MIL or Check Engine Lamp Testing pt.1

The Check Engine Light is on. Why? Lets start at the beginning. Your Check Engine Light, Malfunction Indicator Lamp or whatever name yours goes by currently is actually just a small bulb in the instrument panel. We are going to call it the MIL. When your car is running positive voltage is at the MIL feed circuit. I've made an extremely sophisticated mock-up to refer to.

Depending on make, model and design, the MIL will come on when you turn on the key and maybe stay on a few seconds when first started. Those are bulb tests so you can see that the bulb does work. If the MIL stays on while running or comes on while driving there is a problem. It could be a short on the ground circuit but that is very very rare. It will usually be because the computer has detected a problem. The computer will either ground the circuit directly, or request another computer to ground the circuit, again depending on the particular system. Next then is to determine if the computer sees a reason to turn on the MIL and what that reason might be. This is when a code reader comes into the picture. When the computer determines there is a problem it is not only supposed to turn on the MIL but it is also supposed to store a trouble code related to the problem. 

There are basically two diagnostic systems built in to the MIL systems. There is a federal government mandated OBDII compliant and an OEM (manufacturer) system. OBDII readers are very common and usually not expensive. When you go to a parts store and get codes read, it is usually an OBDII reader. More expensive readers and scan tools can usually read both OEM and OBDII. It depends. There are a lot of diagnostic systems out there and a lot of testers. Over the years I've used and purchased several. So let me introduce you to my little friends. 

I have an AutoXray EZ-Scan 6000 that I've had for years. It does OEM checks on old OBDI systems up to 1995. Rare usage these days for that but it does still work as an OBDII generic reader. Here is a 2002 Taurus with MIL on and the stored code. I'll get into freeze frame and data later. For now, generic OBDII codes.

AutoXRay EZ-Scan 6000

An old OTC Genisys 2.0. Just using OBDII generic here as well. 

OTC EVO Genisys 5.0 & 4.0

Snap-on Solus Pro, Ultra and Edge

OTC Encore

Actron AutoScanner Plus (older and later models)

NGS MachII

OTC Monitor 4000Enhanced

Innova 3110

Alldata Mobile (tablet)

Actron U-Scan

You might have noticed all the codes were pulled from the same vehicle using each of these code readers. You might be wondering why I would have 15 different tools to check for codes. I don't really. I have 18, but was only using the ones that would pull generic OBDII codes from this particular car. Some do more tasks, some do less. Some are very expensive, some are much less so. Some are latest updates, some are better for older cars. Some are for quick basic code reads, some for more extensive testing. Remember that the cars have more than generic OBDII codes and systems.
An automobile manufacturer tool is fully capable of doing many many sorts of tests, commands and updates for a particular make of car but in the general repair shop and aftermarket we usually see many different automobiles and to be loss costly it makes sense to go with aftermarket tools.
This OTC Encore can run Ford OEM self tests.

There are also tests that may not be available on one aftermarket tool, yet present on another. You have to use the strengths of each available tool to help in the diagnostics. When you are using a generic code reader you should be aware that you may not be seeing all the stored codes. There may be OEM system codes that don't show on the OBDII reader. Good to remember if you have an MIL on but show no codes on your reader. A look at OEM codes may reveal stored codes. Also, there are many "sources" that make it seem as though all you need is codes read and immediately know what is wrong with your car. That isn't true at all. When I try to explain there is no machine that tells you what is wrong I don't always feel like I am believed. I think often the person I'm talking to just figures I must not have one. There is a LOT of false information out there so I can understand it being confusing.
So just how useful is this P0136 OBDII code? It gives a starting point. The brief code definition given by the reader says O2 Sensor Circuit Bank 1 Sensor 2. An in-line engine design has all cylinders in a single row on the engine block. As far as "bank" there would only be one. On this Taurus 3.0 we grabbed codes from though it is a V-type block. There are two banks of cylinders. Bank 1 would be the side with cylinder #1. Bank 2 would be the other side.

Bank 1 on the Taurus then is the bank nearest the firewall. The code is for bank 1 and sensor 2. The sensors are numbered as they come from the engine exhaust manifold. Bank 1 sensor 1 would be the first O2 sensor in the exhaust of bank 1. Bank 1 sensor 2 would be the second sensor in that banks exhaust. Sensors in front of the catalytic converter are usually for fuel control and sensors after the cat are cat monitor sensors.  

P0136 then indicates a problem with the catalyst monitor O2 on bank 1. B1S2. Since the post converter sensor isn't for fuel control there will probably not be a performance problem. The post converter sensor lets the computer "see" how well the converter is working, or how poorly. When the cats getting warmed up and are doing their job you can see it by watching sensor activity. Fuel control sensors should be switching well from lean to rich at full range. Monitor sensors, not so much. Sort of a slow and less active version of the fuel controls because it is after the exhaust has gone through the converter. 

This is from a 99 Chevy P/U. Both banks and both upstream and downstream sensors. You can see the difference. 

