2003 3.0L Non-nav

Situation: Climate control works 4.0 for "awhile" whether ambient is 80-90 F or 30-40F. Also doesn't matter if the set-point is significantly above or below ambient. The vehicle delivers cold air in fine fashion on a hot day with a low set-point, and modulates well with a set point just above a cool ambient. Good mix temp and you can very slightly feel when the water valve has been opened and closed with slight variations in discharge temperature.

However, after 15-20 mins of operation, it goes full-hot. And stays that way, until shutdown and some small waiting period. Immediate restart does not seem to clear the problem, but a stop at an interstate rest area, walk in to the water closet to have a wee, and walk back to the car and it may be good for another 15-20 mins.

Setting the control to "LO" or full-cold, will get cool-to-cold air out of the inboard/center vents and hot air out the outboard registers. So you can close the outboards and make the drive "tolerable."

Troubleshooting to date: I have found several posts quoting the same behavior, some were resolved with renewal of the evap temp sensor, others with a repaired trace on the climate control board.

I've replaced the evap temp sensor, and the new one was indicating about 10 deg warmer than the one it replaced, with both at the same temp. The delta remained more or less constant between 50F and 90F. No joy, so pulled the driver and passenger heater sensors (Same p/n as evap temp sensor) and checked them against the removed evap temp sensor. They were both in general agreement with each other and about 10F warmer than the old evap temp sensor.

Hooked up WDS and found permanent failures of both solar sensors. However, I'm relatively certain these have been out since we bought the car in 08. The headlights in "auto" setting are ALWAYS on, regardless of ambient light, whereas on the 05, they go off in daylight - so I attribute that to the solar sensor input - perhaps errantly. Watched various parameters on WDS whilst running - both while CC was "working" and while it was failing to full-hot. Nothing jumped out as a "trigger." Evap air sensor reporting 2-4 deg C, sometimes up to 6. Driver and passenger heater sensors reporting temps in accordance with each other and their set-points. One anamoly, both driver heater coolant control and passenger heater coolant control reported "E035" instead of values. I took the leading "E" to mean "error." I removed the climate control board and inspected it under magnification, noting no traces that appeared damaged.

Hooked WDS to the 05 and those coolant control parameters were not showing up on the pick-list to watch. Otherwise, very similar behavior to that observed on the 03 except it never took off into full-hot faiure mode. Unfortunately, the 05 was only available for a week around Christmas and is now out-of-state with my daughter, so no further comparison-troubleshooting.

Obtained a replacement (Used) climate control for an 03 via eBay. Once fitted, the car exhibited the same behavior, including the "E035" on heater coolant values. One improvement, the display on the replacement doesn't go blank or blink (at least it hasn't yet) like the one it replaced. On the down-side, the fan speed knob is a bit nicked-up.

Lastly, I've unplugged the water-valve and conducted a test-run. This confirmed that with the control set to full-cold, I get cool/cold air from the center vents and warm air from the outboards with the water-valve open. This phenomenon had been vexing me. I had a circuit board failure on the 05 a few years ago, driver's side wv full open, and could never get any sort of cool air from the driver-side outboard nor center vent despite setting, til I mended it. While disconnected, I've confirmed 12V at the center pin of the WV plug, with grounding at the outboard pins at full-cold and open-circuit to ground at full-hot.

Summary: The water-valve is opening, on both sides, at some point after 15-20 mins of operation. The climate control is commanding it closed, but to no avail.

So does this indicate a failure in the valve? The solenoids are unable to close above some temp? Ambient doesn't seem to play a role in how long the control operates normally, which leads me to believe it occurs when the coolant boosts valve temp to a certain point. However, I don't believe a wee-stop at a rest area would allow sufficient cooling to resume normal operation?!? Is there a climate-control ground or power connection that could be dirty, preventing delivery of sufficient electrical potential to actuate the solenoids?

I don't mind replacing the WV, apart from the historical fact that about 1 of every 2 new ones leaks externally......but would hate to do that to find the problem persists....

 

It may not be this, but the CCM can fail with blown transistor(s) and they tend to look fine - i.e. visual inspection can show faults but lack of visual faults can just be that a fault isn't apparent.

 

True. But a second module with the exact same failure mode? (Although i haven't visually inspected the replacement) The eBay seller claimed it as good, only problems as viewable in the pictures. So it couldn't possibly be faulty......

Ideally, I would've swapped control modules between the 03 and 05 when the 05 was present. This would've cleared things up immensely. However.....finite amount of time, generally good weather over the Christmas break...and I had other outdoor tasks that really couldn't be delayed...so missed that opportunity.

