York Furnace Won't Ignite Unless Multimeter Attached

dbrubeck

I have a York furnace. During the startup cycle, the electric ignitor glows red, but no gas comes out. I can measure 24VAC from the circuit board to the gas pump at the appropriate time. The resistance across the gas pump is infinite. I have another identical furnace where the resistance across the gas pump reads 8 Mohms. So it seems that the gas pump needs to be replaced. (please confirm)

The interesting part is when I leave the wires connected to the gas pump, and stick my multimeter probes (set to VAC) into the connector next to the wires that go to the gas pump as they leave the circuit board. So essentially the multimeter is connected in parallel to the gas pump. I measure the expected 24VAC, and the furnace ignites.

Can anyone explain why the presence of the multimeter makes the gas pump work?

Thanks!

Jamie Hall

Bad connection, and the multimeter probes are forcing things together. Check all the plugs and connectors before you go throwing parts at it.

EBEBRATT-Ed

I agree with @Jamie Hall. Loose connection of some sort is the likely cause

dbrubeck

thanks I'll poke around some more before replacing the gas valve.

dbrubeck

So I tested the gas valve by running 24VAC from the C/R terminals and the valve works well.
I unplugged the leads at the gas valve and also measured 24VAC at ignition time. If something is loose I would expect it to not show the voltage.
So the gas valve is good and the voltage to the valve is good, but when I put them together there is no gas flow.
So I stuck an extra wire into the connector and everything is working. I still can't understand why this fixes it. I'll post a picture for the curious. Extra wire is there third from the top.

HVACNUT

And that's where the loose connection is, in the molex plug. Turn off the power, remove the molex plug from the board and make sure the pins on the other side of the plug are properly seated.

ratio

I think the first, second, and fourth comments nailed it.

Sometimes, the connector will get reamed out to the point that it doesn't make physical contact, but another failure that I've came across is a layer of oxide on the mating surface that, while appearing to have mechanical contact, impedes current flow. Symptoms range from no operation to intermittent operation. Fastest fix is a few mating cycles, that generally knocks off enough of the oxidation to allow operation.

Jamie Hall

I would point out -- in connection with the comments on oxide etc. -- that it is entirely possible for a connection to show good voltage with a multimeter (which has a very high resistance) but, when a load is placed across the connection to not be able to pass any current. Don't be fooled by this; a lot of folks have been.

JUGHNE

Sometimes there is a loose solder joint on the back side of the board. You can pull the board off and check.
Your added pressure of jam in devices could be helping a cracked solder joint stay in place.

dbrubeck

great inputs thanks. I did disconnect the molex and inspect the pins and put it back on, but didn't try removing the board to inspect the solder joints. Will try this next. Thanks!

EBEBRATT-Ed

as far as bad connections go, 575 volt or 480 volt it takes some resistance to stop current flow......they will come and get you!!

24v and 12 volt (like your car battery) Low voltage can deliver some power but can't jump through dirty, loose or bad connections like higher voltage will. Doesn't take much to stop current flow

Jamie Hall

I had an interesting one like that the other day -- one side of the usual 120/240 line serving an outbuilding seemed quite dead but wasn't -- lit the neon in my tester gadget. Turn on a light on that side, though -- nothing. No load and test it with a multimeter -- voila, 110 volts (the other side said 122 -- typical around here). Clearly not a bad neutral, as the good side said 122 to neutral no matter what the load, but put a load on the bad side (like a 100 watt bulb) and it dropped right down to about 1 volt to neutral.

Traced it to an aluminium/copper connection where the line from the main switchboard (copper) connected to the overhead line (aluminium) to the outbuilding. Not loose enough to pull apart, but loose enough so that oxide had built up in the connector (and it was a Cu/Al rated connector) just enough... been up there the better part of 50 years.

Gremlins.

unclejohn

@Jamie Hall. I call that ghost voltage. It's there but it isn't.

EBEBRATT-Ed

@Jamie Hall. That's what the oxide inhibiter is for. Usually works pretty good but after 50 years??

Lucky you didn't lose the neutral, that puts your 120 volt loads on the two legs in series (I am assuming single phase power) with 240 volts across them you an burn some things up depending on load resistance.

Probably should re make all three connections.

Had a fluorescent light that if you put a meter between the metal on the fixture and the neutral you would read 277 volt (it was a 277 lighting circuit) you could read voltage but nothing there.....you could touch it. Took the fixture apart, it had never been grounded

Jamie Hall

EBEBRATT-Ed said:

@Jamie Hall. That's what the oxide inhibiter is for. Usually works pretty good but after 50 years??

...
Probably should re make all three connections.

...

I did...

Solid_Fuel_Man

Aluminum conductors do have some serious consequences if aluminum oxide forms, as it's a great resistor! Resistors make heat, which forms more oxide, which makes more heat.... it is claimed that the "new" aluminum alloy wire since the 1970s is much more stable than the pure aluminum used originally.

I've made thousands of connections with aluminum, but I wonder what percentage will fail and in how many years. And 10s of thousands of copper connections, wonder what percentage will fail and when. Who knows?

@Jamie Hall that old tri-plex (assume that's what went from building to building) is good stuff, it's funny how 50+ years ago the plastic insulatuon holds up so well.