Written by Chris Dekker, former co-owner at Tools in Motion.

If you spend any time on the Internet at all, you’ve surely seen ads like these by now. Besides the clickbait mechanic-bashing, what else is wrong with this advertisement? Let’s dig in.

One of the primary aims of our blog and social media presence is to give our customers a behind-the-scenes look at what we do every day, and to educate people as well. A better-informed customer will be able to take better care of their vehicle, after all. If we had to pick the one aspect of our job that is the least understood by customers, it would be the process of diagnosing a car problem. That’s why we re-visit this topic as much as we do.

There’s still a large group of people out there who honestly believe that we just hook up a computer that tells us what’s wrong with their vehicle – like the device in the photo above claims to do. This device is a generic OBD2 code reader that’s meant to work with your phone. You can actually purchase something similar on Amazon for around $10, or a standalone tool at many local stores for between $20 and $250.

What is a code reader? (The simple explanation.)

In 1996, every vehicle on the road adopted a self-diagnostic system called On-Board Diagnostics 2 (or OBD2 for short). It was around this time period that on-board computers – or “modules” – were becoming much more common, and more plentiful. 1980s vehicles usually just had one module that controlled the engine and transmission, while 90s cars had 5-10 modules; and today’s vehicles have 20-30. Most of these modules contain “self-diagnostic” software that is designed to detect bugs in the vehicle. When a fault is detected, a warning light is usually illuminated, and a “trouble code” is stored. For example, if an Engine Control Module detects a misfire on cylinder #2 of your engine, it would set a trouble code P0302 – “Cylinder #2 misfire”. OBD2 codes are always 5 digits in length, with the first letter dictating which system is affected: “P” for powertrain, “C” for chassis, “B” for body, and so on.

An OBD2 code reader is a very basic version of the full-feature scan tools that professionals use, and allows you to retrieve most (but not all) of the “P” trouble codes stored in just your engine control module.

So, how does a code reader tell you what’s wrong with your car?

It doesn’t. A trouble code is just a starting point in the diagnosis: a direction, if you will. That P0302 trouble code mentioned above doesn’t tell you if your misfire is being caused by a bad spark plug; ignition coil; fuel injector; vacuum leak; sticking valve; compression problem; or a dozen other possible causes. That’s where the rest of the diagnosis comes in. We’ve discussed the actual diagnostic process lots in these previous articles, so we’ll skip over that for now:

The “magic computer”? We still don’t have one.

Explaining the diagnostic process.

Real-world diagnostic examples:

To give you some examples of how little help a code reader usually is in properly diagnosing a problem, we want to show you three examples of problem vehicles that came through our shop recently. We’ll list the symptoms, with any trouble codes that were stored. Look at this information first, and make a mental note of which components you’d lean towards replacing based on that information alone – then read on to learn what the results of a full, proper diagnosis were. We’re not cherry-picking these examples, either; these were literally the last three diagnostic jobs to come through our shop.

1)  2010 Mercedes GLK 350 with a rough-running engine.

This Mercedes came to us with a very rough-idling engine, and a “check engine” light on. Diagnostic accuracy is always important on European vehicles, because their parts aren’t cheap and one wrong guess would cost a lot more than a proper diagnosis. Our initial scan revealed two trouble codes in the Engine Control Module (ECM):

• P0300 – Random cylinder misfire detected.
• P2005 – Intake manifold runner tuning valve position error.

It’s important to note that this is where diagnosis with the code reader stops. Do you feel comfortable throwing any parts at this car yet? We didn’t either, so let’s continue…

To an experienced technician, this engine had what felt like a single-cylinder misfire. But which cylinder was misfiring? As we often see, the ECM didn’t know; and stored the rather unhelpful “random cylinder misfire” trouble code instead. Performing a cylinder contribution test with our Mercedes scan tool allowed us to determine that cylinder #2 was our culprit. We did a quick check of mechanical compression in that cylinder – which measured good – then proceeded to check for fuel and spark supply to cylinder #2. We found that the cylinder #2 ignition coil was producing a weak spark, and had to be replaced.

The new coil solved our rough idle, but diagnosis of the variable length intake manifold (a system designed to provide a good combination of low-end torque and high-end power) took a little more work. When none of the obvious causes for the P2005 code were present, we dug deeper. Using a special camera called a borescope, we went inside the intake manifold and found some physical damage that meant the manifold will need to be replaced.

Total cost for this accurate diagnosis: $150.

