1989 Jeep YJ Wrangler 2.5L Coil Voltage Issues

[Tom in NH]

Greetings,

Have a 1989 YJ with the 2.5l TBI.  Have had a sporadic idle, stall problem.  Verified fuel delivery and injector performance, tested and/or replaced all sensors and IAC motor.  Cleaned throttle body, checked for vacuum leaks.  Kept looking over at the coil/icm assembly, but thought "no, just replaced those not too long ago."  Long story short, tested the coil - badly out of spec.  Changed out both coil and icm and finally (slow learner here) checked voltage entering coil.  Constant 14 =/- volts at startup and run.  Checked with my daughter who drives the vehicle and she confirmed after pondering that yes, seems to only happen after driving for quite a while - giving the icm/coil plenty of time to get nice and toasty.  So, I'm guessing that means ecm troubles, as I know this problem has popped up from time to time but am wondering if it is possibly an ignition switch issue?  So, I'm checking with the experts before I choose which pain in the butt part to remove first.

Thank you.

[Moses Ludel]

Hi, Tom...I would run resistance and voltage tests on the circuits you describe.  Ideally, it's best to have the system warm first.  Regarding the ignition switch circuit, we went through a similar test sequence in this recent exchange:

https://forums.4wdmechanix.com/topic/1180-88-yj-total-power-loss-when-you-turn-the-key/

In general, items that tend to fade when hot would include the ECU, the coil and modules.  Often overlooked on the YJ with its 12VDC system is grounds.  The ground lead from body to engine near the dipstick is a notorious high resistance point.  Check the resistance on the ground circuits and pay close attention to this ground in particular.  Oxidation is the usual issue.

You can measure voltage and resistance from the battery all the way through the entire system as an overview.  (Do not create a spark near the battery while testing.  A defective battery produces explosive hydrogen gases.)  See whether the alternator is performing properly.  If necessary, have the battery tested under load.  AutoZone and O'Reilly's are more than happy to test your battery and alternator functions, a chance to sell products.

A place to start...

Moses

[Tom in NH]

Hi Moses,

Thank you for your response and thoughts.  Some more info and answers to some of your questions. I am reasonably sure the grounds are good.  Will check in the am, but went through them after reading another thread on your site when this problem first cropped up.  I paid particular attention to the ground by the dipstick. I also added  grounds a couple years ago and checked and cleaned those. I have had both battery and alternator tested at autozone and they check out good.

I am on my 3rd coil in the last 18 months (warranteed, but still ...).  This last one is what got me to searching for the cause and today found the constant 14 volts.  My understanding is that line voltage (12-14+ volts) passes to the coil at crank, but the ecu is then supposed to step the voltage down to less than 9 volts at run.  Is that correct?  If so, then my constant 14 volts at run means some defective circuitry on the ecu and my problem of frying coils? 

Could the ignition switch play any part in this problem?  It appears to function properly.  No problem at crank jeep starts well (and runs well until coil degrades over time and then stalls/misbehaves when hot).  All accessories run and work properly, tach functions properly.  I'm inclined to look to the ecu, but a new ignition switch is much cheaper than a reman ecu.  I will check more voltage and resistance in the morn.

Thank you.

[Moses Ludel]

Tom...Traditionally, a resistor wire or ballast resistor stepped down coil voltage for 12V coils.  Conventional ignition coils do operate to the 9V range after starting to last much longer.  However, I see no resistor wire or ballast resistor in the ignition coil wiring circuit for your Wrangler.  If there is step-down voltage, it would take place in the Ignition Control Module, which receives 12 volts plus (the line voltage you describe as 14V with the engine running) from the ignition switch to connection A (YL colored lead) at the ICM. 

With the engine off and the ignition switch in the ON position, read this input and output voltage at ICM connection A and compare the A pin voltage to the coil positive lead voltage.  You should be reading near the static battery voltage.

