Trying to Get My Armada’s A/C System through the Summer (Part I)

I’m going to delve into some of the issues of the air-conditioning system in my 2008 Nissan Armada—the vehicle I love to hate, or at least to intensely dislike due to its apparently never-ending needs. Just to be clear about this, I’m not a pro, so don’t take any of this as vetted advice. It’s just me describing what I do.

I’ve done a lot of A/C work over the decades, with a concentration on retrofitting or resurrecting the A/C in my older cars. I wrote a book about it called Just Needs a Recharge: The Hack Mechanic Guide to Vintage Air Conditioning. In it, I define “vintage air conditioning” as a system that may have originally had R12 refrigerant in it instead of the newer R134a; that likely had an old-school upright compressor instead of the newer smaller rotary-style ones; that may need improvements to blow cold if it’s converted from R12 to R134a; and—most importantly—is not part of a climate-control system where there are a bazzilion sensors and the heater core and the A/C evaporator core are co-located inside the same box and a blend door is used to mix cold and hot air to achieve a certain cabin temperature.

This last bit is crucial. On a functional vintage A/C system, when you twist the knob to turn it on, that sends power directly to the compressor. As long as there is more or less the right amount of refrigerant in the system and there isn’t a blown fuse or any broken wires, there are typically only two things that can prevent the compressor from switching on and pumping refrigerant to create cold air. The first is the pressure switch screwed into the receiver-dryer that cuts off voltage to the compressor if the system pressure is either too high or too low. This is intended to save the compressor if the system either leaks catastrophically or has a blockage, but in practice, on vintage cars, I often run without the pressure cut-off switch because it’s a potential source of leakage. And besides, if a hose blows or a connector breaks, you usually know it immediately from the refrigerant cloud and smell. The basic electrical layout of a vintage A/C system is shown below in this illustration from my book.

Typically, there’s also an anti-icing switch in the thermostat that cycles the compressor off if it senses that the evaporator coils are freezing, but my experience is that when those fail, they usually fail in the “on” position. (In a great bit of automotive trivia, A/C-equipped AMC Ramblers between about 1962 and 1966 had a “Desert Only” setting on the dashboard that bypassed the freeze-up switch if you were driving in ultra-low-humidity environments.)

In contrast, on a modern climate-controlled vehicle, particularly a large one like the Armada, with its three separate zones (driver, passenger, and rear), the block-functional diagram is seizure-inducing. In the center of the one for the Armada, there’s the oddly named “a/c auto amp”—the brain of the climate-control system.

Simply to engage the Armada’s A/C compressor, there is a labyrinth of gating functions that go far beyond turning it on, unless there’s a pressure issue. Wrap your head around this description from the Armada factory manual of the decision tree for sending voltage to the compressor clutch:

“When the A/C switch is pressed, or the mode switch is turned to the defroster position, the front air control outputs a compressor ON signal to Body Control Module (BCM). The BCM then sends a compressor ON signal to the Engine Control Module (ECM), via CAN communication line. The ECM judges whether compressor can be turned ON, based on each sensor status (refrigerant pressure sensor signal, throttle angle sensor, in-vehicle temperature sensor, ambient temperature sensor, optical sensor, intake sensor, etc). If it judges compressor can be turned ON, it sends compressor ON signal to IPDM E/R, via CAN communication line. Upon receipt of compressor ON signal from ECM, IPDM E/R turns air conditioner relay ON to operate compressor.” Sheesh.

When I bought the Armada, it was winter, so I couldn’t tell whether the A/C system was working or not. I could tell that there was a rhythmic clicking sound coming from under the dash. I assumed correctly that this was from one of the actuators trying to move one of the blend doors in the climate control box, an issue I’ll circle back to later. In late spring, the weather got warm enough that I began sussing out the A/C system. To my delight, I found that it had refrigerant in it. I shot one can of R134a into it and it began to cool. Another can and cool transitioned to cold. 

