Gas valve or circuit board, short-circuit protection?

JaneH

Wow this forum does have professional discussions! I hope you guys can help me with this puzzle...( I read https://forum.heatinghelp.com/discussion/160527/gas-heating-gas-valve-or-circuit-board and it's kinda similar to my case but not exactly).

Now I look back I realized it's been 1-2 weeks sub-normal, till one morning I woke up to 5 degrees lower than my setting. Cleaned the flame sensor and it seemed improved but not a lot. Currently my observation is this:
1. The gas valve fails to turn on about half of the times;
2. The gas valve stays on when it should most of the time -- the flame sensor LED is always nice and bright, no low sensor signal ;
3. The best part: The probes of a multimeter on the gas valve terminals help tremendously to "jump" it on when it should. But only when the meter is on( I get AC 22.6-23.2v). If the meter is off, it doesn't help.
So the meter acts like a resistor? The gas valve inside has deteriorated and there is some kind of short so the board tripped it from some over-current protection(Is there such a thing for gas valve at all?) but with the meter as a resistor the protection didn't kick in? Or on the control board there is some hair-thin short that triggers the trip? I'm trying to make sense of this to decide if I should replace this 2-year-old board or have someone professional to replace the gas valve -- all wearing masks of course .

4. I tried @Abracadabra 's putting-fingers-on magic. I feel the relay click but not the valve tap(for the failed cases). However, for such case the voltage when the relay clicks goes down to 0 while the normal off condition is 0.8v -- that's why I was thinking of some short-circuit. Or is it a bad relay that instead of clicking to 24v it sometimes decides to click to 0v? LOL
5. If I have not confused you enough, here are what I hear for a successful cycle:
click and the small fan is on; seconds later click and igniter turns orange slowly; 15 seconds click( a big relay), tap(gas valve), and fire simultaneously, meter AC 23v steady, click of igniter off, blower's scary clicks and hot air blowing.
And a failed cycle:
click and the small fan is on; seconds later click and igniter turns orange slowly; 15 seconds click( the big relay), voltage falls from 0.8v to 0, 2 clicks(one turns off igniter one turns off gas valve I think), and try again.

I'm going to unplug the gas valve to measure the board side 24v and the valve side ohm tomorrow as suggested by @ratio in the post mentioned above-- don't want to unplug anything now it's a snowy night and kids in bed.

Thanks for any tips!
-Jane

JaneH

Anything I did? No one would comment?

JUGHNE

Post a picture of the gas valve/burner compartment.
Show the gas valve label. Indicate where you put the meter probes that make it work.
Stick toothpicks in the place.

JaneH

Zooming in

The two yellow wire here are connecting to the gas valve terminals so I’m measuring here:

Thanks for looking!

JUGHNE

There is a fuse on the board, upper left hand corner.
What looks like the letter "E" is actually a 3 upside down.
That protects all the 24 volt wiring.

It might be a loose crimp on the inside of the Molex plug where you put the meter leads. You could unplug it and check both M & F terminals.

Is the physical push of the probe making the valve work or do you think the internal electrical load of the meter does it?

It is best to test the voltage right at the valve. If you pull the crimps up slightly to expose the male terminals going right into the valve and put the leads there.

You could push something in the Molex connector besides the meter leads and see if that corrects things.

pecmsg

That flex line is not allowed in the cabinet.

Ill A S S U M E the fuse is good because of the green LED

JaneH

Sorry pecmsg what flex line?

Yes the fuse is good. But by mentioning that JUGHNE assured me that there are no other transient over-current protection therefore my imagination of that dropping my 24v on the gas valve is invalid.

I measured the ohm between gas valve terminals. It's more than 20M so that doesn't look good?
However it works perfectly last night and today...

I measured the voltage for the gas valve when the gas valve is unplugged, almost 24v. However while playing there I feel the wires/connectors might be fishy. I'll follow JUGHNE's advice to reinforce the weak points and report back.

Thank you so much guys! I hope it's just a shameful connection issue!

pecmsg

The flexible gas line connected to the steel pipe!

Only Solid pipe is required inside the cabinet. No Unions

JUGHNE

Yes to have flex gas pipe inside the furnace cabinet is not to code, that is connected to your gas valve.
It would not cause the problem you are having.

Gas valves are not easily measured with ohm meter, I never have luck trying that.

JaneH

Ahhh what do I do? Sue the previous owner or get sued by the county ? Maybe Virginia is different? Flex pipe poses safety concern? I now leave the cabinet door open day and night since some superstitious observations told me that it fails less frequently when I leave the door open -- maybe just easier access when it fails so not very objective. I guess that's considered breaking some code too.

pecmsg

That flex line can be changed at a later date but has to be done.

Agreed that's a separate issue.

JUGHNE

If you have the gas valve replaced that would be a chance to have black threaded pipe installed inside the furnace. It should also have a "drip leg" on the outside of the furnace after the gas stop.
Flex and unions are more prone to gas leaks than solid lengths of pipe. If the furnace has some vibration it could also eventually wear thru the flex where it enters the cabinet.
A leak within the cabinet may not be noticed as the burner fan may pull the leaked gas out and also a build up of gas could be ignited when the furnace fires.

Flex used like that is the lazy "git er done" installers time saver.

JaneH

Reporting back -- I hope it is fixed. Let's see.

Waking up to another morning of 5 degrees lower than my setting, I was actually having a firm plan -- find the (sometimes)bad connection and fix it now!
Finally the measuring points me to the Molex plug F terminal(board side) of one of the two yellow wires(in my picture 1&2), where I put the black meter lead (in picture3). Re-soldering that pin fixed the shaky ohm I measure between the two yellow wire F terminals, from a sometimes-78-sometimes-infinity-sometimes-jumping-around to a steady 78 ohm . I hope that helps.

Now it's running as normal. Fingers crossed, no more drama please.
Thank you guys a lot for the technical and moral support! It means a ton to me. Especially to @JUGHNE: you're right about everything! All the little dos and don'ts and names of little thingy gave me so much to analyze and they directed me to the right path.

mattmia2

Was there a broken solder joints on one of the molex pins through the circuit board?

JaneH

Not really broken, but sketchy, might be cracking. I'm not positive. But yes, it's about a "solder joint on one of the molex pins through the circuit board". Thanks for the clarification.

Here is what I observed:
I was measuring the ohm between the two F terminals (where those two yellow wires go in. They supply 24v to the gas valve). You know the tiny cup of the F terminals? One of them would give me infinity if I touch only the rim, but 78 ohm if I push the lead into it. The other one would show the same(78) for the inside and the rim. For a second I was like: this metal cup is not a conductor LOL? Then I realized it's the connection to the board -- when I touch the rim lightly it's not connecting to the board; when I push down deeply it is connecting.