2) 2008 Hummer H3 with Check Engine light, ABS light, and Traction/Stability control lights on; four wheel drive will not engage; rear locking differential will not engage.

Because of all the different issues on this truck, starting diagnosis involved checking for trouble codes in all of the on-board computers; not just the engine computer that a generic code reader can access.

• The Sensing and Diagnostic Module (which controls ABS, traction and stability control) and the Final Drive Control Module (which controls the four wheel drive) had both stored trouble code C0045 – “Left rear wheel speed sensor signal fault”.
• The only code stored in the engine computer was P0300 – “Random cylinder misfire detected”, like our Mercedes above.

Since the ABS, stability control, four wheel drive, etc are all dependent on knowing the wheel speeds, our C0045 trouble code explained all of the customer’s issues except the Check Engine light.

Does our Hummer need a new wheel speed sensor? In this case, that would have been a bad guess. The sensor tested fine, (and here’s where it’s important to point out that every part on a vehicle can be tested before it’s replaced) so we investigated further. The issue turned out to be a chafing wiring harness underneath the vehicle, where several wires had rubbed through on the frame. We repaired the wiring and moved on to the Check Engine light.

The engine in this Hummer ran fine until we got it hot, when it started misfiring. Whipping out our GM Tech 2 scan tool for a power balance test, we determined the offending cylinders were #4 and #5. As we’ve already discussed, a misfire can be caused by dozens of different problems, so we ran a battery of tests, using many different tools and pieces of test equipment. Having ruled out a lot of the more common causes, we installed an in-cylinder pressure transducer (super-sensitive electric pressure sensor) into cylinder #5 and wired the transducer to our oscilloscope. Using this tool, we were able to detect the cylinder #5 intake valves were sticking. A little more camera time with the borescope, and we determined this was due to carbon build-up inside the cylinder heads, which would need to be cleaned.

Total cost for this diagnosis (including the wiring repair): $300.

3) 2012 Toyota Tundra with Check Engine light on.

This diagnosis was pretty straightforward. Scanning the Engine Control Module only resulted in one stored trouble code:

• P0441 – “Evaporative emission system incorrect purge flow.”

This trouble code indicates that when the ECM applies vacuum to the fuel tank by commanding open the purge valve, it is not seeing the expected drop in fuel tank pressure, as reported by the fuel tank pressure sensor. Let’s stop for a minute again. This is as far as the code reader takes us. Do you feel comfortable replacing a part yet? A faulty purge valve would be a good guess, but that’s a $200 part. What’s a proper diagnosis going to cost?

Possible causes for this trouble code include a faulty purge valve; a failing fuel tank pressure sensor; a leak in the system somewhere; even a loose gas cap! After checking the gas cap – you ALWAYS check that first! – we moved on to testing the purge valve using a Toyota scan tool that can command the valve open/closed, and a handheld vacuum pump. The valve functioned fine and all of its plumbing looked good, so we moved on to looking for leaks in the system. The quickest way to do this was to use a smoke machine, a tester than pressurizes the system and fills it with a thick smoke. Once you find the smoke escaping somewhere, you’ve found your leak.

We found the smoke escaping from the canister vent valve underneath the vehicle, which we had used the scan tool to command closed for the test. Removing and bench testing the solenoid revealed that while it was receiving a good power and ground input from the ECM, the solenoid itself was not closing properly – and was the source of our trouble code.

Total cost for this diagnosis: $150.

Hopefully these examples help illustrate just how valuable a good diagnosis is, and how much work (and technician skill) goes into one. Do you have a vehicle problem that needs trouble-shooting? Let us know!

Written by Chris Dekker, former co-owner at Tools in Motion.

“Do you have one of those computers that tells you what’s wrong with the car?” This has got to be one of every mechanic’s least favourite questions to get from a customer. As we’ve explained before, while scan tools can provide a starting point for a proper diagnosis, it’s very early in the process that the human brain must take over. As we often tell people, there is a big difference between pulling out a trouble code and actually diagnosing a problem.

Here’s a great example from this week. This little Honda CR-V came to us with a “check engine” light on, and the engine idling rough.

We connected a scan tool and retrieved the stored trouble codes from the engine computer, or ECM. As we often see, the stored codes were of no help, as the engine had set misfire codes for all four cylinders. We already knew the engine was misfiring, and the trouble codes don’t tell us why the misfire is happening.

Even though this vehicle’s rather simplistic engine computer is flagging misfires on all four cylinders, the issue really felt more like a consistent misfire from a single cylinder. Removing and shorting out the spark plugs wires one by one, we determined that the engine was misfiring on cylinder #4.

