Sunday, October 28, 2012

Korean Chaos

Well, it has been a long time since I last posted and I would like to catch up. Thankfully, business has been good and I have had little downtime. Here is a 2004 Kia Optima with a 2.4 liter 4cylinder motor. I see a ton of Hyundai/Kia vehicles in my area. They are very popular. The shops concern is the vehicle seems to run pretty decent with a slight lack of power but once warmed up has a very shaky idle. There no codes. I first look at some fuel trims which seem to be acceptable. I road test the vehicle and no doubt it is slightly underpowered and indeed has a shaky idle. I decide to look at engine vacuum to start with because the vehicle feels like the EGR valve is stuck on at an idle.
As you can see 14" of vacuum at an idle. I normally see these vehicles produce 19" of vacuum on a good running vehicle. The low vacuum is a clue. The next thing I do is feel the EGR valve with the vehicle running. It feels pretty cool the touch. Now, this is not an absolute foolproof diagnostic test. However, typically if an EGR valve is flowing some exhaust gases it will be warm to hot to the touch at an idle. Warning, do yourself a favor hit the EGR with a non contact laser thermometer first before putting your hand on it. Ok, EGR not flowing at an idle, what is next?

I like to scope things. I feel it gives me the best bang for the buck. In my business remember I need to be fast and efficient. I want to look at crankshaft sensor to camshaft sensor patterns and how they correalate to each other. If you notice the cam sensor is driven off the exhaust camshaft of this car.
Here is the CKP sensor pattern in yellow on channel 1 and the CMP sensor pattern on channel 2 in green. Call it my OCD or whatever I typically always use channel 1 for CKP sensors, it keeps captures consistent. Ok, is this a good pattern? This is where data acquisition comes into play. So I pull up a known good waveform from my library.

Here again CKP is channel 1 in yellow and CMP is channel 2 in green. What I can see immediately is that the Cam sensor pattern either intersects the Crank sensor pattern halfway or straddles it midway. Lets revisit the vehicles waveform at a different timebase.
It is shifted no doubt. Be careful using this method. The pitfall is that the Cam sensor is only on one camshaft. So we could have a situation where one camshaft that drives a sensor is in time and the other that has no sensor is off. When presented with that situation I use running compression with a pressure transducer and look at exhaust and intake valve opening events. Here, I didn't have to. I advised the shop to check timing belt alignment. Sure enough they called me back later that day after resetting the timing belt. The vehicle performed and idled well.

I recently was at a shop that just did a timing belt replacement on a 2004 Hyundai XG350 3.5 liter v6 with 75,136 miles as part of routine maintenance. The vehicle ran well into the bay. Unfortunately, it ran quite poorly after the service. It was setting misfire codes and had a terrible lack of power and a rough idle. These cars normally idle like glass and perform quite well.
The tech that did this job was sickened. He is a very thorough tech and I have been there a couple of times in my life where the vehicle ran worse after I serviced it. It is a bad feeling. I really felt for him. I went over the timing belt procedure with the tech and on the car several times. I often find that more can be found out asking questions than looking at the car. After some head scratching the tech said something that stuck with me. He said he had a hard time seperating the crank pulley from the crank sprocket. With this I broke out the scope and decided to check CKP/CMP correalation. I had a hunch.
Here again CKP in yellow and CMP in green. I didn't have a known good pattern handy. But, I felt very confident that it would follow suit like the Kia previously. Here, is another shot.
Cursor 1 is where the rising edge of the CMP sensor is and cursor 2 is where it should be. I know what happened. I tell the tech to get the timing belt off and remove the crank sprocket and CKP interruptor ring behind the crank sprocket.

Here is the interruptor. Those blades pass through the CKP sensor causing the square wave pattern. If you notice there are two dimples on the ring that should not be there.
Here is the backside of the crank sprocket. The two protrusion tabs that should go into the holes on the CKP interruptor ring are worn off.

Here is how it fits together. What happened here is when the tech was seperating the crank pulley from the sprocket it pulled the crank sprocket forward and disengaged the CKP interruptor ring from the crank sprocket. When the tech reinstalled everything it moved and therefore corrupted CKP/CMP timing. This caused the vehicle to run like the timing was retarded drastically. This was very easy to do. Hyundai does not extend the crankshaft woodruff key all the way in and relies on these two small tabs that are easily broken to index everything. One new crank sprocket and interruptor ring and away it went.