I would also check available voltage from Pin 27 (OR color) at the ECU or at the ICM Pin 27 connector (may be easier to access).  The #27 pin from the ECU is the input to the ICM for the firing signal.  This is a square-wave 5-volt signal for the firing pulse.  You will only get this reading with the engine running, which will require a jumper wire to your meter.

I'm not clear whether you have been replacing coils or the entire ICM and coil assembly.  Test the Ignition Control Module.  Not surprisingly, the ICM ground is actually the ground near the dipstick.  Use an ohmmeter to verify adequate ground resistance and check the negative voltage (DC system) at the ground connection with the engine running. 

If the ICM is not at fault, consider any sensor that would affect the system when hot, including the coolant sensor, the IAT or the crankshaft position sensor (CPS) at the top/back of the engine.  I have addressed both the IAT (recently) and the CPS in several 2.5L forum exchanges.  See these other forum topics.

An oily or fading CPS can fault when warm.  Also consider the oxygen sensor.  If the O2 sensor needs replacing, make certain you use an OE direct replacement, either Mopar or the supplier/manufacturer's own product (NTK, Denso, etc.) using the part number that crosses over from Mopar .  I will not use aftermarket Brand-X oxygen sensors;  they fit a variety of engines and do not align with the needs of a specific engine.  Any of these sensors can cause a rough idle when warm.

Don't waste money by replacing sensors arbitrarily.  Each sensor has resistance tests, hot and cold.  If they perform well when hot, they are well.  Test specifications are available in the factory FSM.  I've provided some of these specs as PDF copies in other forum exchanges.

I found the wiring diagram for the Model 80 (YJ Wrangler) ICM at page 8W-106 of the 1989 manual.  Note that there is no "step down" voltage to the ICM.  (The alternator uses a resistance wire for protection, but it does not limit system voltage.)  The Pin 27 from the ECU is a square wave 5-volt signal that triggers the coil to fire.  (Some mistake this for a ballast within the ECU:  It is a trigger and does not adjust the primary or source voltage.)  Pin 27 is not a steady voltage source, it pulses in a square wave form.  I have also included the plug connectors to the ICM so you can determine which poles to check for voltage:

YJ Wrangler 2.5L ICM Wiring.pdf

When writing my books and articles, I always rely on first generation factory source information from FSMs.  You'll find this helpful, too.  If you do not have a manual, the FSM covering 2.5L TBI YJ Wranglers between 1987 and 1990 would be valuable.  I have the Mopar two-book set from 1989 (Engine/Chassis Body & Electrical).  Winging this work is pointless.  Check out eBay and used automotive book stores for a FSM set.  

Moses

[Tom in NH]

Hi Moses,

Wow.  Thank you for the thought and detailed comments.  A lot here.  A few comments and then I will test what I haven't yet tested and report back .

With key on, engine not running, I get 12 + volts (basically static battery) at connection A (yellow lead) and at the + lead on coil.  With engine running, I get the 14+ volts at both.

I was aware of the 5 volt sine wave function (though didn't really understand it's operation until your explanation - electrical is not my strong suit), but I have not tested it running.  I got zero volts at key position run with jeep not running earlier - now I know why.  I will back probe it and test and report back.

I have replaced 3 coils and 2 icms.  The last 2 replacements have been both together as a unit.  I replaced the original o2 sensor with an off brand a while ago, but in reading other threads of yours, realized it probably wasn't a good fit.  Have since replaced it with an NTK, which I understand is appropriate.  Other sensors have tested good (TPS, MAT and MAP).  tested EGR solenoid and it tests good.  Replaced the EGR valve as it was original andrusted, filled with carbonized crud, etc.  All of these things have resulted in Jeep running quite well.  However, it just stumbled yesterday for the first time in a couple weeks, which tells me the issue is starting up again, which makes sense.

It is interesting what you say about the Idle Air Control Motor and the CPS.  I check the CPS periodically and it is almost always oily.  (valve cover has a crack at one back corner bolt hole).  I clean it and replace.  Didn't realize that could cause issues when hot.  Very interesting.  Any fix (besides trying to keep it from getting oily)?