OK, now what? We’ve got lots of possibilities here: the issue could be a bad spark plug or plug wire; a problem with the distributor (yes, this car still has one); a faulty fuel injector; or about a dozen other things. Removing and inspecting the cylinder #4 spark plug seemed like a good place to start. As it turns out, the spark plug and wire were both in good condition.

We noticed, however, that the spark plug was a bit wet with fuel. Having already ensured we had a strong spark supply to the plug from the distributor and coil, this could only mean one thing: The spark plug was firing; the fuel injector was firing; but the combustion event was not taking place inside the cylinder. The next logical step seemed like performing a compression test. We installed our compression tester in the #4 spark plug tube; disabled the ignition system; and cranked the engine over. As it turned out, cylinder #4 was only making about 40PSI of compression! (A good cylinder on this engine measured around 160 PSI.)

The low compression was definitely the cause of the misfire. Every engine needs at least 100 PSI per cylinder to “get the fire going”, so to speak. Now it was time to determine why cylinder #4 had low compression. Like before, there are lots of possibilities: it could be a burnt/bent valve; worn out piston rings; or a handful of other things. How do we determine where all that lost compression is going? We install a cylinder leakdown tester. Out came the compression tester, and in went this next tool.

We rotated the engine until cylinder #4 was on its compression stroke, with all the valves closed. Using the leakdown tester, we filled the cylinder with compressed air. As you can see, this cylinder has about 85% leakdown. (20% is the most we’d ever like to see on a good engine.)

We can also use the leakdown tester to determine where the leaking compression is going, by listening for air leakage at different points on the engine. Air coming out of the intake manifold or throttle body indicates a leaking intake valve on this cylinder. Air hissing from the tailpipe indicates a leaking exhaust valve, and air leaking from the oil cap points towards a leak into the crankcase via worn out cylinders and/or piston rings. This vehicle had none of these leaks. The compressed air was actually leaking from the cylinder #3 spark plug hole, indicating there is a blown head gasket or other combustion leak between these two adjacent cylinders.

The next step in diagnosis will be to remove the cylinder head for inspection, and likely replace the leaking head gasket.

Almost every warning light diagnosis works this way. The trouble codes (sometimes) provide a starting point, and then there are usually many other tests that must be performed – using even more specialized equipment – to “zero in” on the route cause of the issue. With some issues, there are no codes stored at all, and the technician must let the symptoms and their experience lead them in the right testing direction.

Every good diagnosis goes like the one on our Honda this week: A well-trained technician knows exactly what test to perform next based on the symptoms at hand, and lets the results of that test tell them what test should be performed afterwards. There is no wasted time troubleshooting parts that don’t need to be checked – and more importantly, no money wasted replacing parts that won’t fix the problem.

This is the value of a good diagnosis by a qualified professional.

Written by Chris Dekker, former co-owner at Tools in Motion.

During the 1980s, there was an important shift in the automotive industry: carburetors were out, and fuel injection was in. While a carbureted engine basically sucks in a stream of liquid fuel, a fuel injected engine uses an electronic fuel injector (which is controlled by a computer) to deliver a burst of high-pressure fuel mist at precisely the right moment. Think of a Windex bottle; the stream setting mimics a carburetor, while the mist setting is your fuel injector.

Most fuel injected engines from the last 20 years use port fuel injection, where fuel is injected upstream of the cylinder into the intake manifold. The fuel spray is carried past the intake valve, and into the cylinder (where it is burned) by the flow of air into the cylinder. Remember this part about fuel passing through the intake valve, as it gets important later!

Liquid gasoline has a hard time burning; it’s the vapor that really gets things going inside your engine. The mist delivered by a fuel injector atomizes easier, and burns more completely. This more complete burn – combined with the improved control from having a computer running things – translates into better fuel economy and decreased emissions. Now imagine that we could take that mist from your Windex bottle, crank the pressure way up, and shoot that pressure through tiny holes so small that the human eye can barely see them. The fuel spray would be super fine; how well would that burn?

This idea is part of a new technology called direct injection. Direct injection also offers another advantage: it delivers its super-high-pressure fuel spray directly into the cylinders of your engine, right beside the spark plugs. Compared to port injection, direct injection offers even better control of the combustion event, and even better performance. Gasoline direct injection (GDI, for short) is getting more common in our industry, as car manufacturers continue to chase better fuel economy and cleaner emissions. GDI can help a small engine to make as much horsepower as significantly larger one, as Ford has demonstrated with their popular Ecoboost engines.