Another from a Jeep that was in recently. You can see that the bank 1 cat is not working at all. Sensor 1 and sensor 2 on that bank are mirror images. The bank 2 cat is bad as well but working a bit. When you see a monitor running like the one in b1s2 here you will get a P0420 for poor cat efficiency. 

This one is from the same Ford we captured the code from. P0136 B1S2 Circuit Problem. 

To figure this MIL problem out we have to find out what actually sets a P0136. The code would indicate the sensor is staying near 0 volts. Ford is pretty vague on the details. But we can see in the data waveform the sensor is actually near 0 volts. 

At the most basic, if all your tool does is read codes you would need to inspect the wiring, sensor and exhaust relative to the b1s2 sensor and then monitor the signal with a meter. If the wiring, etc is good then watch the signal while feeding propane to the intake. The richer fuel should make the sensor react. If it doesn't and all else was good then you do have a bad sensor. None of that is easy or simple. Point being, a code doesn't tell you what is needed to fix. The code, when it isn't misleading, is about what needs to be tested. 

I'd like to continue this as a series. I hope it is coherent. 

To start to find out why your MIL is on you will need to check for stored trouble codes.

There are OBDII self diagnostics and there are OEM self diagnostics.

You can start with a basic OBDII code reader.

You may need more to do useful testing.

I think the next level up from OBDII code reading only would be OBDII freeze frame data. Hope to cover that next. 

Thanks for reading!

Kenneth Hayes

G&G Auto Repair

2002 Taurus: Overheating?

Came in to work this morning to find a 2002 Ford Taurus SES parked along with a note. Overheating problem, it says. To get started on the diagnosis I drove the car into the shop. The A/C was on when I started the car. The cooling fans should be running with A/C so seemed like a good time to check the fans can work. The fan motors are both running so no problem with fan motors themselves.
Next would be a look around under the hood for anything obviously a problem such as coolant leak, sensor wiring frayed, disconnected or shorted. Most obvious is the coolant reservoir is not full.

Next, top off the coolant level and position this car on a lift. Because gravity is not a suggestion, it's the law. I'm going to pressure test the system and look for leaks. That will include looking in areas more easily seen from below. On starting the engine to move the car, I looked inside the reservoir for "turmoil". The car is not at engine temperature and has been sitting for hours so the thermostat should be closed. If I see coolant trying to come up out of that reservoir it is a sure sign of a bad head gasket. But it looks calm. 

With the car on the ready to lift when needed I want to put pressure on the system. This a 16 psi system. 

After several minutes the pressure has dropped slightly. I don't see any obvious leaks from the top and no drips underneath. Time to take a look from below after pumping it back to 16 psi. 

Leaking from under a hose, at the clamp end. Actually the clamp is most accessible from above. The clamp was slightly loose. Now the leak is fixed. I don't know about the overheating though so not finished yet. It could have been overheating due to low coolant level but there could be other problems. I have to run the car, let the thermostat open and fill to level. The cooling fans run with A/C and I saw that working but the fans also are supposed to come on when coolant temperature gets around 220F. I've checked oil for coolant contamination and did a combustion gas test at the reservoir. The system is filled, the fans come on at 215F and off at 210F. Good circulation. No signs of overheating. Time for a road test. 

Didn't see any problems from the cooling system on the road test. Customer called and said he was having boil-over and overheat at idle conditions but never at road speeds. I explained what I've found and done so far. He also said he has been working on the car and only added water to the cooling system. I need to test coolant strength after this cools enough. Then I'll give him a call back.


A few things while I'm waiting:
Water as a stand alone coolant is not good. It will cause rust and it has a lower boiling point than a 50/50 mix of water and antifreeze.
Flushing the cooling system is done as maintenance when changing coolant. If your radiator is plugged it will need replaced or removed and cleaned by rodding.
If you put a stopleak product in your cooling system, do it once. Don't keep doing it over and over and follow the directions. Mainly though, don't use a stopleak product.

The engine coolant tests as mostly water.

Drain and fill with a 50/50 mix and this car should be good to go. 

Thanks for reading

Kenneth Hayes

G&G Auto Repair 

2000 Toyota Sienna CE : Intermittent No Crank

Got this Sienna in the shop that sometimes has to be boosted with jumper cables to start, according to the customer. It has been looked at by other shops but no cause found. I noticed the battery is a Toyota brand dated 11/17 so not very old. The car had just been jump started to drive here and I pulled it inside still running. After parking and shutting off I put the charger on and saw a 69% charge. I let the charger stay on until reaching 100% late last night. This morning I did a quick reconnect and saw the battery back to 90% so I'm already suspecting there is a drain.

That hunch turned out to be wrong. 

Back to plan A then. I should have enough charge to do the battery testing. I'll start the engine a few times with lights and accessories on to dissipate any surface charge and get a more accurate test. Also, I like to disconnect the battery terminals and test directly from the posts with no possibility of the car wiring affecting the result. Since I'll be disconnecting the battery I like to put a memory saver in place to help the car retain system memory. 

