And so we come to the latest1 chapter in the sorry Saga de Dryer.
As reported here, the dryer was once more hors de domestic usefulness due to some unspecified annoyance. To recap: It was tumbling and blowing air out of the vent, but said air was cold and has little going for it in the speedy drying of the clothing department. Action was called for2. The burner appeared to be non-working after the usual lying-on-the-floor-with-my-eye-to-the-spyhole session.
I took stock of the situation. I had replaced the coils that turn the gas on and off. I had replaced the most complicated thermostat and I had tested them all during the last session with the damned machine. A failure in one of them was unlikely. Part of the decision-making process involved in these diagnostics was my admitted reluctance to remove the back of the machine again. Or to put it another way: I wasn't taking the back of the machine again unless hell itself froze over. Research showed that the next likely culprits would be the igniter (sort of like a small kettle element, it glows red hot on command to set the gas on fire) or the thermal sensor (which makes sure that the gas doesn't get turned on if there is no chance of it being lit).
Getting at these items, both components of the burner assembly, would involve essentially the same process as that used when I replaced the burner solenoid coils last July. The two screws securing the fan/lint filter housing to the top of the casing would have to be removed, the front levered up and hinged back and the front of the casing removed. This would involve uncoupling the door sensor, the one that turns off the machine when you open the door so all the clothes don't get thrown on the floor. This in turn would necessitate separating one of those multi-wire bayonet-style nylon couplings that have an integral snap-lock built in.
Everything went about as expected, which is to say I slashed myself to ribbons on the razor-sharp panels and found that no matter how I squeezed, pulled or swore the nylon electrical coupling would not separate. I ended up using my trusty Leatherman tool, which had a sort of screwdriver-cum-paint-can-opener on it that I could use to pry the nylon catches apart. It worked like a charm, and in some time at all I was able to hurl the front panel, door and all, away from me and the machine with a triumphant howl of manly triumph. I took it as a good omen that the drum did not fall out of the machine like last time, scattering bits everywhere, but remained properly mounted on its rear bearing.
I mentally made a note of the position of everything and reluctantly concluded that I would have to dismount the drum anyway in order to get at the burner, when fate intervened to save me much valuable time and the drum fell off its rear bearing, striking me smartly on the right kneecap with a resonant bong!
I took the opportunity to gather certain tools as I hopped around the basement clutching my knee to my chest with my mouth. Once the agony had subsided to manageable levels, it was time to start.
I whipped off the igniter, but it stupidly showed no signs of having burned out. No black marks, no gaps melted in the element and, most damning of all, the correct no-load resistance when measured with my new multimeter3. So the burner was nominally all right. This was disappointing in that I hadn't fixed the problem at the most likely fail point, but I was buoyed up by the fact that there was another component still to be tested, and that at least I hadn't had to remove the back of the machine.
Next up was the sensor. I disconnected it by gripping the little brass "spade" connectors with my pliers and heaving with all my strength. In only ten or twenty minutes of cursing I had both wires off and was able to measure the resistance of the sensor, which is essentially a switch with a resistor in it. The reason the wires have to be pulled off is that if you don't you measure the resistance of the component in parallel with all the other circuit components, which can throw off the reading, sometimes by a lot. It also acts as a guard against destruction of cheap meters and the possible loss of life caused by poking the test leads onto live circuits mistakenly left plugged in due to the murderous perfidy of Mr Brain. No stranger to electrocution, me.
The sensor checked out as okay. There was only one thing for it: I would have to run the machine and test the various voltages floating around it when it was in operation.
This raised several important concerns. Firstly, there was the concern that when testing for volts they are always one distraction away from applying themselves to one's body with potentially lethal results. Not optimal. Next there was the concern that the machine was in parts, but there should be no problem running the motor for a short time with no working load (sometimes they are designed to work under load and can be burned out by running them freely). Finally there was the issue of the bloody door sensor, which would have to be reconnected for the test since the disconnected wire was pretending to be an open door which prevents the machine from running.
This was a bit of a poser, since the wires in the socket were not straightforward enough to short out with bits of wire4 and the wire in the front panel of the cabinet wasn't long enough to reach with the panel dismounted unless I wedged my foot under it.
