endgroup\$
    – JRE
    Nov 14, 2020 at 20:55
  • \$\begingroup\$ Yea, I get that. I shouldn't have made that comment, but I am going to leave it in the post for the sake of the existing answer making sense. Nonetheless, I think there is an optimal solution here that strikes a happy medium between expense and reliability. I am willing to put in the time myself - is my general approach correct? \$\endgroup\$
    – VSO
    Nov 14, 2020 at 20:58
  • 2
    \$\begingroup\$ OK let's troubleshoot. Does it only spark when the stat is on (call for heat?) Can you put a test light or meter across the spark-spot. The 24 voltAC transformer may have a short between the windings. During testing you could shut off the gas or remove the wires going to the gas valve. These are first steps to finding the problem. \$\endgroup\$
    – John Canon
    Nov 14, 2020 at 23:49
  • 1
    \$\begingroup\$ This should definitely be closed as a consumer repair question - it's designed for cases exactly like this. \$\endgroup\$
    – pipe
    Nov 17, 2020 at 1:25
  • 2
    \$\begingroup\$ @VSO Everything these days has electronics, but consumer repair questions are off-topic when OP doesn't already know electronics. Such questions always turn into a long discussion about "did you try X?" and are way too broad, often requiring proprietary documents (basically to stop the flood of "My expensive TV broke, help me fix it" questions). I don't know where to take it. Stack Exchange is not meant to cover every question anyone can have, so maybe there's no place at all. It sounds on-topic on Home Improvement. \$\endgroup\$
    – pipe
    Nov 17, 2020 at 1:59

1 Answer 1

1
\$\begingroup\$

As both an experienced furnace man and a licensed auto mechanic, I am doing the math: a $2 fix that takes 30 minutes means I could make $4 per hour, if I had the time and inclination. If the furnace can have one worn-out part, it can have more worn-out bits, probably hidden until they bite. Looking for the problem will probably uncover other deficiencies, as both those professionals suspect. If you need to keep this furnace going, it needs professional TLC.

\$\endgroup\$
9
  • 2
    \$\begingroup\$ Not a useful answer. This site is a spin from stackoverflow, which wouldn't exist if everyone replied with "if you want a reliable site, hire a professional programmer". You are an auto-mechanic? This will sound familiar to you then - was told to replace ECU, turned out to be an O2 sensor. Was told engine needs replaced, needed to clean throttle body. I realize it's easier for you to sell me a new furnace than to troubleshoot some wiring - but I would rather pay you $80/hr to troubleshoot a short than pay $5000 for a furnace. Anyway - with no hostility - your answer does not help. \$\endgroup\$
    – VSO
    Nov 14, 2020 at 20:49
  • 3
    \$\begingroup\$ @VSO : but it's the right answer for this case. Stay safe and good luck with it. \$\endgroup\$
    – user16324
    Nov 14, 2020 at 20:51
  • 1
    \$\begingroup\$ @VSO To remove transformer wires use electrical tape to hold the bundle together before unscrewing the wire nut all the way off. Then you will see the copper ends twisted together. Untwist and remove only the wire going to the transformer. Tag and mark everything. Twist the wire nut back on the remaining wires, but not loose. Do the same with the other transformer wire. Both wires on the PRI side of the coil are now disconnected. Now you have disconnected the transformer out of the circuit. Complete step 1 as I had described. I will edit my earlier answer if you have more questions. \$\endgroup\$
    – John Canon
    Nov 16, 2020 at 0:54
  • 1
    \$\begingroup\$ @VSO Use history and symptoms to understand the specifics. Use the schematic diagram to see the building blocks of operation. Don't just change parts by guessing. Divide and conquer: disconnect certain subsystems to see if the problem stays or goes. By confirming the good, you get closer to the bad. Good luck! \$\endgroup\$
    – John Canon
    Nov 17, 2020 at 4:11
  • 1
    \$\begingroup\$ @JohnCanon See top of post!! Ty. Sorry to bother you again, but had to share. \$\endgroup\$
    – VSO
    Nov 18, 2020 at 1:30

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.