Ultrasonic cleaning to restore the YH-409 HVAC sensor and park brake vacuum switch

[Hillbillycat]

Hi Lutrova,

the six-pointed piece is a push in locking star washer. Just use your dental pic and gently pry up on the tabs. It will come loose. Thereīs not much tension.

Just finished a bench test of the reassembled sensor. I had hooked it up to my electric vacuum source and my spare HVAC plenum box, so I can watch the actual servo motor move.

Results:
Sensor doesnīt work at all. So itīs worse then before. It appears that the sensor bimetal has either lost its tension (need to disassemble again and check the bimetal alone with heat and ice spray) OR the stem I had cleaned the tiny grooves actually needs to seal tight - NOT bypass some air.
Like you said - they all tend towards full cold. But isnīt rubber prone to harden over time? I donīt think the orange seal had softened but the bimetal lost tension. Isnīt this what happens to all these choke springs - they loose tension over time, donīt they?


Please do remove the star washer and look inside yours. Maybe I lost a part undernetah the orange rubber diaphragm (which I doubt). Maybe there should be a rubber seal that had disintegrated on mine and had been rinsed by the ultrasonic cleaning.

Reassembling the spring works smooth if you use two arterial clamps. One to hold the spring end in place, feed it through the hole and the other to grab the extruded end once itīs fed trough. Make sure to unscrew the setting screw at the bottom end. I took pics but need to upload.

[Hillbillycat]

Took it apart again and checked the bimetal with heat and cold. Flexes like it should:
Applying heat flattens the metal plate, applying cold arches the bimetal.

Now waiting on your answer on what you found underneath the diaphragm, Lutrova.

If it looks like mine and thereīs nothing else inside I think we can conclude that the bimetal had lost some of itīs arch (tension), hence biasing towards full cold. Iīll try and re-arch it then.

[kishy]

I think you're missing a piece...as I mentioned there is a tiny ball-bearing part that hides somewhere in there. I immediately lost mine when I took one apart, but I saw it exit before it was gone forever (presumably into the gap between the concrete pads of my garage floor). Either that or I imagined it.

[Lutrova]

Looks like Kishy was right. You do appear to have a piece missing. I don't believe it has anything to do with the sensor failing toward cold, but it might have an impact in the sensor functioning at all, or the ATC system holding vacuum on hills. This piece that comes out of the brass part of the diaphragm seems to plug the vacuum supply when the sensor calls for full heat, allowing servo vacuum to leak through the vent hole right beside the servo vacuum port. I think that if it's missing, you'll never fully plug the vac supply, which could lead to a system-wide leak. But it might also deprive the rest of the sensor valve of enough vacuum to operate the servo.

So this tiny moving piece seems to be an important part of the sensor's full operation. But I wonder if you just glued the hole where it should be shut if you might still be able to get some function out of the sensor.



Here's a rough section of what I think is going on with this sensor. I'm inclined to agree with you that the orange rubber part of the diaphragm is likely not the point of failure, but the bimetal losing tension. Maybe bending it to give a little more arch would do the trick, but giving it the right amount of tension to properly counteract the spring and operate within the correct range seems like an impossible task. Then again, if bending that piece does restore some function to the sensor, that would still be an improvement over the status quo. And we would at least understand why these things fail.

[Hillbillycat]

Hi Lutrova,

ASWEOME!

Now I know why mine doensīt work anymore after reassembly.
Strange that I did not see that tiny ball-plunger exit the brass shaft during disassembly. I worked slowly and carefully but it was not there. The only thing I can think of is that it fell through the end when the spring came off and me not knowing somethingīs inside to watch out for. But like I said I didnīt notice any parts falling down.

Bending the bimetal actually works to get the diaphragm move later. I tried that with mine yesterday (not knowing the ball plunger is actually missing). Thanks to the adjuster wheel on the butt end you can compensate if there was too much bending.

Sealing up the brass stem doesnīt work. You canīt control anything anymore. You get full vacuum that way. Tried that too by plugging the brass stem.



