How To Repair
  2002 Toyota Camry SE 2.4L 4-cyclinder

© 2013 Brian Mork


Home  •   Site Index  •  Wiki  • Blog


Introduction

This web page contains a narrative log and pictorial essay to maintain a 2002 Toyota Camry SE with 2.4L 4-cylinder engine with 162,000+ miles.  At the bottom, I also have some simple graphs aggregating lots of data on costs.  This page is meant to teach you, document technical data, and perhaps most of all motivate and encourage you as you try to figure out your non-working car. I also have web pages about my 230,000 mile 1994 Suburu Legacy with 2.2L and a 1989 Dodge B250 Ram Van with 5.2L engine

I
always appreciate your link back to this page so Google thinks what I say is important!

Before you start work on a car project that may take several weeks,
consider canceling insurance or doing "storage" insurance. My insurer used to provide a "storage" option that provides only comprehensive coverage and drops the cost of a newer vehicle from $52.50/mo down to $5.00/mo.  More recently, they've changed the definition of "storage" to mean "state mandated minimum insurance," and for a Dodge van discussed in the other web apge, that change dropped cost from from $27.17/mo to $6.70/mo. Or, if you can handle the risk, you could remove ALL coverage including comprensive and save a boatload of money.

Air Conditioner
New Tires & Tie Rod NOT connected!
Rear wheel bolt stud replacement (remove badly rusted on rotor)
Air Conditioner Relay Again

Fall 2012 - Air Conditioner

The air conditioning started operating intermitently.  A green light would flash on the panel button, and the compressor failed to operate, so of course no cool air came out.  Eventually, it totally quit working and all I got was a flashing green light after about 5 seconds of "on time".  Big tip for you up front:  you can bypass all the hassles I had by noting if the compressor comes on during those 5 seconds and then stops, or whether it does not engage the clutch at all.

I took this to the shop and they quoted a bill of $600 to replace a sensor on the compressor pump - $300 in parts and $300 in labor.  They said the sensor cannot be purchased separately from the compressor.   Uggh.. I wasn't ready for this, so paid him a $70 diagnostic fee and took the car back home.  We'll live without air conditioning for the winter.

March 2013

I had good luck with Omega Environmental Technologies (retail covered as AC Compressor Pro) when buying a compressor installation kit for my old Dodge van, so I ordered a compressor kit for the Toyota.  I later found out the switch on the compressor is not a pressure switch but an inductive speed sensor like an ignition pickup coil.  Plus I later found out it did not come with the compressor!  I figured at least I'd pay the repair shop to install it because I didn't want to get into recovering the freon with a home-made dry ice and alcohol cold bath and old propane bottles.

I took the package to the shop and he said he'd get back to me with a quote. I reminded him about the diagnostic work he had already done, and told him I already purchased the replacement compressor kit.  He said, "Oh, you already know what you want done?  Let me go get a quote for you."  He came back with a quote of $700.  What?!  Labor was $300 last fall, and now it's $700.  Including parts, that's slightly over $1000 to replace a broken sensor!!  Something is not right.

Well.. he explained that it's not a pressure sensor, but rather a speed sensor.  To replace the sensor, he has to replace the compressor.  To replace the compressor, he has to replace the expansion valve,  to replace the expansion valve, he has to replace the evaporator, and to replace the evaporator (which I hadn't purchased yet), he has to disassemble the instrument panel.  Uggh.. really.. this is gross.   The story was changing quickly.  I didn't know if he was intentionally trying to squeeze me, or whether he didn't care to estimate with any degree of fidelity better than what was required to get my work signature.

I took the $1200.00 repair problem home a second time.

April 2013

I called the local Toyota dealer and asked if I could buy the sensor and fix the problem myself.  Sure, they said.  It's a $117.73 part.  What?!  You ~can~ buy the sensor separately?  Why did the repair shop say otherwise?  I researched on the web with a proper search for "a/c speed sensor" instead of "a/c pressure sensor" and discovered that the entire freon pressure needed to be emptied to replace the sensor. 

I had already paid $300 for the compressor kit.  They will take it back, but charge a 25% re-stocking fee.  Does anybody believe the repair shop might reimburse me for taking me unecessarily down this path?

I decided to go to a different shop and get a quote to simply empty the system, replace the sensor, and pump up the system again.  I asked for a quote and they called back to say the sensor could not be purchased separately.  Hmmm... this story is sounding repetitive.  I let him know Toyota offered to sell me the sensor and he said that if I brought it in, he'd be happy to evacuate the system and replace the sensor and pressurize the system for $173.10 in labor costs.