This afternoon, I've teamed up with a chap some of you know, Rube Goldberg, and we've fashioned a "connector" of sorts for the water-valve. Now to look up in the electrical guide a suitable fuse to tap in the engine bay junction box and the plan is to run with the WV forced closed and see if it opens against direct pwr/ground connections.

 

This scheme seemed good, for temporary testing:



The idea was to run the red wires to chassis ground, and the middle wire to +B which should command both solenoids to close the water valve, both sides, and hold it closed.

Admittedly, much easier to fit it to the old WV on the bench than to the installed one:



THen connect ground:



and seek battery voltage - according to my meter, all of these locations are hot with key off:



May have assumed too much, and didn't pay sufficient attention to detail to check that terminal with key on, but generally, hot with key off stays hot in run. In any case, when I plugged that wire into the source, I heard a click down at the wv. So far, so good.

Started the car and set temp at 69F vs. 60F ambient as reported on the climate screen. I did not expect discharge air to get any higher than ambient, as the water valve was commanded closed on both sides, and the coolant was cooled to near ambient. I let it idle a bit while completing some other tasks, then noticed elevated temp in the discharge air, all across. Not hot, but warm, and the coolant just coming up to temp. This indicated either at least one side of the wv had opened, or one or more of my makeshift connections had failed.

I had it up on ramps, so I inspected from both above and below and the wires still appeared to be in-place. Whilst underneath, I successfully resisted the urge to reach in and touch it (the undertray is off while I troubleshoot) with visions of the FEAD grabbing my arm and tearing it off. So back on my feet and shut the engine off, no point in going for a test drive when the result will be failure and is easily seen at garage idle. Thought about pulling the power wire from the junction box but decided I wanted to feel the tops of both solenoids, expecting them to warm-to-hot. One (driver's side, I believe) was hot while the other was barely warm, but vibrating slightly.

Like a kid in a museum, among a bunch of "Don't touch the artifacts" signs...I was drawn to fiddle with the wires......sizzle, flash, bunch of white smoke....searing pain! The pain-part was perhaps consciously self-induced as I grabbed the glowing wire and yanked it in an effort to ensure the entire engine bay didn't go up in smoke. Then, quick, roll out, stand up and pull the power wire....Still...roasted a fair portion of thumbkin and pointer. (Don't try this at home)

Decided I'd better put everything back right and make sure I hadn't done any damage, so I reconnected the proper wv connector, lit the mill, and took it to Bridgeport and back - about 11 mi round-trip. HVAC performed flawlessly until about 3 mins beyond the turnaround point. Then hot air. On the way home, I set driver-side to 61F and passenger side to 75F. Cabin continued unabated heating. Swapped the settings to 75 Driver and 61 passenger and it got reasonably comfortable. At this point, ambient was down to about 55, and I was at this setting for perhaps the last two miles.

I'll take it on the daily commute this week as many days as I can stand it and play with the settings and try to determine if there is a particular coolant temp (Via ELM327/OBDII) or time that it fails, and if split-settings moderate the failure. For now, I'm most suspicious that one-side of the valve is failing, and that likely the passenger side.

 

Great detective work, Zane. Keep us posted....

 

I employed it (the 03 S-T) Monday morning for my 50+ mile commute. Had low 30's (F) ambient at start-up and frost on the windscreen. So against my wishes, I set defrost mode, then went back to the split 75 Driver/61 pax setting when the wind screen was clear. I was amazed that the climate control worked 4.0 until I hit the traffic slowdown on 35W, about 40-45 mins deep into the trip! Then, customary full-heat. That afternoon, returning home, with ambient about 65 F, I got failure within 5 minutes!

After cogitating on it a bit, and more importantly, finding the DCCV on eBay for $62 vice the $92 I'd seen everywhere else, I decided that it had to be the valve failing, as the climate control should command both sides closed at "LO" and both sides clearly weren't staying closed. As an aside, I forgot to mention, when it fails to hot and I run the climate control set to "LO" with outboard vents closed to stay reasonably comfortable...the rear air discharge is also full-hot.

I took it to Bridgeport this afternoon to pick up my daughter from school and since it was only 54 ambient, elected to leave the climate control "OFF." Now, this should leave the DCCV wide-open, and should bring it up to temperature. By the time I'd parked in the school lot, still in idle, I reckoned it had been running long enough to fail to full-hot. i turned on the HVAC and got good, cold air from the vents as befitted the set point of 61F. So I clicked the HVAC back off, opened the sunroof for comfort, and waited for her to show up, continuing at idle. When she arrived, we set off and turned on the HVAC again. It then failed to full-hot within 90 sec or so. This further convinced me it was a failure of one or both solenoids - it was not the coolant heating it above some certain temperature, but more likely the solenoid could only stay engaged so long, before letting go.

