6 ignition coils in 9 months- 2000 ob 2.5 mt!!! Please help

[plain OM]

I've attached a diagram of the coil assy which is for my 07, but is for the assy with the integrated ignitor. (It's electrically similar to the one ETC posted.)

I suspect using the analog meter might be somewhat misleading. Between the tower pins 4 and 5, and between pin 6 and 7 on the diagram, you should get around 12 kiloOhms (12,000 Ohms). Also, the spark plug wires should each read in the range of 10 to 20 kiloOhms.

The primary side of the coils themselves inside the integrated-type assy can't be checked.

There are 4 pins on the input connector of the coil assy. One is 12 V, two are signal inputs from the ECM, and the 4th is ground. (I believe the signal inputs from the ECM are on pins 1 and 4 of the 4 pin connector, pin 2 is where the 12 V comes in, and pin 3 is the ground.) With the connector disconnected, the latter three should all read a high resistance to ground because they go through the triggering transistors in the coil assy. (If you still have one or more of the failed coils, these resistance measurements might be helpful. It might show if one of the transistors, or output coils, is shorted.) If you have a digital meter with a diode check test range, measuring between the inputs to each of the coil's transistors and the ground pin might give a lower reading in one direction (polarity) and a higher reading in the other direction.

[SubieDrew]

Hi plain OM,
thanks for the diagram- easy to understand. Looks the same as mine.

Think I need to get a digital multimeter. a new toy I can use it anyway for other electronic projects so it's a good excuse to upgrade my ancient one.

I'll re-check the plug wires.. I think they were at 8kOhms.

So if I'm understanding this diagram correctly then I think the ignitor for cyl 1+2 is what was smoking in the last coil. It was the little part of the coil pack when you look underneath. That side toward cyl 1 was dripping onto the manifold and that side of the pack only at cyl 1 was bulging outward.

With that being where the short probably was does that lead anyone to further pin-point the electrical issue upstream toward and/or including the ECU.. vs a bad plug or wire creating too much resistance?

is the primary coil also know as the ignitor?

I'm in good shape on the tower resistance. I don't have the old coils, but maybe my mechanic still has one I can compare. I suppose I could have him take this used coil and test is on a car in his shop to see if it's working. Once I get the digital multimeter I can check it this weekend.

thanks everyone for your help.

[plain OM]

I don't think we can conclude that it's a problem related to cylinder 1 solely because most of the damage on the coil assembly is closest to that cylinder. We, or at least, I, don't know what's where inside. In many cases, windings of the two transformers (seen at the right in the diagram) can be concentrically wound, so we can't tell, for example, if one or the other is overheating. But if each time the coil failed there was a trouble code (CEL) related to the same cylinder misfire, this could be an indicator.

There are too many variables and not enough hard diagnostic information. The causes of overheating include a fault in the ECM, a short in the switching transistor in the coil, a short/arc-over in the transformer, a high voltage pulse in the transformer, etc.

It would be unusual for 6 coils to fail the same way on their own. It would have to be an external cause. The output side the coil -- spark plug cables and spark plugs, were checked and changed as I understand it, so these are not likely problems. But it is important that the wires and plugs be good, that the wires are fitted on the coil towers and spark plugs properly and securely, and that the metal part of the spark plugs are indeed grounded to the block through the threads.

Also, as noted earlier, one of the four pins on the coil input connector is the ground. This wire should be checked to make sure it is good, that is, there's little or no resistance between the pin on the wiring harness connector and both the engine block and the battery negative post. (A digital meter with a low Ohms range would be helpful here. Many will read as low as 0.01 Ohms on their lowest range.)

Beyond those, it's possible that the ECM is delivering a faulty signal to the switching transistor(s) in the coil assembly. If a transistor is turned on for too long, it will cause the transformer primary to overheat. This can lead to deterioration of the wire insulation, and shorting between the windings. It could also stress the transistor leading it to short, which would connect the transformer primary directly across the 12 V line, and this would lead to heavy overheating.