At least this was reasonably priced, so all I had to do was make sure the sensor was bad.  I no longer trusted the first repair shop.  I poked around on the web and found an article by Dan Marinucci specifically addressing blinking green dashboard lights and a/c speed sensors (Motor Magazine copy or local copy).  This article also confirmed that replacing the sensor would require depressurization of the system. [later note - Dan Marinucci's article is not correct; the green light blinks for other reasons I found out; he has not replied to my email giving him examples.]


Click on any picture to see a close-up view.



O'scope on the two pins from the a/c compressor speed sensor.
  Positive and negative excursions were 350 mVolts into a 1 Kohm load resister. Removing the resistor allowed the sensor to produce about 400 mV.  You can watch a video of the A/C speed sensor when I manually apply +12 volts to the compressor clutch to make it run.


After the pressure was brought up to about 200 psi, I disengaged the clutch and the pressure settled back to about 165 psi.

I decided these readings didn't mean mutch since it was about 45 degrees F at the time.


This shows the high pressure side pressure meter hooked up and a yellow test wire in place, being used at this time to energize the clutch coil.


Leaning over the driver side of the  front wheel looking at the relay and fuses.

I  removed the relay and stuck the blue electrical connectors in place so I could grab the connection better with a pair of aligator clips.


The fore/aft terminals are the relay coil, with 12v on the forward terminal.  The inboard/outboard terminals are the relay output.


Once I confirmed the relay was not working, I decided to cut it open and investigate.

Anvil, chisel, hammer...

Notice the original one is 90987-02022, which has a coil that the computer can barely turn on.  It tries, but then quits with the blinking green light on dashboard control.

Toyota service bulletin EL011-05 (6/3/05) says to replace with a 90987-02028 relay, which activates easier.


Silver legs energize the coil to magnetically operate the contacts.  Put 12 volts across the 170 ohm coil (silver legs) and you should hear a "click". The copeper colored legs are suppose to close with low resistance.  This relay clicked fine, but  made no contact between the copper feet.


Fuse and relay map inside the lid of the fuse box. The relay we care about is labeled "MG CLT" for MaGnetic CLuTch.

.

.

.

Thursday (Scoped the sensor)

With confidence that I now understood what the sensor was, I decided to remove all the fancy car computers and controls, and manually spin the compressor (+12 v to the clutch coil with the engine running) and measure the inductive sensor with my oscilloscope.  This was fine to do since I was sure that the system had sufficient gas pressure and oil to not burn up the compressor.

I connected a scope directly to the speed sensor and recorded a speed sensor video.  Look at the photo in row1 column1 above first, and then watch the video.  Right at the beginning of the video, you can hear the compressor clutch engage, and see the engine load down, and then it stabilizes back at a higher idle RPM (pulses closer together).  When I release the clutch, the pulses go away.  Everything looked good!  Contrary to the repair shop quote, the sensor was not bad.

Let me repeat: 
The repair shop was going to charge me $1200.00 to fix a phantom they fabricated.

Still, the compressor won't run on it's own.  Apparently the green flashing light and no-run will occur if the compressor speed doesn't track well with the engine RPM due to slipping belts.  I diddled with the belt and other little things but became convinced that when power was applied to the clutch, the compressor immediately engaged and the compressor worked fine.  No slippage even though the green light came on.

Monday

Since the compressor worked fine and the sensor worked fine, I went back to see if there is some other cut-out that was preventing the system from running.  I figured I better check the hi/lo and operating hi/lo pressure more carefully.  Everything measured low.  I referenced a generic pressure chart that came with the pressure gauge and connectors.  The chart didn't even go below 75 degrees and I was working in 45 degree weather.  I decided the pressures probably ~should~ be low!

I decided to wait until warmer weather.

Friday after Work

Came home, changed clothes and got outside before the temperature started dropping in the evening.   Pressures looked good.  I monitored the high-side pressure as I manually clicked on the clutch coil while the engine ran.  Pressure slowly built up to about 225 psi and then the radiator cooling fan turned on and the pressure immediately dropped down to about 170 psi.  Repeated over and over.

The small diameter tube going into the passenger compartment evaporator was a bit warm.  The larger diameter tube coming out of the firewall was a bit cool.  All seemed correct.

I went back and clicked the a/c on and off a few times from inside the car.  And then it hit me. It was taking about 5 seconds for the green light to begin blinking, but I couldn't hear the clutch even trying to come on!  Wait a second.  Dan
Marinucci's Motor Magazine article was clear that the only reason for a blinking green light was a compressor running at the wrong speed.  Mine wasn't running at all!  When the 3-wire connector was back together and I sat in the car pushing the a/c button, if it was a speed problem, I should have heard the clutch at least come on for a few seconds before the car complained about mismatched speeds.