However, moving the fuel spray from the intake manifold to inside the cylinders had an unexpected conseqeuence: bad deposit build-up on the intake valves. It turns out that the continuous fuel spray onto the intake valves in a port-injected engine did a good job keeping them clean. Deposit build-up on the intake valves of GDI engines has become a big problem; causing misfiring, hard starting and poor performance.

This nasty build-up is composed of:

• Carbon from the combustion of fuel, which is pushed back out of the cylinders.
• Trace amounts of engine oil that leak past the valve seals, or are carried into the intake manifold by the crankcase ventilation system.
• Exhaust soot that is recirculated back through the intake manifold by the EGR (an emissions control) system.

This all combines to create a rock-hard, clumpy layer of junk on the intake valves that gets thicker and thicker over time. Eventually it builds up to the point where airflow through the valve is restricted, or the valve cannot open and close properly.

The photo above is from a Mazda that we serviced this week, and is the worst example of this build-up that we have seen to date. This engine had less than 150,000 kilometers on it, and a bad misfire that was due in part to the valve issues. With this Mazda, we decided to do a little experiment. As GDI becomes more common in our industry, so too do the problems with intake valve deposits. Several companies have responded with various products designed to clean off these deposits, including CRC with their GDI-specific “Intake Valve Cleaner”. But do any of them work?

Most of these products are a liquid designed to be “fogged” into the intake manifold as the engine runs, where they will travel through the intake valves and hopefully remove some of the deposit build-up as they pass by. The trouble with these products is that you normally never get to see how well they work (or don’t work), since the valves are hidden deep inside the engine. With this Mazda, we had the intake manifold removed and full access to the valves – so we decided to try several popular products to clean the valves directly. We chose CRC Intake Valve Cleaner, Sea Foam Motor Treatment, ACDelco Cleens Combustion Chamber Cleaner, and a last-minute crazy idea: Wipe-Out, a very powerful gun cleaner designed to remove carbon and copper fouling from rifle barrels.

First we sprayed each of the valves with one of the cleaning products, let them soak for a couple minutes, then blew out the intake runners with compressed air. Initial results were disappointing: none of these products magically dissolved or removed any of the build-up through contact with the valves alone. Experience told us not to expect this anyways, but it would have been nice!

Next, we tried a longer, ten minute soak and then a good scrubbing with a small toothbrush before blowing out the ports again. A tiny amount of deposit was removed from each valve, with all of the products working about the same, but we were barely making a dent. We could tell that completely cleaning the valves this way would probably take days; not an economical choice for the customer!

It was the end of the day, so we decided to fill each intake port with cleaning product and let the valves soak overnight. The next morning, we gave each valve a good picking-at with a dental pick, and then a scrubbing with the toothbrush before blowing out the ports. The Wipe-Out seemed to work the best, but barely. We were hoping that “thinking outside the box” would produce an industry-leading breakthrough here, but it was not meant to be. All of the products loosened up an outer layer of the build-up, but 80% of the deposits remained.

One important note: Most of these products are designed to work in the presence of heat on a running engine, which should make the carbon build-up easier to remove. This was impossible with the engine disassembled, of course. However, we’d like to believe that if overnight soaking, scrubbing and brushing will not remove much of the build-up, 2-3 minutes of misting onto the valve will not work either.

At this point, it seemed that the various cleaning products, like much of the snake oil bottles on the parts store shelves, were a chemical solution to a mechanical problem, and just weren’t going to work. No chemical was going to remove these deposits, and pulling the cylinder head to remove the valves would cost thousands of dollars. There had to be a better way. It was after a little research that we came across another popular solution: sand-blasting the valves with crushed walnut shells.

The crushed walnut shells are abrasive enough to remove the deposits, but not enough to remove metal from the valves or cylinder head. After a quick phone call, we had a big 30 pound bag of crushed walnut shells to play with, and decided to try this technique out. Now, we were making progress! It still took a lot of careful work, but the sandblaster did a nice job of cleaning the valves.

The walnut shell grains proved a bit too large for our recovery system to handle, though. A few hours later, poor Richard had walnut shells everywhere – in his hair, his clothes, and even his nose – but the Mazda had some much better looking valves!

The crushed walnut shell technique offers a good value in that it’s much more effective than most cleaning methods, and still relatively quick; definitely the way to go in a case like our Mazda here.