Smoke and Mirrors

By Richard McCuistian

Smoke can be friendly.

1986 Toyota Cressida

2.8L Engine

A42DL Transaxle

96 ,549 miles

Taking in Jobs That Make Sense

At the college where I teach Auto Mechanics, there are some people who seem to think we run a regular shop in my wing of the campus. We do get a lot of live work requests, but the person who is requesting the work has to qualify; we don’t want to compete with local shops, so we basically only work on vehicles belonging to employees of the college or students who are enrolled there.

Photo:  Checking for exhaust leaks using smoke.

I know a few college Auto Mechanics instructors who tell me they just don’t take in live work any more because it can be a headache to deal with in regard to paperwork, not to mention the stressful exercise of having to deal with the fallout of students’ mistakes.  Be that as it may, from my years of field experience, it seems that students learn some things a whole lot faster on real world repair jobs than they do simply hacking their way through shop worksheets (a student that completes my program has done about 300 worksheets by the time he or she graduates), as valuable as those can be. 

Worksheets can (if properly written and executed) be structured to teach in a single exercise what some professional technicians who never went to technical college still don’t know, but there isn’t too much pressure to finish a worksheet in a timely manner.  As for live work, I push the students to work their way through those jobs with the same urgency and time consideration they will be expected to display at their future place of employment.

When I do take in live work, I generally try to make sure it provides timely benefit to the students.  It needs to be, in most cases, on a vehicle that is newer than a 1990 model.  I don’t have time in my curriculum to help people restore mid-sixties pickup trucks, and believe me, there are plenty of those jobs to go around. 

But there are other considerations. For example, if I’m teaching Transmissions, Brakes, and Steering & Suspension in a given semester, it doesn’t make sense to burn valuable lab time on a head gasket job; there are other semesters where time-consuming work like that can and will be covered. 

But the line becomes blurred in a case like the one I’m about to describe.  This semester I’m teaching the first of two Engine Performance courses, and when one of the college maintenance men asked if we could look at a warm idle misfire on his 86 Cressida, I felt the benefit to the student that drew the ticket would be measurable, and as it turns out, it was.

Old Fashioned Diagnostics (sort of)

            This semester I’m teaching the first of two Engine Performance courses, and the Cressida provided a perfect opportunity to unravel a mildly unusual driveability concern.  There was a clue to the nature of the problem in that the car ran very smooth when the engine was cool, but developed a consistent idle misfire at operating temperature. Experienced driveability guys will recognize those indicators, when they come as a package, generally point to a vacuum leak.

            Some of the earliest instruction I give my students is how to determine which cylinder is misfiring, and an old fashioned power balance test is a good way to do that. While the more seasoned techs (like me) have pulled spark plug wires for years at the risk of experiencing a volt jolt, it makes more sense to disconnect the injectors (or the COP coil primary leads) one at a time to find the dead cylinder.

Another alternative is to backprobe each spark plug boot with the point of a test light (the clip should obviously be connected to ground). The test light bulb won’t light (all volts there, almost no amps) but the dead cylinder can be located that way without the danger of a nasty shock.  Some guys actually puncture the boot instead of backprobing it, but that’s an egregious practice, and always makes a path for the spark to reach out and touch someone or something, particularly on moist days.

            The long and short of it was that as the engine warmed up, Lamont found that the #1 cylinder was dead on arrival.  Connecting the o-scope produced no conclusive results; the pattern on that cylinder was practically the same as all the others.

Digging for the Cause

            A quick look at the spark plug revealed nothing particularly interesting other than the fact that it was showing some wear and the center electrode resistance on that plug was higher than the others by about 2000 ohms.  It didn’t cost much in time or cash to throw a set of spark plugs at it, but to no avail.

         Listening to the injector with a stethoscope gave no indication that the injector was doing anything wrong, but we could do an injector flow test later if need be. 

The engine sounded as if it had normal compression on all six cylinders while spinning, but in the interest of gathering data, we fetched a compression gauge and measured the front three cylinders to see how they were puffing.  All three were pushing about or around 170.  A running compression test (which is always a lot lower than a spinning test) on those same three cylinders revealed about 75 lbs on cylinder #1 and 90 on the two behind it, but that in and of itself wasn’t very convincing.  We hadn’t exhausted all possible non-intrusive tests yet, and when I’m calling the shots, we’ll pull no rocker arm cover before its time.

Zeroing In

          Lamont was running out of ideas, but I hadn’t scraped the bottom of the barrel yet.  I handed him a can of carburetor spray (knowing that there are fire extinguishers close by) and told him to safely, carefully and sparingly mist the non flash-point prone areas of the engine compartment. Propane works for this, but tends to blow away if the fan is running. Lo and behold, spraying the cleaner down under the intake manifold on the driver side front of the engine, Lamont found that the skip evaporated and the engine ran smooth.

                If we were looking at fuel trim readings (which we had no way to read on this old Toyota), we would have seen the short numbers dropping back toward zero.