And, of course, it didn’t last.

If you haven’t figured it out, the Just Needs a Recharge title of my book is a joke (like Ran When Parked, the title of one of my other books). An A/C system NEVER “just needs a recharge.” A/C refrigerant isn’t used up like engine oil. If it needs a recharge, it’s because the refrigerant leaked out. The leak needs to be found and fixed. Part of my whole vintage-car A/C rejuvenation mantra is that if you leak-test the evaporator assembly by pressurizing it before you install it under the dashboard (the hardest part), then procure all-new components, make new hoses, and leak-test the entire system before you charge it up, it simply comes up working and stays that way. And if it does leak, the leaks are typically easy to find due to the simple nature of a vintage system and the amount of space in the engine compartment. But on a modern system in a large vehicle with rear air, due to the number of components, with how tightly things are packed in, and with the amount of plumbing required, finding a leak can be far more difficult. 

But with it sort of working, I can just keep topping it up to keep it working, right?

Well . . .

Despite certain reports to the contrary, it is not in any way illegal to top off the refrigerant in a car’s A/C system. If it was, anyone buying a can of R134a, particularly the expensive ones with the integrated hose and gauge, could be followed home, surveilled, and arrested. (What is illegal is intentionally venting A/C refrigerant into the atmosphere—e.g., pressing down the little valves in the middle of the service ports.) If you pay a dealership or a qualified A/C shop to test your system, what they’ll do is pump all the refrigerant into a pressurized container and compare its before-and-after weights. It’s from there that they deduce whether the system was low on refrigerant, infer the presence of a leak, try to find and repair it, and recharge the system with the proper amount of refrigerant. Most DIY-ers, including me, don’t have this kind of suck-it-out-and-weigh-it equipment, so we rely on either doing it from scratch with an integral number of whole cans of R134a that add up to the total system capacity, or by topping off a system while looking at a set of manifold gauges to make sure the low- and high-pressure sides are within some reasonable range (about 20 to 45 low and about 150 to 275, respectively), all while measuring the vent temperature with a thermometer or simply feeling the crack of cold air with our hand.

To me, there’s a clear and consummately practical line between topping off a slightly leaking system and hopelessly emptying $12 cans of R134a into it as if they’re Axe deodorant before an out-of-your-league prom date. The crucial issue is whether the A/C system has any pressure left in it at all. To determine this properly, don’t just take a matchstick to one of the service valves and see if it spits (illegal, remember?). Do it correctly with a manifold gauge set. Make sure the knobs on both gauges are rotated closed, then hook up the low- and high-side hoses to the vehicle’s service ports. The gauges should read the resting (ambient) pressure of the refrigerant. While there’s no strict metric relating resting pressure to whether the system is under-charged or over-charged, I’ve always used the rough rule of thumb that the resting pressure reading in psi should roughly match the air temperature in °F (e.g., 90 degrees out, 90 psi). The exact number, though, is unimportant. The issue is whether the resting pressure is zero, because if it is, there’s a major leak that does need to be fixed first, and unloading every can of R134a from the shelves of AutoZone into it won’t change that. But if it’s got a good amount of resting pressure, whatever leak it may have is small, and you may get away with topping it up. For a while.

That appeared to be the case with the Armada, because, unfortunately, a few weeks later, the A/C seemed to go intermittent. My assumption was that enough refrigerant had leaked out quickly enough that it rocked back over the line into non-functionality, so I tried to find the leak. I used Big Blu soap bubble solution as well as sniffing with a refrigerant detector. Neither found anything. I usually don’t regard dye as a go-to solution, but I shot in a can of R134a with dye in it. This, along with a U/V light, revealed a leak at the joint between the condenser in the nose and the receiver-dryer. Unfortunately, on this vehicle, those two components are integrated, so to fix it, the condenser must be replaced. As in nearly every car, the condenser is in the nose in front of the radiator, so the radiator has to come out first. On one of my vintage cars, this is maybe 20 minutes of work, but on the Armada, it’s a bear, and I’m hoping I can put it off until September. 

