I tested the Plug wires, and found one that was suspect... I replaced em all.
Battery voltage was 12+ volts at the connection to the ICM. I didn't want to lay under the car and test the voltage while it was being turned over....
I did find (I think) a connector by the injectors that had the same black and pink wire that is hooked to the ICM. I noticed the voltage drop to 8 volts while cranking. I tested one injector again (with the used ICM installed) it flashed my test light.
Does the volatge drop mean I have a bad ground when turrning it over? Or is that normal? Perhaps the ground to the ICM is broken while turning the ignition.
I guess I may have to crawl under the car and do the test at the actual ICM connection (black and pink wire).
When you (Comboverfish) talked about isolating ground, did you mean running a ground wire from the battery to the black wire? I'm having a bit of trouble figureing out which wire in the harness (black & pink) actually runs to the ICM
At least the weather here in Minnesota has been warm!
Like I said, get piercing leads and attach them first, then run extension leads out from under the car and then test -- if you feel the need to get out from underneath.
You'll save yourself time with a wiring diagram. Go to the library and make copies from the GM factory wiring schematic. If the GM book isn't available, the Mitchell or Motor publications are adequate. If your library is particularly progressive, then they may have service info on CDROM or DVD from one of the above companies.
Voltage drop means the difference in voltage in one circuit when compared to actual system voltage. IE, if you do my first suggested measurment across the two ICM wires, there should be nearly the same voltage there as if you were to measure across the battery terminals. Taking it a little further, let's say you see 8.2 volts there while cranking, and then you test the battery terminals while cranking and see 8.5 volts. That isn't a significant amount of voltage drop. (The voltage in that instance is just plain too low to start a car, so you would want to charge the battery and inspect the battery terminals for tightness and corrosion). Example 2: You read 8.2 volts cranking at the ICM and 10.5 volts at the battery. NOW you have either a power or ground feed issue in the ICM wiring. Example 3: 10.4 volts at the ICM cranking, and 10.5 volts at the battery. No problem with the wiring or battery in this instance.
The above is a quick test to see if you are barking up the wrong tree regarding the ICM. If there is a significant voltage drop, then reread what I said about checking the ground side voltage drop with a DVOM. I think I explained it as clearly as I can. Otherwise I can only suggest you get someone to help you that is already familiar with DC voltage diagnosis. If you don't know why you are doing the tests or what to expect, you may misinterpret a poorly executed test (probe not making contact, DVOM set to wrong scale, etc) and run around in circles indefinately.
Sounds like what I was saying. See above regarding piercing leads and extensions.
See above regarding acquiring a quality, accurate wiring diagram. You
*could* tie in a ground wire and *see* if that fixes the problem, but that's not what I said. That method will lead you down the path of butchering your car for a lack of understanding.
The simple fact is that, despite the difficult location of this set of suspected components, a decent mechanic would have the basic cause of no spark located in under 30 minutes real time. A diagnosis would probably cost about one hour of flat rate. Furthermore, if the mechanic installed a part that he suggested as bad and there *still* was no spark, he would be confident enough in his (proper) diagnosis to exchange the part for another *working* part. It's not uncommon to get bad electrical parts brand new out of the box. I can only imagine that that percentage of failed parts is much higher when buying from the junkyard.
First off Thank you for all the help given. I really do appreciate it. Your explanations have been very clear, I just don't know the terminology.
I figured out what pierceing leads were, but didn't pick up on attaching them to an extension wire. I thought you meant I could test the circut at the wireing harness.
I'm virtually positive I'm in the right circut. I do have a wireing diagram as well. I didn't know that to test the voltage drop I had to compare battery voltage drop as well.
The motor turns over just fine, so I didn't suspect the battery or terminals. I was pretty sure I understood about voltage drop. I just didn't know if it pointed to wireing or the components. I think I'll find example 2 will explain my situation. Then I test which wire is suspect using the prevously mentioned method.
Yup I understand a "real mechanic" could find this problem rather quickly. I probalby should have had the car towed in a week ago. Some of the reason is money, some is pride, and more is pure stubbornness. The last time I took a car in for a suspected wireing problem, it was three hours of shop time and no solution.
I tried the volatage test again. I found the battery was bad. Replaced the battery, and still No Spark.
I tested the leads to the ICM and (with the new battery) the voltage droped to 10.8 volts while cranking. The battery dropped to 10.8 volts while crankning as well. If I understand correctly, this means the voltage to the ICM is correct.
No corosion on the ICM connectors. I suspose that it's possible I have two bad ICM's?
would there be a test I could do with one or more of the coils removed? To make sure the ICM is actually getting voltage?
I would guess since no spark on any plugs, or at the coils (tested with a spark tester). That no voltage is getting to the coils.
Yes, that shows you that all available voltage is making it to the point where you were testing while under a load (the task of attempting to fire the coils). And 10.8 volts is fine as a system voltage during cranking.
It sure is.
As I remember, the coils have terminals that mate with the ICM when you secure them into place. If you took off a coil, you could connect a test light across the two exposed terminals on the ICM, then crank the engine and look for the light to flash rapidly. Don't confuse the light 'dimming' with flashing. If it's staying on all of the time and dimming with every compression stroke of the engine, then that indicates a bad ICM. Actual flashing of the test light indicates a good ICM (for the most part).