    We definitely had some unmetered air making its way into the intake, and while I didn’t reveal it at the time (students learn better by discovery than by my words), I knew there had to be an intake leak right at the point where the injector delivers its fuel.  There were two possibilities.  One would be the injector o-ring, but that wasn’t in the area where the spray garnered results.  The other would be an intake gasket problem at the #1 runner.  I told Lamont to raise the vehicle for a closer inspection of the underside of the intake.

It was crowded under there, with a lot of stuff in the way that made it hard to see, but it appeared that the intake gasket was split right at the bottom of the #1 intake runner, which would make perfect sense.  But before we tore it down, I wanted to be sure I had exhausted the opportunity to teach Lamont everything I could.

Got Smoke?

            Smoke is usually considered a bad thing, whether the source of it is a cigarette, a tail pipe, a hot and oily exhaust manifold, a wiring harness, or an electronic component.  Many of us have joked about “letting the smoke out” when we see some expensive part give up the ghost in a pungent cloud.

But in the world of emission controls, where, by today’s rules, a 0.020 inch breach of an evaporative system is outside federal standards, smoke is the best way to find a pinhole leak. Air and hydrocarbons are invisible.  Smoke is visible.  Since we can’t submerge the whole system in half-drum of water like we do tires to find otherwise invisible leaks, the benefits are obvious!

On an evaporative system, you’re supposed to seal the canister vent (by whatever means is easiest), connect a smoke machine to the evaporative test port (the one under the hood with the green cap), pump the system full of smoke, fire up your halogen light, get out your inspection mirror and you’re in business. (It’s like we’re magicians. We once had only inspection mirrors in our toolboxes. Now we can have both smoke and mirrors!)

One problem is that a smoke machine of any decent quality costs as much as a fully functional scan tool ($2000), and if you’re going to invest that kind of money in a piece of equipment, you’d better find other things to use it for.  The literature that comes with the smoke machine will give you ideas you probably hadn’t thought of about how and when to use it.

Smoke can be used to check for vacuum leaks, EGR leaks at the pintle shaft, leaks under the dash, throttle body leaks, brake booster leaks, intercooler system leaks, HEVAC air management control vacuum leaks, injector o-rings, wind/water leaks, oil leaks (especially axle and pinion seal leaks!), air injection leaks, MAP sensor leaks, and a whole lot more. Having a smoke machine opens up a whole new world of closed cavity diagnostics.  There are limitations; leaks that only show up under higher pressures aren’t smoke machine friendly, but there are other less expensive tools that can ferret those out.

Smoke is a beautiful way to check for exhaust leaks, and you’ll usually find leaks you weren’t even looking for.  Ever had a job where you fixed one or two blatantly obvious exhaust leaks only to discover that there were two more you couldn’t previously hear?  The smoke method is a shoe-in for finding all the leaks the first time around, and it’s a whole lot easier to re-test for a leak after repairs while the exhaust system is still cool instead of firing up the engine.

Most of these units come with rubber cone-shaped adapters for this purpose.   Shove the cone in the tailpipe, connect the machine, fire up the smoke, and what may have been a difficult leak to find suddenly becomes visible as a neat white plume spurts from the guilty area.

My smoke machine is an EVAPro by Vacutec®, and it requires connection to the vehicle battery along with shop air. It has a flow meter and different settings to determine the size of the suspected leak.  The machine comes with enough smoke solution to do 500 tests, and when the smoke is flowing, it has a pleasing lemon scent and leaves a UV trace behind to pinpoint the exact point of the leak.

Cinching the Diagnosis

            Checking an intake leak with a smoke machine is downright simple.  Remove a vacuum line and connect the smoke hose so that it pumps the intake full.  In thirty seconds or less, you’ll see smoke if there’s a leak, and this time was no exception.  A cloud of smoke came wafting up from beneath the intake in the same area where the carb spray had produced its results.   A steady stream of smoke was spewing from the almost unnoticeable split in the intake gasket at the # 1 runner.  My photo is a zoom-in, and it was a lot harder to see this gasket problem with the naked eye (ever thought of putting digital camera in your toolbox? It’s amazing what you can see by snapping a photo and blowing it up on the computer for a closer look!).  A new intake gasket got the Cressida purring like a kitten once again.

Conclusion

            A couple days later during some fuel system exercises, the 1997 Oldsmobile trainer vehicle was the subject of a worksheet I had issued that dealt with ferreting out a miss caused by a faulty injector, then removing and replacing the injector to find that the spark plug on that same cylinder had been bent shut.  In the process of different students working their way through the exercise, the Olds had developed a running problem that I hadn’t originally inserted, which was a skip on #4 cylinder. The first thing the two students did was grab the smoke machine.  Smoke came spurting out around the base of the #4 injector, and it became apparent that the o-ring had been damaged.

            One odious part of this scenario is that if I place these students in a shop where there is no smoke machine, they may be a bit handicapped.  I’ll say one thing; if I were to return to the field and had choose between the purchase of a smoke machine and a scan tool, the scan tool wouldn’t win the contest by much!                                                           R.W.M.