So, I find myself trying to do what I joke about in the title of book. And, again, for a while, the additional can of R134a seemed to be all the system needed.

Unfortunately, last week, it went intermittent again. Specifically, I’d turn it on and it would blow cold, but then after driving for a bit, it would get warm in the cabin. The light on the A/C switch would still be on, but when I stopped and looked under the hood, I could see that the compressor wasn’t engaged (the pulley was spinning, but the nose wasn’t). The odd thing was that the next time I drove the Armada, the problem repeated—the A/C initially worked fine, then got warm. So it didn’t seem to me that the system was simply low on refrigerant—it seemed that something was shutting off the compressor. Nor did it seem that the compressor was “short-cycling,” which is normal if it’s low on refrigerant. The length of its on and off cycles seemed much longer. On my vintage A/C-equipped cars, I’d simply hot-wire the compressor to the battery as a test, but the Armada’s compressor connector is very difficult to reach and separate.

The Armada and other Nissan factory service manuals are available online. I found that the manual has an exceptionally thorough climate-control troubleshooting guide, but its level of complexity mirrors that of the illustrations above. It would take me forever to step through it all. If I had diagnostic software that pointed me to the source of a fault code, that would be much faster, but I haven’t bought or appropriated the Nissan CONSULT factory diagnostic software or purchased a good scan tool with access to Nissan-specific diagnostics. 

So I was thrilled when I learned that the vehicle itself has some diagnostic capabilities that are accessible on its console display via a Byzantine process of starting the car and then, within 10 seconds, holding down the “settings” button while twisting the radio control back and forth. I continued to be joyous when I located the climate-control diagnostics and heard and felt the actuators and fans being put through their paces. But then it mysteriously reported nothing other than the outside ambient temperature readings going wonky. I read up on the Armada forum (clubarmada.com) and learned that the error codes are actually flashed in the temperature field. I wrote them all down, came inside, pored over the factory manual and the forum, and learned that my car, for reasons unknown, had posted three codes whose meanings aren’t listed anywhere. So much for the in-car diagnostics.

I read more on the forum and saw a hail of posts describing symptoms remarkably like mine that were fixed by a $60 easy-to-replace resistance-based temperature sensor that sits up under the dash in the inlet of the evaporator assembly. The posts said that the sensor usually works, or else is shorted (low resistance), or is open (infinite resistance). I checked mine in both hot and cold water with a multimeter, and it basically seemed to work, with resistance decreasing with increasing temperature as it should, but it did seem to be off from the table of values in the factory service manual. And it made sense that, when comparing the evaporator temp to the cabin temp, erroneous data would cause the compressor to be shut off, right? I’m usually willing to try one of these “maybe it’s this” things, so I ponied up the 60 bucks, ordered it online, waited for arrival, and installed it. 

And it made no difference. Whenever I drove the Armada, it continued this frustrating behavior where the A/C would initially work, then mysteriously shut off, then come back alive the next time I used the truck. Oh, did I forget to mention that this Armada is black with a black interior? Combined with a hot summer in Boston with stretches of 95-degree heat, it made me think that I couldn’t rely on the Nissan if I needed to actually use it for anything. Not that I really was, mind you (and that’s a larger story of the folly of my having bought it in the first place), but the entire scenario just added to the way that the thing glowered at me in the driveway. My wife certainly wouldn’t ride in it with me knowing that it could turn into a sweat lodge at any moment.

So, how do you proceed? Is there a middle ground between a nosebleed-level repair bill at a dealership and hot-wiring the damned compressor to the battery?

I’ll talk about that next week.

***

Rob’s latest book, The Best Of The Hack Mechanic™: 35 years of hacks, kluges, and assorted automotive mayhem, is available on Amazon here. His other seven books are available here on Amazon, or you can order personally inscribed copies from Rob’s website, www.robsiegel.com.