Matrix in Your Mind

By Richard McCuistian

When the shop manual comes up short, the only trouble tree matrix available may be the one you build in your own mind.

1992 Lexus SC300

94,658 miles

3.0L Engine

A340E Transaxle

When “the BOOK” Doesn’t Cover It

It can be quite easy to get lost in a shop manual troubleshooting matrix to the point that you don’t even know why the next step is necessary or why you are performing it. And in too many cases, we can come to the end of the matrix to discover that we haven’t fixed anything. 

The point is that you can’t switch your powers of logic off when you start following the steps in a trouble tree.  The instructions at the beginning of a matrix may be where the answer lies, i.e., check fuel quality, etc.  I worked on a brand new 1989 Taurus rent-a-car for a few days before discovering that some yahoo had pumped some diesel into the fuel tank before returning the vehicle. 

I saw a Power Stroke diesel come through my service bay with low fuel pressure due to a weak fuel pump, and a resulting hard-start concern.  After the fuel pump was replaced and the pressure was normal, the truck wouldn’t start but would puff out massive clouds of white smoke due to faulty adaptive learning numbers borne of the previous low pressure concern; the PCM was ratcheting Injector Control Pressure up to about 3000 psi trying to compensate for a low fuel pressure problem that wasn’t there any more.  There was nothing in the book about that.

One strong point that can be made here is that engineers don’t always write good shop manuals because there is so much they have known about their product for so long, that they think everybody knows it almost as well as they do, and if that weren’t enough, problems can develop in the field that they never encountered or even considered.  That’s what TSB’s and special service messages are all about.

Hot Potato

This month’s ride is a 1992 Lexus SC300 that had, according to the owner, been to half-dozen or more shops, none of whom had been able to do anything except replace parts and make suggestions, an experience that will empty a fairly rich bank account in short order.  While the car did run, it didn’t run well, and you could have thrown a match at the gasoline mist that was coming out of the tailpipe and made a nice pyrotechnic display. 

This young fellow had purchased a throttle body for his Lexus from a salvage yard and installed it himself, and the salvage yard part made him more than five hundred bucks poorer, but the gasoline mist was still blowing out the tailpipes in a poisonous cloud.  In true Lexus form, this throttle body has two throttle plates stacked in it with a throttle plate position sensor for each plate, an actuator for the Traction Control throttle plate and a stepper motor style Idle Air Control.  So much for that!

As a result of having experienced one disappointment after another, he had a jaded view of virtually every shop in the area (even those he hadn’t visited), especially since all the ones he had hired worked on the car and charged him for repairs that weren’t repairs at all. This wasn’t an intermittent concern, either! It still boggles my mind to see shops taking in jobs and then turning out cars that run just as bad or worse while charging the customer as if they had fixed something.

At the college where I teach, jobs like this are a powerful challenge for burgeoning students; furthermore, my pledge to this customer was that we would try to find his problem without spending any more of his money if we could.  It might cost him to have it fixed, but since my students don’t get paid labor, it wouldn’t cost him to have us check it.

Digging In – Building the Matrix

        This little machine is equipped with a 3.0L I-6 similar to the power plant Toyota stuffed under the hood of the Cressida, and it’s a peppy little mill when it’s running right.  Well, as aforementioned, this one would run but not well, and initially it seemed that whatever was causing the problem was affecting all the cylinders equally, because it didn’t seem that any one cylinder was misfiring.

“When a fuel injected engine is getting too much fuel,” I explained to the students, “what we have to determine is whether the PCM is making the decision to put too much in there or whether the fuel is coming from an unauthorized source over which the PCM has no control.” 

For an easy first check, I pulled the fuel pressure regulator vacuum line, and looked for fuel there, which would indicate that the regulator diaphragm had been breached and fuel was passing through that vacuum hose into the intake.  The line was dry.

A peek into the 1992 Lexus symptom chart was rather vague; I had been to an identical chart a few months ago on a 1995 LS400 no-start we tangled with and found it to be no help at all, so I built a trouble tree matrix in my own mind.  Those of us who are diagnosticians at heart have all done that without labeling our thought process as such, and in many if not most cases, it’s the best way (if not the ONLY way) to find a problem like this one.  A plan of attack that doesn’t involve expensive parts swapping is always wise.

In so doing, we work a little harder to check the stuff that makes sense and find out what’s NOT wrong so we can home in on what IS wrong like missiles on a target.  In some cases, a known good part is the only choice, and I hate those situations, especially when a known good part is costly and/or isn’t on the shelf.

A.                Should we pull the flashout codes?  Maybe, but with a fuel dump like this (this wasn’t black smoke, it was a white mist of raw gas) one it made more sense to check the fuel pressure.  It was a foregone that there would be oxygen sensor/fuel mixture issues.

B.                 Should we check the fuel pressure?  Good choice under the circumstances.  It didn’t seem likely that it would be an ignition concern.