I did some more web searching about speed sensors and there were multiple rumblings about blinking green lights being fixed with a new a/c relay in the fuse box.  Toyota Technical Service BULLETIN, June 3, 2005 EL011-05 says, "Locate the A/C magnetic clutch relay.  ... If the part number is 90987–02022, please replace the relay with the countermeasured relay P/N 90987–02028.  ... If the part number does not match 90987–02022, then please diagnose problems [normally]."  Turns out the car computer is too weak to drive the original OEM relay, so they are now specifying an easier-to-actuate one.

I decided to check min "magnetic clutch relay" or "a/c relay".  I found the relay using the key printed on the fuse box top over the left front wheel well.  I pulled the relay and poked and prodded with a voltmeter on the car and an ohm meter on the relay.

With the relay out of the vehicle, I monitored the relay coil connection (the fore and aft terminals) from the car computer and indeed it was trying to turn on the compressor for about 5 seconds before cutting power.  The forward terminal measured 12v  all the time and for about 5 seconds the a/c was turned on the aft terminal was pulled low, sinking current through the ~170 ohm relay coil.  If the relay was being commanded on, why wasn't my compressor turning on?  I pulled the relay and used a 12 volt battery across the 170 ohm coil.  Nice strong click, no sticking. I went back to the car and used a wire to close the coil output contacts.  Instantly the compressor came on.  That's good.  Hmm....

I went back to the house and measured the relay output contacts and I clicked the coil.  Nothing.  Megaohm resistance across the output.  Nice click, but no electrical contact. 

Back to the fuse box, I pulled the fan relay and stuck it into the a/c relay position (it's coil measured almost exaclty the same resistance, so I knew they were compatible).  Voila!  the a/c system was working perfectly.  I didn't leave the engine on long because I had disabled the cooling fan by taking its relay, but I had positively confirmed the problem.

Required: one new a/c relay.  I did pop the old relay apart to see if I could learn anything.  Maybe a slight amount of visible corrosion in by the contacts, but nothing gross.

The air conditioning problem had been one total fiasco by a commerical repair shop:
Was it malicious intent to rip me off?  I think it was simply laziness.  They didn't care what the problem was.  They figured if I wanted to get it fixed, they would just fiddle around putting new things on the car until it was fixed.  They were factually wrong about buying a sensor, and they just didn't care to courteously save me me money by doing the minimum repair possible. 

Need I say how offensive that is to me?   They would have put a new $1100.00 a/c system in the car and it wouldn't have worked any better.  Eventually they would have found the relay.  Then what?  Say I'm sorry, and bill me $1100.00 + $17.00.  They probably would have eaten the $17.00 cost and never told me of their mistake.  I'm sort of fed up with car repair shops right now.

Last time, they quoted me $1600.00 to fix a Dodge van and really all it needed was a new intake gasket (but they put the vacuum line wrong so I had to fix my ventilation control system when I got it home), and they wanted to bill me $450.00 for an exhaust manifold I found for $75.00 on the internet.  I have wanted to believe these guys are a home-town good deal.

They are proving to me that I need to go elsewhere.  I did go back and talk to the manager. He said, "Oh, yea, I think that was Mike (or some other name I can't remember).  He should have checked the relay first before giving you the estimate.  He no longer works here."  Pass the buck.  Okay.  I got it.


For all the hobby wrench turners out there - I have been tempted to believe it's not worth my time to always be fixing cars.  And then I learned again, it IS worth my time to know enough to figure out what's wrong.

[update: nearly 4 years later in 2017, the a/c system is still working fine.]

Spring 2014 - New Rear Tires

Road trip.  new tires.  tie rod bolt sloppy loose AFTER traveling 400 miles.  Took it back to the repair shop. Fixed it with a new tie rod.  When done, they slammed tires on with a power impact wrench, which turned out to be carlessly over tight compared to the spec'd 80 ft lbs, causing a latent effect I didn't find out about until 3 years later.  ]

January 2017 - Wheel Bolt and Stuck Rotors

bolt to pop off rotorsI planned to only rotate the tires.  When I got to the rear passenser side, I unscrewed one of the wheel nuts and the entire bolt snapped.  The only way that could  happen was if the last person who put on the nut way overtightened it and stretched or fractured the bolt.  In any case, I had to get the brake rotor off so I could get the wheel bolt stud out of the hub.