So I think we can conclude the following for restoring to proper operation.

(1. an ultrasonic cleaning to remove any debris inside.) - not mandatory but helpful. You could clean everything when fully disassembled.
2. careful disassembly (ball-plunger!) and re-arching of the bimetal plate.
3. re-setting the adjuster wheel on the back before install in car. This can be done while the sensor is hanging loose from the glove compartment and just connected to the corrugated input hose and vacuum lines. If it checks out OK in operation then the sensor can be reinstalled.

Iīll try to get one or two further sensors from parts cars for further testing and verifying our claim.


P.S. of, course: very impressive sketch!

[DerekTheGreat]

Lutrova, you sketched that diagram of the ATC sensor? Niiice.

[Hillbillycat]

You will not believe what I just found on my workbench under all the tools.........

the missing tiny ball plunger!!!

Oh my god, canīt beleive the luck I had. Will see if I get that sensor working again now.
At least I can dis-and reassemble the servo blindfolded now under a minute, hahah )))

[Lutrova]

I'm not sure how that ball plunger falls out. Mine looks fragile, but seems to be flared on both ends to keep it in place. Hopefully you can get yours back in. I made a bit of an error, I think, on that portion of my diagram. The ball plunger should be fully extended from the diaphragm by the vacuum supply, whose passageway tapers to a point to allow the ball to fully seat and cut off servo vacuum. So yes, as you say, without that piece you're always going to have full vacuum and full cold.

Good catch on the spring adjuster wheel. I had wondered about that gear on the back long ago, but then forgot all about it. What a clever piece of engineering.

I feel pretty good at this point about the odds of getting one of these things back in action. It'd be helpful to find some sort of benchmark for how arched the bimetal should be before reassembly, e.g. 3mm height at 70*F. The true adjustment probably goes down to tenths of a millimeter, but if some number can at least get people in the ballpark the spring adjustment wheel seems to provide a lot of wiggle room.

[Hillbillycat]

And itīs working!
I re-arched the bimetal again before reassembly.

Bench test on my spare HVAC plenum moves the servo fully to cold when the servo gets heated with a hot air gun and full hot when sprayed with ice spray. By moving the temperatur selector lever you dial in between.

In car test has to follow, but since temps are very hot here Iīll get full cold anyway (AC not working to provide a real cold airstream to cool off the sensor.)

The tiny plunger falls out when you unclip he bimetal from the tabs. This is the point that you need to remove the spring to get the bimetal off of the diaphragm stem. The spring prevents the tiny ball plunger from falling out.
My advice during reassembly is to first install the bimetal and then insert the plunger, ball end pointing down. Then immediately secure with the spring and youīre safe not to loose it during the reassembly process.

The tiny plunger has two different ends: One ball end. This goes into the valve and gets sucked into the hole in the black plastic valve stem, sealing it off. The other end is flat and minimal of larger diameter. This prevents the plunger from falling through the diaphragm stem and acts like a stop, limiting the travel.

Establishing a benchmark for the arch would be great for others to follow.
I did forget to measure the arch I did but it was more around 5mm I think.

[Lutrova]

That's great news.

Did you find the vacuum supply line holds vacuum at full hot, or does the plunger still let some through?

[gadget73]

I'm not sure what function that little plunger bit would serve, other than to stop vacuum from leaking through the open end of that brass piece. Seems to me you could seal that up with RTV or whatever and have it work since the vacuum would need to pass around it, not through it.

Otherwise this thing looks to work like a needle valve in a carb, just it uses a spring to pull the needle out of the hole instead of a float. Pull the thing out of the hole, vacuum passes to the servo, and it pulls to the cold side. Release the tension, the needle seats, vacuum vents, and the servo moves to hot.

Also based on the diagram (Excellent work BTW, that really helps a ton to see how this BS thing works) it would have to constantly leak some amount of vacuum when set to "cold" since the vent never seals, otherwise it would never be able to shift back to hot / no vacuum at the servo.