Well the current draw is in within specs and the battery tests good (but is still a bit less than a full charge). I'll check the alternator output, put the charger back on to get 100% again and monitor this thing for a day or so. Meantime I can check for related bulletins. Hold on. Found a problem. I put the old battery load tester/meter on for a quick check. Battery still tests good with load test but the alternator is barely registering. With no accessory load it is not a good charge and with accessory load it sometimes drops to no charge at all. Retesting with a more accurate tester and yes the alternator is indeed bad. Belt is still original as well. 

Summary: I need an alternator and a drive belt, battery charge and a follow-up system test. 

Thanks for reading

Kenneth Hayes

G&G Auto Repair

A/C Testing

Just a quick walk through and reminder of A/C system testing. The complaint on the 2006 Sonata was the A/C was not cold. Checking the settings, they were fine for testing. You want the system set to coldest temp, a low blower speed and recirc (max) airflow. Put a thermometer in the center dash vent and look for A/C output to at least be 50F. Depending on variables like how long the car has been sitting in the sun, how hot the outside temps are, and such. You might have to drive a bit to reach coldest vent temps but this particular car was getting down to 40F with no driving at all other than from the parking area to inside the shop.

I don't see a problem. That is certainly cold enough. A test drive would be in order to determine if the problem is related to conditions that happen while driving. But while the car is inside it would be a good time to check the actual refrigerant charge. The customer had mentioned adding refrigerant himself. Usually that involves one of the A/C charge kits sold over the counter at so many parts stores and box stores. One of the problems with using the kit is that you have no idea whether the system is properly charged either before or after you use it. Personally, I don't see the point then. But if we didn't spend money on something we'd all be wondering what to do with it right? Anyway.... I connect the recovery/recycle machine and open both gauges. I recovered 1lb 9oz of refrigerant. The system was overcharged by 6oz.   

With a proper charge and a road test the A/C still operated as designed. Could the overcharge have been the problem? Yes. More is not always better. The pressures would not be correct and could cause the compressor to cycle more often than designed. Or, there could be another problem lurking in the shadows that has not shown up while the car was here. I just advise the customer and we see how that goes. 

Thanks for reading 

Kenneth Hayes

G&G Auto Repair

How Does That Work Series: Testing A Battery Drain Problem

A common complaint is a battery going dead after a car has been parked for a day or so. I just happen to have a 2013 GMC Sierra in the shop with that problem. I'll walk through the steps to diagnose this particular truck. First, we want to be sure the battery is fully charged and is a good battery. The electrical system tests aren't reliable unless we are using a good, fully charged battery. The battery had a 90% charge when the truck came in. A slow charge is needed to bring it up to 100% and that can take a few hours. Sometimes I switch batteries to let the original charge while I use a known good battery to move ahead with the tests. This truck was staying overnight so I just left the charger on the truck until it reached 100%.

After the battery is charged, it should be tested. Cranking the truck a few times with lights and accessories on will get rid of the surface charge so the test will be accurate.

Once the battery has been verified good and has a full charge it is time to measure the actual current draw. I like to use an amp clamp along with the meter so that I don't have to disturb a circuit.

You have to let enough time pass that the computers should go to sleep mode. If you start testing before that time you will be chasing a draw that isn't a problem. After waiting an hour I put the clamp on the negative battery cable and I can see excessive current draw. 

At this point the job is to find the circuit the problem is on. I had a small ground wire from the battery post with no draw and the large one with 377mA. The positive post had three cables from the positive post. I found the draw to be on the cable from battery to underhood fuse block. 

Next step is to start removing fuses from the block one at a time and watch the meter. When you see the draw fall to normal you have found the circuit and are a step closer. I like to take a picture of the fuses while all in place in case I have to refer to it when re-installing the fuses. As I remove each fuse I lay them out in a similar pattern but the backup of a picture is a good idea. If you remove a fuse that has no affect on the draw don't put it back in immediately. Just keep removing fuses and laying them aside until you find the one that affects the problem. Putting a fuse back in before you have found the problem can sometimes wake a computer that was in sleep mode and you will have to allow the timeout all over again. 

Eventually I did find the fuse for the problem circuit. I like to put the other fuses back after I know where the problem is and just leave that one out. 

Next up is a wiring diagram. It is sort of a roadmap for tracing further. We want to see what this fuse feeds. This fuse turns out to be the battery feed to the instrument panel fuse block. So I put the fuse back in place. We'll have to wait for the timeout again but I also had to leave the driver door open to access the instrument panel fuse block. The door is open but tripping the latch closed is enough to make the computer believe the door is closed. 

Just like the procedure for the underhood block, we'll pull fuses in the instrument panel block until we find the circuit with the draw. I found the problem circuit and left that fuse out, replacing the rest. 

Then it is back to the wiring diagram. The fuse powers the Vehicle Communication Interface Module. (OnStar Module). 

Having gotten this far I was able to find a related service bulletin. 

You will find different paths on different automobiles but the procedure for tracing battery drain is going to be basically the same. Thanks for reading!

Kenneth Hayes

G&G Auto Repair