What needs to happen is a *break* in voltage at the coil negative terminal for just the right time duration to create a spark.
You can test for B+ supply to the coils. One of the two terminals for each coil should have B+. The other terminal for each coil is the negative feed that is triggered off by the ICM when spark is commanded. That is how a functioning ICM is capable of flashing a test light.
I Tested for B+ at the coil terminal. I had B+ volatage. However turning the car over did not flash the test light. The opposite terminal if I understand correctly is an intermitant ground, used to fire the coil.
I also tried connecting the test light to battery + to see if the ground terminal ever worked. This didn't flash the light either.
So, i know that the ICM is getting volatge. But do I know the ICM is bad if there is No ground (on the other termial)? I think you were saying that the brain box is only used to time the spark. The coil should flash without the brain box connected. Wouldn't the computer need to tell the ICM when to create Ground which would fire the coil? With only two leads into the coil, how does the ICM tell one side or the other of the coil to fire?
If I explained it again, I would be repeating myself. It sounds like you have a bad ICM again, assuming all of your test results are valid.
I think if you reread the last half dozen or so posts it will become clearer. I don't know how else to explain the system any clearer. If I'm missing something, then I do apologize.
Sorry, I have read all the posts several times. I didn't really need to know how two leads into the one coil could fire two seperate plugs from the same coil, I was just curious.
You've been more than patient with me and taught me a great many things about diagnosis.
If the brain box is truly not needed to generate a spark except for timing. Then the Crank Sensor must tell the ICM when to fire.
I will get a new used ICM over lunch, and try it this evening.
Each coil fires both of it's wires at the same time. Each of the two coils therefore fire twice as often as needed, but it is really not a significant waste of energy. The cylinder that doesn't need spark is not under compression, and draws very little current from the coil.
Not a problem. I just can't say anything new at this point. To go further would be for me to physically test the system in person and see if there was anything either of us were overlooking.
Well, you definately wouldn't want to drive around on ICM timing alone! It is only used to get the engine running. The ECM sees that the engine is spinning faster than ~ 400 RPMs, *then* it tells the ICM that it must use EST timing info from the ECM to time the coil outputs. If you had said that the car *almost* starts every time you crank it, but doesn't remain running, we might have needed to investigate the bypass and EST wires. Since it isn't producing *any* spark during cranking, I was focusing on that issue.
Then the Crank Sensor must tell the ICM when to fire.
Yes. Look on your wiring diagram. The crank sensor is wired directly to the ICM. The ICM then sends a square wave reference signal to the ECM so that it knows engine speed also.
It's raining tonight so I didn't get much time to play. I did hook up the new used ICM with the same result, Didn't start. I havn't actually tested if there is spark yet. I could have wet plugs, but I suspect it would have aat least fired.
One thing you said earlier is puzzeling...
Doesn't a break in voltage mean that the ground goes to open to cause the spark? When I tested ground terminal of the previous ICM it was open (it wouldn't light my test light with the test light on B+) Does the ICM need to be grounded through the motor? I wouldn't think so since both B+ and a ground are supplied.
That sucks. I tend to think at this point that I'm missing something. I will think more about it because this has me curious.
The ICM sends ground to the coil until it needs it to create spark -- at that point it cuts ground momentarily and the coil induces that stored energy from it's primary winding into it's secondary winding which results in a spark at the sparkplug.
It's possible that the coil negative terminals are supposed to be open until the point where the engine is being cranked / run. I would have to 'reverse engineer' the system with an oscilloscope to figure out details like that. I don't have any factory GM literature on the basic operational strategy of the system.
No, but the ICM case may be grounded for one reason or another anyway. The supply ground is where the internal electronics get their ground path, if you look at the schematical sketches drawn inside the ICM in the wiring diagram. (at least on my diagram - it's a Mitchell)
I emailed a regular R.A.T. poster who is extremely knowledgable about domestics and has GM dealer experience. He told me that, despite my schematic which clearly shows the Black/White wire supplying ground to every function inside the ICM, he is certain that they get ground through the ICM case to the block. IOW, he is sure that my wiring diagram is inaccurate on that count.
You weren't testing for spark with the ICM unbolted or otherwise off the engine, were you?
He told me he has fixed a couple no starts by sanding the rust and corrosion off of the engine block and the ICM mating surfaces on GM DIS systems just like yours.
I would try that suggestion before going any further.
I crawled under it to remove the ICM and clean up the contacts, only to discover I had installed the original ICM instead of the new used one.
It was dark, and slight rain, but I wanted to at least try the new used coil. I grabbed the wrong one. That's my excuse. I did sand the contact areas on the ICM and block before I tried again today.
I have learned many things through this ordeal.
3 month old plug wires aren't necessarily good. The same goes for a 3 month old battery. I now know how to test if a battery has enough output even if if turns over the motor easily. Never assume you know what's wrong when you are diagnosing (I should have found the bad plug wire with one of the first tests). Don't replace parts without finding out they are bad first...I knew this one but the ASSuming deal convinced me I was right.
To recap: 1bad plug wire; 1 bad battery; two bad ICM's
I am indebted to you comboverfish. Thank you for your patience, and shareing your knowledge. I feel like I just completed a school mini-couse.
Glad to hear it! In all fairness to the battery, it may have just needed a complete and proper charge. Most 3 month old batteries, regardless of how cheap, are still in working order.
MotorsForum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.