[Lutrova]

I think the plunger is the needle valve. It takes a very small amount of play in the plunger to lose all vacuum in the supply line. Without the plunger, the bottom of the brass diaphragm doesn't sit against the plastic valve in a way that will stop much of anything. The only thing restricting vacuum seems to be the ball at the end of the plunger, which sits against a narrower part of the plastic valve.

I've been doing a bit of testing of this, applying vacuum to the supply port without the spring attached to see how well the plunger seals up at full hot. The results have been a little inconsistent. When pressed from the hole at the top of the brass diaphragm, my plunger will hold vacuum indefinitely. Otherwise, the best unassisted case is a loss of 1 inHg every 3 seconds, which is probably just fine. But if you pull the diaphragm out a bit to simulate vacuum to the servo, the plunger will oftentimes not reseat itself quite right. Then it'll either leak rapidly, maybe 5 inHg/sec, or not seal at all. I had hoped that the application of vacuum would be enough force to pull the plunger into position, but this may not be the case.

I have three thoughts on getting the plunger to seat more consistently. 1) There could be some debris in the brass part of the diaphragm that's causing the plunger to stick. I haven't done ultrasonic cleaning, so if this is the case then maybe Hillbillycat will have better performance here. 2) There's a natural amount of friction between the diaphragm and plunger. Maybe applying a little silicone lubricant would get the plunger to go where it needs to. 3) This could just be the way the design works, and plunger position has always been erratic. I wonder whether gluing the plunger in its most extended position would do the trick. Either a bit of superglue on the outside, or packing the inside with RTV could keep it from getting pushed up and losing its seal. The RTV might still allow some flexibility too, which could be helpful if the spring pulls the diaphragm a little off center. It would also keep the plunger from ever being lost again.

[gadget73]

Ah, so its not the tapered face of the brass thing that does the sealing? I guess that makes sense that its a smaller doo-dad. We aren't talking a lot of volume here.

[kishy]

I've gotta say, bravo for digging in on this topic...I've meant to for a long time but I am not good at staying focused on tasks and my to-do list is frightening in all areas of my life

My concern about the line of thinking that [the sensor primarily fails due to a loss of tension of the spring and could possibly be corrected by bending the spring or turning the adjustment screw to add more tension] is that the failure mode, at least as I've seen it in my cars, is not an outright refusal to stay in the hot position. It drifts gradually over time, and if I jiggle the temperature lever around (generally setting it somewhat hotter in the process), it may resume functioning reasonably correctly but then drift cold again gradually.

It's entirely possible there are perhaps 4 or 5 different failure modes and we all sort of assume that we all experience the same thing, but in my case, I am not certain that more spring tension actually fixes the issue. Like, sure, we can make it be hotter, but it already gets hot, it just drifts over time. At the same time, if a 1/4 turn adjustment on the screw every 30 years is all it takes to keep these things working happily, that's a wonderful possibility.

[Lutrova]

....My concern about the line of thinking that [the sensor primarily fails due to a loss of tension of the spring and could possibly be corrected by bending the spring or turning the adjustment screw to add more tension] is that the failure mode, at least as I've seen it in my cars, is not an outright refusal to stay in the hot position. It drifts gradually over time, and if I jiggle the temperature lever around (generally setting it somewhat hotter in the process), it may resume functioning reasonably correctly but then drift cold again gradually....

I agree that there are likely several different methods of failure for these things, and loss of tension in the bimetal is only really compelling for a constant cold condition - which is the way my sensor had failed. Assuming I'm correcting that with a little bending, I think I might now be encountering the failure mode you've described. And my working theory for that is that the plunger is wandering around and not seating properly. Just handling the diaphragm I've noticed a tendency for the plunger to retract into the brass housing, and it seems to keep doing that when installed.

So if the problem is the plunger losing position, the solution ought to be to fix that position. As far as I understand, the plunger has no need to move independently of the diaphragm, which should only be moving based on what the bimetal sensor and the spring are doing. The question then becomes how best to keep the plunger's position.