In the old days of points and distributors, we would use an ignition scope to examine the input signals to the coil transformer for it's shape and duration (dwell). I'm not sure if this is applicable with these fully electronic systems, but it is possible the ECM is sending an improper signal.

Finally, there's 4 wires going to the input connector. Two are the signal wires from the ECM, and one is the 12 V line. Although unlikely, if the wires are frayed, and there is contact between the 12 V line and one of the signal wires, it would turn on the corresponding switching transistor and cause excessive current through the primary of the related transformer. So a close inspection of the wiring, and ensuring that there's no dirt or grease across the connector that could cause such a connection, would be a good idea. (This is what ETC referred to earlier.)

One thing you could try, especially if the coils are failing almost as soon as they are installed. With the car cold (the failures do seem to be related to the car being cold), disconnect the 4 pin input connector at the coil. Turn the car ignition to ON (but do not go to Start). Then, measure the voltage at each of the pins of the connector (to ground). One should have 12-13 Volts coming from the battery. Another is the ground. The other two are the signal wires from the ECM, but as the engine is not turning, I believe there should be little, or no, voltage at these two pins. If either of them does show voltage, the ECM could be faulty. Again, it would be one of those rare faults, but not impossible.

The ignitor is not the primary coil. The ignitor (separate, or integrated, as in the two diagrams) is a circuit that turns the current through the transformers on and off. In the diagrams, the "primary" of each of the two transformers (at the right) is the spring-like spiral on the left of the two vertical lines, and the high voltage secondary is to the right. (If this isn't clear I'll mark up another diagram.)

I realize this isn't providing any clear fixes. But I hope it might move things along to a solution.

[SubieDrew]

Hi plain OM,
Wow great info thanks a TON! Once I get to the car with a digital meter I'll check all that. I'll look for a meter with a low Ohms range. I'm glad you spelled it out so clearly. I was wondering what kind of reading I should expect with ignition ON but not cranking. With ignition off what would you expect the readings to be?

I'm also going to double check the plugs and wires.

My guess on cyl 1 was based on multiple CEL codes P0301 and P0302.

Also, I have noticed mice evidence in the engine, so maybe at some time the little buggers did some classic wire modifications.. as you and ETC said could be some investigation needed. I would love it to be that "simple". maybe the cold issues are also related to moisture/condensation on a faulty/frayed wire?

Thanks for the clarification on the primary... but the ignitor is integrated in these coilpacks right? not like i can check them separately.

I was wondering if the ECU got damaged from a high resistance situation from a bad plug or wire on the first few coil failures. When my mechanic changed the plugs finally in Dec 2010 he said the platinum's in there didn't have the correct part number... he didn't like that. Weird cause I think he put them in at the last major tuneup like 30K miles ago.

On AA1CAR.com there was mention of possible damage to the ECU if a short lowers the normal resistance in the primary windings, OR an open or higher than normal resistance in a coils secondary windings could allow feedback induction through the primary circuit. Not sure if any of that would apply here. I'm hoping the dealer did actually get their diagnosis right of the ECU being ok.

No worries on no clear solution, just glad to get such helpful guidance. Thanks again.

[plain OM]

I'll look for a meter with a low Ohms range.

Many digital meters will have a resistance range that is 100 or 200 Ohms maximum. If the readout of the meter is 4 characters, the 200 Ohm range would read as low as 0.1 Ohms, which for automotive purposes is fully adequate. Meters with resolution down to 0.01 Ohms will be more expensive -- they might have a lower maximum range (e.g. 10 Ohms) or use more digits in the readout.

I was wondering what kind of reading I should expect with ignition ON but not cranking. With ignition off what would you expect the readings to be?

I don't have any specific information on this. But to begin with, the 12 V and ground lines are obvious. My understanding of how the ECM interfaces with the igniter circuit in the coil assembly suggests that the two signal lines should be at zero or a low voltage (perhaps less that 0.5 Volts) when the engine is not running or cranking (which of course is this case if the connector is removed and the key is only at ON, and not turned to Start.) If there is a positive voltage at either signal line, it would turn on the transistor switch in the igniter circuit, causing current to flow continuously through the primary of the related transformer -- not good. With the ignition off, there should be no voltage on any of the pins of the 4-pin input connector. (Incidentally, be very careful when measuring the voltages at the connector. You don't want to inadvertently short one pin to another.)