So the new valve showed up this afternoon, and after watching an episode of "Louder with Crowder" followed by "Grand Tour's" latest episode, I finally mustered the gumption to go fit it this evening. Now I've got NO HEAT!

OK OK, I did wrap on both solenoid cans as sharply as I could, given the limited space with a brass mallet, and managed to get lukewarm air from the passenger side and rear vents. Driver's side still stone cold. It was late and I'd been assisted in the fitment by Corona, so all testing was conducted on the ramps in idle, but I did have full operating temperature.

Tomorrow's test plan will be to take it down the road in hopes that increased temperatures and coolant flow rate at elevated rpm will "break things loose" and set everything right. If not, unplug the DCCV and take it back out to confirm it is a stuck-closed valve and not the climate control commanding it closed.

For the sport of it, I fit the makeshift harness responsible for last week's 3rd deg burns to thumb and forefinger back to the old valve, just removed and connected the other ends of the leads to the battery. I got faint, clicking sounds, not the solid bangs I expected. Looking into the ports with a flashlight, I noticed one side was not fully closing. I dis-assembled the valve, and found both sides to be free and easy to manipulate. I suppose it is just a decayed coil or winding of some sort. Mechanically, at least, I din't see any problems....but of course, now, my concern is more for why the new valve won't open, in lieu of why the old valve won't close...

Next day edit: I gave it a spirited booting this morning with the climate control set to "OFF." This was apparently what was needed to blow both sides of the new DCCV open. When I activated the climate control, I had warm air on all sides. Ran the setting down to "Lo" and got cold air everywhere just before getting back home. This was enough running to normally result in failure to full-hot but I'm not entirely sure it is mended. I'll use it for the commute some next week before declaring full success.

 

Apparently my problem was one or two week solenoids in the DCCV that would work fine for a spell, then let go, allowing hot coolant to pass to the heater matrix. Supports my postulate that ALL water valves leak; the only variables being "when?" and "internal or external?"

I put 300+ miles on it commuting Wed-Fri and the climate control performed flawlessly, whether in afternoon highs of 82F or morning lows in the 30's.

This is the car that I changed the DCCV pre-emptively several years ago while renewing something else in the cooling system. That one leaked externally within a week. The just-removed one was its replacement. Not sure its age offhand, but think it has no more than 50k miles on it. It has been failing in this manner for nearly a year.

Edit: Just did my research and determined my initial rounds of musical DCCV's began in early October, 2012 and concluded in late December, 2012 at nearly 105,000 miles. So the chronology is this: Replaced the functional OEM DCCV in Oct, '12 while renewing the radiator. New 4-seasons valve leaked within a week, but I didn't have access to the car until Thanksgiving, when I returned the OEM valve to service and contacted Rock Auto which provided a 4-seasons replacement, which I fit in Dec '12 after the OEM started leaking. I'm at 162,500 now, and as mentioned, it's been failed for maybe a year and 10-12k miles, I'd reckon.

I changed the DCCV on the 05 in Sept, 2014, (@103k mi) also "pre-emptively" as part of a radiator renewal, or so I thought. Fitting a new Motorcraft YG378 revealed that the OEM DCCV had a stuck-closed drivers' side solenoid and a failed climate control module! Not sure of the mileage on that one right now, but my last record was 133k mi and my maintenance app predicts it should be at about 138k by now.

For the record, I fit the motorcraft to the 03 this time...

 

Congrats, Zane. Let's hope your issue remains resolved. These DCCVs can really be troublesome, even brand new right out of the box. I'm still fortunate enough to be running the factory DCCV in my 2005 S-Type 3.0 and I do not plan to remove it unless/until it begins showing signs of failure. I've been wiping its exterior with a white paper napkin every weekend for many years during my regular fluids & pressures checks in hopes that when it finally fails, it will leak externally and I will catch it in plenty of time. No guarantees, I know. But that's about all we can do to monitor these damn things on a regular basis....

I'm glad that neither my wife's 2006 XK8 nor my 1999 Ram have dual climate control. Our 2013 Outback does, but the associated valve does not seem to experience the all-too-common failures that our S-Types do. I have never seen this issue mentioned on the Subaru forum that I utilize....