2.3L ('83-'97) Sudden loss of spark, but have coil power and signal

[lowspeedpursuit]

'94 2.3L. I like to think I'm not a complete idiot, but I am having an absolutely baffling problem over here, and I'm at my wit's end.

Truck ran and drove yesterday morning, zero problems. Parked, then in the afternoon suddenly I've got strong crank, no start. Fuses and relays are all good. Double-checked the NSS/clutch switches (couldn't remember if those cause no-crank or not), and they're good. EDIT: Failure here would cause a no-crank.

Check spark. No spark anywhere. Swap coils, no change. Swap in a spare ICM, no change. Check CKP, I'm geting both signals. I have power at the ICM and the coils. I'm getting signal all the way to the coils, but I'm still not getting any goddamned spark.

How is that even possible? Did both sides of both my coils fail at the same time? I'm hoping I'm secretly an idiot, and I'm missing something super obvious.

Thanks.

 

[RonD]

May only have spark on the exhaust side spark plugs in 1989-1994 2.3l so only test on that side

Sudden no start on 2.3l SOHC can be broken timing belt, just have a look behind cover and see if its there and turning with engine

50/50 test is the best way to determine reason for No Start if engine cranks(starter works)
Spray fuel into the engine, gasoline or Quick Start(ether)
Try to start
If it starts runs then dies you have a fuel delivery issue
If it doesn't fire/start then its a spark issue(or compression issue in the case of the 2.3l)

Spark testers don't work well on Waste spark systems, and fuel pressure doesn't mean fuel delivery in the cylinders
50/50 test was first used in 1890's on the first gasoline engines, lol, and is STILL the best first test to do when you have a No Start


When Ford went to dual spark plugs in the 2.3l in 1989 they thought activating BOTH spark plugs in each cylinder for start up may cause a weaker spark on both
So the ICM "inhibites"(turns off) spark pulse to the intake side coil pack when starter motor is activated, well actually below 400rpms on crank sensor, engine cranks at 200rpm
So only test for spark on exhaust side

In 1995 and up ICM was dropped and so was inhibitor

 

[lowspeedpursuit]

Good afternoon; thanks for the advice.

Timing belt is intact. No change manually putting a bit of gas in the engine.

I'm only testing off the rear/exhaust coil, and I'm definitely not seeing any spark. I don't know which contacts are paired, but I've got no spark from any of the four, trying both coils connected to the rear/exhaust plug.


I'm also reading now that the CEL is meant to turn off while cranking, and that staying on indicates a bad CKP. Mine is staying on, even though I'm getting output from the CKP doing this test. Is that test valid?

I see people talking about looking for 1VAC directly from the CKP, but I'm not clear on which pair of wires to probe. Seeing a lot of "google it", then the top results are the forum threads I'm already reading.

 

[RonD]

Yes, CEL going off while cranking means the Computer is getting a good timing pulse

1989-1994 2.3l uses a Hall effect crank sensor, 3 or 4 wire, with a 12volt on and off pulse as the signal

Most Ranger crank sensors were VR(variable reluctance), 2 wire, generates its own voltage, AC voltage, would show 0.5v to 1v AC at 200rpm, cranking

If CEL stays on then either crank sensor is not outputing a good on/off signal or ICM is not sending it to computer
Computer just does the fuel mix, ICM and coils do spark on their own, no computer needed
So if 50/50 test was a no fire/start then ICM doesn't have good crank sensor signal

This page tells you have to test the hall effect crank sensor on a 1989-1994 2.3l SOHC: https://easyautodiagnostics.com/ford/2.3L/icm-and-crank-sensor-tests-4#12

 

[lowspeedpursuit]

Yeah, I'm on easyautodiagnostics now. For the CKP, I'm probing ICM pins 4 and 2 and seeing back-and-forth 0-12VDC on manual cranking, as expected. Does the other test looking for 1VAC somewhere not apply to the '94 2.3L at all?

Following allong on easyautodiagnostics, I'm also seeing signal while cranking on both pins 1 and 3 of the rear/exhaust coil connector, as expected.

So basically, I'm confirming CKP signal output, but my CEL also stays on while cranking.

And I'm also confirming signal input at the coil, but I'm not getting spark, even when swapping the coils.

Makes no sense, right?

 

[RonD]

1vAC would only apply to CAM Sensor testing in 1994, if there is one, not spark related as far as startup spark, it was a VR sensor

And rear axle ABS sensor in 1994
And VSS speed sensor on speedo cable hookup, also a VR sensor

 

[lowspeedpursuit]

Sweet, thanks.