Mice -- very possible. Quite a few posts here about the damage they do.

I was wondering if the ECU got damaged from a high resistance situation from a bad plug or wire on the first few coil failures. When my mechanic changed the plugs finally in Dec 2010 he said the platinum's in there didn't have the correct part number...

In this type of circuit, if the plug doesn't fire, a high voltage spike can be impressed back on the primary of the related transformer. This would then impact on the related switching transistor of the igniter circuit, which could cause the transistor to short out. That's why it's important that the wires to the plugs be properly attached, and that the ground at the plug be good. I don't think the ECM would be damaged -- there should protective circuits for that -- I can't discount the possibility.

My guess on cyl 1 was based on multiple CEL codes P0301 and P0302.

If these codes showed up again each time a coil failed, then it does suggest that the problem is related to cylinders 1 and 2. However, both spark plugs are fired from the same transformer and igniter circuit so it could be one (or both) of the spark plug wires or the spark plug itself, or it could be related to the ECM signal to that particular igniter circuit.

[SubieDrew]

plain OM
thanks again for the advice.
I'm looking at this multimeter for $20 us should be pretty good and looks like enough accuracy:
RadioShack® 15-Range Digital Multimeter - RadioShack.com

would it make sense to get one or both of these spark testers below? or is a spark plug with a jumper cable grounding it ok? Or are these too dangerous with a distributor less ignition system? I don't want to get zapped or blow up the car, but the later would take care of the situation I suppose

seems like this would be helpful to see if there is an issue with any of the wires:
Advance Auto Parts: IGNITION TESTER AMPRO by Performance Tools - Part W86553/AC805

This one looks like you can adjust the gap:
Advance Auto Parts: HI ENERGY IGNITION TESTR by Performance Tools - Part W84600/AC813

Once I get some more checking done, I'll let you know what I've found out. thanks again for the GREAT diagrams and the added notes in red

[plain OM]

I'm looking at this multimeter for $20 us should be pretty good and looks like enough accuracy:

Looks okay as far as ranges and accuracy; however, there's a note at the bottom saying "you'll also need 12 V battery". I'm not sure, but this suggests to me that it uses a 12 V battery, which isn't your everyday run-of-the-mill type. This could be costly, relative to a more expensive meter that uses a 9V battery, a type C or D battery, or AA batteries, all of which are readily available and relatively inexpensive.

I would add that some of the smaller, almost pocket-size meters also have relatively short leads, which can be a challenge when trying to measure at a connector with one lead and at the same time hold the other lead to a good ground. On the positive side, most meters use "banana plug" connectors to attach the cables, and it's easy to substitute different leads, or probes with clips on the end to hold to wires and ground points on their own.

I don't think the plug testers are necessary; if there's any doubt about the plugs I'd just replace, and this has apparently already been done.

I finally found the input/output table for the ECM on my 07, and the voltage measurements at the 4 pin connector at the coil. The two wires that carry the ECM signal to trigger the igniters each should have zero Volts with the key at ON but the engine off, and 1 - 3.5 Volts with the engine idling (obviously this is with the connector attached to the coil). The 1 - 3.5 V that the meter would be reading is actually not D.C. -- it's a pulsed signal, with a max of about 10 V and min of near zero. But the meter will read it as an average voltage, varying with the speed and other parameters of the engine control. Your 2000 might be the same.

(When measuring at the coil connector, it might be an idea to wrap the longer, un-insulated section of the test probe tip with tape, and just leave the tip itself exposed. This will help prevent the possibility of causing shorts between the pins on the connector.)

Edit: Diag below is from 2000 wiring diagram; interestingly, the pin #s and their functions at the 4-pin coil connector are the same on my 07.

[SubieDrew]

good catch on the battery for the meter. I'll double check and see what's up with that. Great idea on the tape on the leads.

Another great diagram. Your explanations help immensely.