            Connecting the fuel pressure gauge to this Lexus was an adventure. Toyota and Lexus engineers (and other Asian makes) have had trouble understanding the need for a fuel pressure test port, and even some of the domestic platforms have been moving away from that tech-friendly feature of late for some reason.

On the mid ‘90s Lexus LS400 V8, you simply watch a small screw on the pulsation damper to see if it moves when the key is switched on.  If it does, and the stuff in the rail is gasoline and not water or diesel, you have enough pressure to get the car started if all else is okay.

Not so with our SC300.

The fuel supply on our problem car runs through the filter underneath the driver side rear with simple inverted flare nut fittings (one possibility for a pressure test point) and travels to the fuel rail through a line that connects at the engine with a banjo bolt.  In the $500 MAC fuel pressure test kit we have on hand in my department, I found a fitting that would replace the banjo bolt and provide a connection point for the MAC fuel gauge quick coupler.  It was fairly simple to make this connection on a lift in our shop, but it would be a drag on a rainy highway at night somewhere in the Utah wilderness.

If the pressure was really high, the fuel pressure regulator or return line would bear further investigation. 

With the gauge installed and the key switched on, the pump ran for 2 seconds and the fuel pressure bounced up to a healthy 40 psi, but it didn’t hold pressure after the fuel pump’s 2 second run.  It should.  Even with no fuel leakage into the engine a loss of residual fuel pressure causes hard starting with a hot engine.

Vectoring In On The Problem

            Now that we had identified the fuel pressure loss as a concern, we needed to find out where the fuel was going; since there were no external leaks the fuel had to be going one of three places, and the data had already gathered pointed toward an injector or injectors.  On our SC300, I wanted to do a systematic diagnosis before arbitrarily diving into a nuts-and-bolts injector inspection.

C.   Block the return line and check for fuel pressure leakdown.

   We pinched the return line near the fuel rail (it was plain old rubber hose) with pliers to see if the regulator was dumping the fuel back through the return.

Result: The fuel pressure still didn’t hold at pump shutdown with the return line pinched off.

D.  Block the pressure line to see if fuel is bleeding back through the fuel pump check valve.

    I already suspected a mechanically leaking injector but for instructional purposes we took measures to pressure up the fuel system and block the pressure line at the filter to see if we still had pressure loss. If there was no pressure loss here, then the pump was the source of the leakdown, and the plot would thicken.

Result: The fuel system still wouldn’t hold pressure.  The problem had to be a leaking injector.

5.        Remove the fuel injectors and mechanically inspect each one for leakage.

Removing the injectors on this 3.0L is no small feat.  The upper intake has to be removed, and some of the fasters aren’t much fun to access. Furthermore there are more than a few hoses and connectors.  Finally, we peeled our way down to the injectors, removed the fuel rail and the injectors, and checked each one by attempting to push compressed air through it with a rubber tipped blower. Energizing each injector a few times to check for intermittent sticking made no difference.

Result:  No mechanically leaking injectors.

6.      Reinstall the injectors and re-test the system for leakage.

This was where I moved away from conventional test procedures.  With the injectors installed but electrically disconnected and with the fuel filter removed and the fuel pressure gauge still connected at the banjo bolt point, I attached a 3/8 fuel hose to a cylinder leakage tester, raised the lift, and fed 40 psi of regulated shop air through the now-empty fuel line and fuel rail just forward of the fuel filter. 

This would give me an audible leak point that would be easier to pinpoint.  The air pressure showed up as 40 psi on the fuel gauge. I pinched my supply hose to trap the air and it didn’t leak down.  This was getting interesting.  I left the air pressure on the system, lowered the vehicle, and electrically connected the injectors.  With a drain pan in place to catch the fuel from the pump, I switched the key on and saw the gauge reading drop to zero.  I had just dumped my trapped air pressure and I could clearly hear it hissing through an injector into the intake.  Disconnecting the injector connectors one at a time I found that #4 was the open nozzle.  It could be a shorted injector control wire, but when disconnecting the PCM eliminated the concern, I knew I had a PCM with a shorted driver.

Closing Out the Mental Matrix

     List price on a new PCM destroys 10 Ben Franklins and one Andrew Jackson.  Checking with a networked salvage yard revealed less than 10 used Lexus SC300 PCM’s nationwide.  We opted to send this one off for repairs (there were no cores available) and in ten days it came back repaired.  The bill was just under $400.  With all that fuel dumping, an oil change was in order, but we did that while the PCM was away.  I started the Lexus and ran it for awhile to clear the puddled gasoline from the exhaust system.  When the learning process was done, it ran like a sewing machine.  

While ‘the book’ mentions possible problems with the injector circuit in its troubleshooting table, our set of home made tests brought real-world success to a problem that had stumped more than a few other techs, but let’s be honest:  If the problem had been an intermittent one, it would have been a whole lot harder to find.

R.W.M.