In the end, the need to be far from the volts along with the fact that my foot did not articulate in such a way as to allow me to be prone on the floor in optimal test-lead poking position while at the same time supporting the front of the dryer - together with the annoying realisation that my arms were four feet short of being able to reach the controls of the dryer from that prone position - provided the answer: I would attach the test-leads to the bit that needed testing and do all the other stuff while standing. First up: testing that the igniter was getting volts when it should.
For a wonder it worked like a charm and in no time I had the rather puzzling datum that when the motor was started there were 18 volts RMS5 on the burner all the time. Now I don't know why 18 volts when there should be none, but there should have been a swing to 110 at some point and there wasn't. I triumphantly concluded that this meant there was an undiagnosed fault in the electrical switching circuitry, which in turn meant I would have to remove the back of the machine after all.
Naturally, this epiphany brought with it strong feelings and I was so o'ertaken with emotion that I spoke in tongues for about ten minutes straight while walking around hitting things with other things.
I pulled the eviscerated machine from it's niche and began the tedious business of removing the back of the damn thing. Once again the removal of the last screw caused the liberated panel to undergo some impromptu topological explorations of the planar form and I was struck smartly in the head as it reshaped itself unpredictably with a musical bong! Once again I sustained ten or twenty small slashes on my hands from the edges of the bloody thing. Many and wondrous were the words of power uttered on that day I can tell you.
Once the panel was off I hobbled to my feet, squeezed out of the niche behind the dryer and pushed the dryer back into its hole so I could retrieve the multimeter I had forgotten to bring with me in my rage. Then I was able to pull the dryer back out of its little space and squat back down to begin the business of electrical circuit diagnostics to the tune of my knee joints popping.
Which was when I saw the dangling wire.
Investigation showed that one of the wires connecting the thermostat I had replaced three weeks before to the circuit had disconnected itself because when it was removed a tiny locking tab in the "spade" connector had folded back and collapsed. The result was a connection that would slowly shake loose when the machine was running. Since this thermostat was the one that determined whether or not the burner would turn on initially, it was obviously the problem.
It was the work of a few minutes with my needle-nose pliers to re-form the connector so it would work properly, and to reconnect the thermostat. It was the work of a bit more than a few minutes to get myself out of the tiny confines of the dryer's space, move the dryer back into it, dummy-up the door sensor and run another test.
Bingo! The burner lit and once more my lungs drew in volumes of air tainted with the combustion byproducts of natural gas (mainly carbon monoxide and carbon dioxide, all good for many things but sustaining life isn't on the list of either). I had triumphed over adversity in a World Gone Mad and fixed the dryer and like that.
It took forever to get the bloody machine re-assembled. First there was the back panel, which was somehow now two screws short of the required ten but I no longer cared. I refuse to waste any more brain cells on the whereabouts of two machine screws that are verifiably not in any of the moving parts of the machine (I checked). Then there was the remounting of the drum, which involves mounting the drum on its rear bearing (a sort of huge felt hoop) while lying on the floor, then holding it in place with my feet and knees while simultaneously attempting to attach the belt tensioning idler to its mounting lug in the base of the machine while hooking it up to the belt itself using the hands, forehead and teeth. There was the reassembly of the casing which only takes two screws but uses up dozens of swear words. It was all very tedious, but the machine was finally working and still is as I type this some two and a half weeks later.
All in all a Great Triumph.
- One hesitates to say "final" since that word could only summon a swarm of anti-handyman demons to further confuse things↑
- The burner igniter works by passing electricity through a metal ribbon. If you use enough force, defined in this case by volts, 110 to be precise, the electric current passing through the material causes it to get hot. Red hot in this case. This is pretty much how any pilot-light-less gas system works↑
- i.e. Why were there three cables when two were all that was needed for switch duties? No man could say without access to the wiring diagram which was in the back of the machine which I did not want to have to remove again↑
- Root Mean Square, the only way that measuring a voltage that goes backwards and forwards sixty times a second makes any sense whatsoever. If you don't measure it like that you can end up deciding there are no volts on any live wires and experience shows that is a path to burned hands, flashing lights in the eyeballs, singed hair, unintended levitation and possible death↑