I was just sorting through the pdf FSM I found online and trying to sift through all the diagrams. The info I found says 2000+ and maybe this diagram is a 2002, but I would think 2002 would be the same as 2000. mine is a MT but I wouldn't think that would change the coil wiring.

Either way the wiring is different than the one you have illustrated... see attachment. Is the diagram you have for a 2000 or 2007? I'll double check the wire colors when I get to the car.

If there is a different 4pin coil that looks identical that might explain a few of the coil issues. But from what I can find online the gen 2 coils are the same, but the gen3 like yours are different visually.

I just found the I/O table for my 2000 it says zero volts in the on position and 1- 3.4V idling (waveform) (basically the same as yours). how would I test idiling, using the DC v meter? where would I put each +/- probe while everything is connected?

Thanks again.

[plain OM]

Either way the wiring is different than the one you have illustrated... see attachment. Is the diagram you have for a 2000 or 2007? I'll double check the wire colors when I get to the car.

Actually, they're both the same as far as electrical connections is concerned. I've attached the complete page from the relevant 2000 wiring diagram (MT). In post #17, I had attached only the bottom part.) Comparing this diagram and the one you attached, we see that the wires coming from each of the pins of coil go to the same places.

Pin 1 goes to a25 on the ECM
pin 2 goes off the page, but when traced, it goes to a 12 V line
pin 3 goes to ground
pin 4 goes to a26 on the ECM.

The only difference I can see is that in my diagram pin 2 has a Yellow wire, whereas in yours pin 2 has a Yellow with stripe wire.

Note that at the bottom of both pages, the diagram show connector E12 as flat with the pin numbers going from 1 to 4. However, in the wiring diagram itself, the order of the pins across the coil connector is not from 1 to 4. Yours shows 3,1,4,2, and mine has 2,1,3,4. This is only for drawing convenience -- to reduce the number of times the lines coming from the connector, which represent wires, cross each other. The pin numbers still run 1,2,3,4, across the connector. Consequently, the third wire from one end (and second from the other end) should be black (pin 3, ground).

I just found the I/O table for my 2000 it says zero volts in the on position and 1- 3.4V idling (waveform) (basically the same as yours). how would I test idiling, using the DC v meter? where would I put each +/- probe while everything is connected?

It might not be possible to measure with the coil connected and the engine running. On my 07, the connector is off to the side and there isn't enough room to get a test probe into that area. But even if there were, it looks as if the back of the connector, where the wires go in, is sealed, so a test probe can't be pushed in to make contact with the pins inside. That's why what I originally suggested was intended to check for 1) the 12 V supply at pin 2, the ground at pin 3, and little or no voltage at pins 1 and 4.

If your connector is physically different, and would allow you to measure what's at each pin with the engine running, then that's fine too.

When measuring, your negative lead should be connected to a good ground. The positive test probe tip is then connected (touched) to each of the connector pins and the voltage read -- but again, this assumes the connector is removed from the coil itself and the pins on the harness connector are then easy to access.

I'm not sure if these measurements will reveal anything, but if pin 1 and/or pin 4 has a significant voltage (i.e. not zero, and also if they are different, such as 4 being 0.2, and 1 being 2.3) with the ignition On but the engine not running (it can't be if the connector is out), there is clearly a problem.

Beyond this, and if there are no signs of physical damage to the wiring (e.g. miceworks), then it might be a job for the dealer, or someone with additional test equipment that can analyze the ignition system and see if anything is amiss in the cylinder 1 and 2 side.

But let's first see where this goes.

[SubieDrew]

Thanks again for the clarification. Since I don't have the car easily accessible, it hard to just run out and do a visual check.

As it turns out, I didn't get to where my car is garaged this past weekend, so no progress was made. I'll do as you said, check it without the connector attached, cause other than that- there won't be much that I can tell with just a multimeter.

Is there anything specifically that I could use a code scanner to check if I get it running? Like info on the air/gas mixture on each cylinder? Maybe that might further explain the misfire codes and the P0420. I suspect that at this age the cat is no good but more curious what the cause was upstream for the cat getting killed (if it is).

Thanks again plain OM. All your help is REALLY appreciated.