My brake rotors were really rusted on.   The only way I was able to get it off was to follow an idea I found on youtube to remove stuck rotors.  The only difference is that my caliper bolt holes didn't clearly line up with the rotor.  They bumped the inner rain shield and their wasn't room for the "foot nut".   So used 3/16" bolt with only one nut.  Instead of the nut that touches the rotor, I used a flat piece of steel about 1/8" thick for the bolt to directly push against, pinched between the bolt and the rotor.  I used two of these - one of the top caliper bolt hold and one on the bottom.  I turned the nut to apply pressure , which wuold have skated all over the steel plate.  Torque was getting pretty high, I'd estimate about 60 ft lb on the nuts and the caliper flanges were starting to slightly bend.  Enough.  With that much pressure applied with two bolts, I tapped on the rotor central part a few times with a hammer and suddenly the entire rotor jumped outward and was free.

mount new stud Once the rotor was out of the way, I tapped on the wheel stud with a hammer to slowly back it out.  I positioned the new one and in order to pull it through the hub and lock it in place, I piled a stack of washers and then torqued the wheel nut onto the bolt.  As I tighted upward of 100 ft lbs, the stud pulled into and mounted firmly in the hub.


July 2018 - Air Conditioner Again

I climbed into the car and pushed the A/C button. After about 5-6 seconds, the green light on the button started flashing.  A/C was not working again.

Based on my experience five years ago (read above), I immediately checked to see if the compressor tried to come on during those 5 seconds.  It did not, so I knew the relay was not delivering 12v to the compressor clutch at all.  Incidentally, I did notice this time that when I pressed the button, the engine idle slightly rose, expecting the load of the A/C compressor.  When the light started blinking, the engine idle went back down.

I pulled the A/C relay from the fuse box and I used an ohmmeter to test the relay coil.  Instead of 170 ohms across the fore/aft connections (in the picture above), the relay coil was about 250 ohms.  In light of the Toyota service bulletin indicating that the relays were troublesome, I expect the 220 ohm one too hard to turn on. Tapping it on the bench didn't improve the situation.

I used a voltmeter and confirmed the 12 volts that is suppose to come from the car to the relay - both to the relay coil and to one of the relay output terminals. Lastly, I momentarily tapped a wire short across the connections the relay was suppose to short (the top two connections in photo above).  The compressor immediately came on. 

I need a new relay. 

After purchasing and installing a new relay, I can see and hear that the clutch works and the compressor spins.  But only for about 5 seconds.  Could it be that I had two problems go wrong at the same time?!
Ughh.. several things to check....

shorting wire

This shows the wire I used to short the outputs of the A/C relay.
ac speed sensor plug

The black wire goes to the compressor clutch.  The two grey wires are the compressor speed sensor.
clutch connect

The paperclip receives 12v from the A/C relay and the red jumper delivers it to the compressor clutch coil. 
oscope connection

Red wire enables the compressor clutch; the white and black jumpers bring the speed sensor out to an oscilloscope.
.

.

After confirming that the system does have freon pressure (about 90 psi yesterday when it was hot, and about 75 psi this morning when it was 74 degrees).  It looks pretty bubbly and foamy in the glass window, but I decided it would not ruin anything because there was oil and freon.  Running the compressor for about 20 seconds and checking the pipes showed that one got cold and one got hot.  The A/C was "working", but the car refused to keep the compressor running.

The car must be sensing low pressure or no compressor speed, so I began to check more.  The speed sensor connector is easy to get to.  The connector has 3 pins - one is the black wire that carries the 12v going to the clutch coil.  The other two are the grey wires connected to the sensor.   I used a red jumper to connect the 12v clutch wire, while I used the black and yellow jumpers to connect the inductive sensor to my meters and oscilloscope.

I measured infinite resistance from both sensor wires to ground.  I also measured infinate resistance between the wires. This doesn't look good.  The Youtube video link above suggests it should be 165-205 ohms.

I started the car and turned onthe A/C.  After 5 seconds, the light started blinking.  I used the wire to short across the relay outputs in the fuse panel to force the compressor on even though the car quit commanding it on.  Unlike 2013, the o'scope showed only random 60 Hz noise signals. Nothing at all like last time (see pictures from 2013). I think the speed sensor is broken open circuit. 

One last check.  I put a 220 ohm resistor across the a/c relay input side (simulating the relay coil) to see if the computer would apply 12 volts to try to turn on the compressor.  Yup.  For about 5 seconds.  But without a speed sensor signal, the computer drops the 12 volts across the relay coil and that causes the relay to switch open, and the clutch disengages.  The green light on the dashboard a/c control blinks green.

This is going to be an expensive project to pump out the system, remove the compressor, install a new sensor, and put it all back.  :-(

Additional Resources

Video of hi/lo pressure readings, and bubbles going by the sight window.
Pressure switch for sale.
Numbers look right, but weak cooling.
A/C relay is the simple fix!


© 2017 Brian Mork. Please contact me using the copyright link prior to commercial use, or reproducing for distribution in a commercial context.