So, then, as long as I'm not misinterpreting the test results, the only reasonable explanation is total failure of both coils, right?

That doesn't sit well with me. Do both physical coils in one pack stop working together? And I guess maybe my secondary failed a while ago and I've just been driving on the primary, but that feels like a lot of assumptions.
We've also got the fact that these coils are only ~2.5yo, and that it doesn't make sense for the CEL to stay on while cranking but the CKP and ICM signals are present.

In any event, I can't find a single spare coil even though I should have a stack of them, and Autozone thinks they're worth $76 a pop, which is insanity, so I'm gonna have to wait for shipping.

If coils don't sort it, I don't really know what else to do but to parts-cannon another ICM, which also feels stupid.

 

[RonD]

No, not yet
Unplug 3 wire connector on exhaust side coil pack
Put test light on center wire(12v) and either outside wire(ground)
Key on
Should see the light come on
Crank engine
Light should flicker, as ICM pulses ground wire to discharge coil
If so Crank sensor, ICM and wires to coil pack are OK, replace coil pack

 

[lowspeedpursuit]

Okay. No change with new coil. Just fired the truck with a "new" (plastic wrap came with the perforation torn off, and I've read that they're supposed to come with heatsink paste, which wasn't in there) Motorcraft ICM. Old coils. From this I can draw three conclusions:

1. My "backup" ICM was no good. It was untested off a parts truck, so that's fair.
2. Standard Motor Products LX230 ICM can blow me. Lasted just a hair under 2.5 years before complete failure, leaving my truck stranded.

3. Most importantly, apparently seeing signal at the coil plug doesn't necessarily rule out a bad ICM. I'm not sure if I wasn't getting enough voltage, or the pattern was off (I feel like that would still make some spark, right?), or what, but I confirmed the ICM was pulsing pins 1 and 3 and yet still had no spark. Not entirely sure what to make of that, but that's what happened.

RonD, thank you for your time, and for walking me through the steps.

 

[corerftech]

From the electrical side, IIRC the ICM is closing a switch to short the low side of the coil to ground (or opening depending on where it is in firing cycle). That FET or IGBT device that is doing the switching in the ICM has basically zero resistance when closed and infinite when open, with a good device. A voltage divider is formed by the IGBT device and the coil primary. Primary is quite low resistance so when IGBT is conducting there is like a .05 ohm resistance at most. 1/20th of your Batt supply voltage is dropped and the balance is seen at coil primary. If the coil load is removed from the ICM, then the divider is gone and now a “failed” IGBT with only a DVM or LED test probe (1 to 100 Meg Ohm rather than the coils 1 ohm), attached will likely behave normally visually. When the failed device has a full 1 ohm load (coil primary)) on it, it’s output impedance (as failed) will cause a massive voltage drop and the coil primary will not fire/saturate.

Basically ICM test open loop might not indicate output failure (pulses still visualized while probing) but closed loopfails due to heavy loading of the switch (IGBT).

Not sure of the conditions you tested in. Conjecture only, but quite accurate.
It would be like replacing a light switch with a heater element. The heater eats the drive voltage and the light bulb is still in circuit but doesn’t light. A bad IGBT is just that, a heater (unless it fries fully open).


If you measured the voltage at the bulb while build removed, you’d never know the heater was there. Measuring the voltage with bulb installed, you will see the effect of the heater.

Did you back probe the coil primary wires when testing with coil connected or just probe the open contacts?

I know you have it fixed, I’m just curious. I have a standard LX230 as well, a year old this last December.

You have raised longevity concerns.
A motorcraft ICM is quite expensive!!

 

[lowspeedpursuit]

IIRC I backprobed the ICM, but probed disconnected contacts at the coilpack because the connector there doesn't lend itself to backprobing, and I don't have a purpose-built piercing tool.

Now that I reread it more carefully, the directions for the test on easyautodiagnostics say this towards the end: "IMPORTANT: Both the coil pack and the ignition module must be connected to their connectors for this test."

So that seems like a case of user error on my part running the test, although since RonD also mentioned running the test unplugged in post 8, I think it might be a common mistake. Or maybe whether the test works with the connector unplugged varies depending on exactly how the ICM fails?

Regardless, thank you for your insight as to why the ICM might fail that way!
Interestingly on price, Motorcraft DY959 is actually ~$20-25 cheaper than Standard LX230 on RockAuto and Amazon both right now. I think historically it's been around ~$10-15 more expensive.