I replaced the IAC motor and regret it.  Should have cleaned and rebuilt the original- which I may still do and reinstall. The newly installed IAC seems ok, just not sure it is bang on.

I guess I have been assuming that there was a failure on the ecu relating to pin 27 or pin 3.  I haven't found an option for repair of the ecu and thought a ceramic resistor installed on the yellow lead (connection a), while not part of the original set up would at least reduce voltage to coil to keep things functioning and not frying coils/icms. I guess a reading of the orange lead from pin 27 to the ICM while running will help clarify that?

[Tom in NH]

Moses,

Another thought.  So, if 14+ volts on the yellow lead to the ICM is normal and expected, then is it possible the Autozone brand ICM could be the culprit and is just not up to the task?  As they are warrantied parts I have been simply replacing with the same.  Because, as I think this through, I am also getting 14+ volts at the + lead of the coil, so the ICM is not doing anything to step the voltage down before it sends the power to the coil.  Like I said, electrical is not my strong suit - swimming as fast as I can here to keep up, but learning and enjoying it.

Thank you.

[Tom in NH]

Hi Moses,

OK, backprobed the orange wire at the ICM that connects to pin 27 on the ecu.  When running, I get a kind of bouncy reading of between .5 and .6 volts.  No pattern or regular timing to the variation.  So if it is supposed to be a 5 volt sine wave, I'm assuming that means I should get a 5 volt reading alternating with zero at regular intervals?

Also pulled my cps and it was oily like it usually is when I check it.  Cleaned it and area around it, retorqued valve cover.  Ran it for a good half hour, and even though the coil is getting 14+ volts at the + lead, it was not hot I would say just warm to the touch.

Keeping my eye out for a used FSM.   Thank you

[Moses Ludel]

Tom...It's doubtful anything would be gained by using a ballast resistor.  The voltage from the ICM to the coil seems like it should be the same as the ignition switch voltage level.  It would be very unlikely that the original AMC/Mopar ICM had a step-down voltage and that the rebuilders are somehow, for some reason, bypassing that function.  There's no reference to step-down voltage in the FSM. 

As for the low .5 and .6 volts reading, this may be the meter's inability to read sine wave.  Assuming that you're using the correct range for the meter (meaning this is not actually 5 and 6 volts rather than .5 and .6 volts), you probably cannot pick up the full voltage reading because the voltage rapidly cycles on and off.

When you cleaned the CPS this latest time, did the idle stabilize?  Was that the source of your rough idle?  Have you run a resistance test on the CPS—cold and hot?

Moses

[Tom in NH]

Hi Moses,

Yes, double checked the scale on the multimeter when I checked the lead from ecu pin 27 and it read a legit .5-,6, but what you say makes sense.  My multimeter is a relatively cheap version and is digital.  I have read that analog meters are sometimes better at reading signals that change quickly.

Idle didn't change after cleaning cps, but it was idling well prior.  The Jeep runs quite well ... until it stumbles which is sporadic and infrequent.  Even less frequent stalling and no start immediately after stalling only a couple of times in the last 6-9 months since this issue was first detected.  Each time there was a no start, I replaced the coil and icm, cleaned cps, but have not ever replaced cps.

Resistance on cps is about 206 cold and 244 hair dryer warm.

New info - checked battery last night and got a reading of 11.9volts.  It did start and I checked with Jeep running and alternator and got 14.4.  This morning battery reading of 12.1.  Weird, as I had battery and alternator checked at one of the chain stores.   Battery was 7 years old, so Got a new battery today.  All seems well and alternator still checks good.

Thank you

[Moses Ludel]

Tom...Sounds like a defective cell or open in the battery.  That's why I brought up the battery issue early-on. 

Anecdote:  My wife drives our '99 Cherokee 4.0L.  She came home from a grocery run and said the engine stalled at an idle (never happens!), it was hard to start, once running she made it home okay.  Cause:  A single dead cell in the battery had caused low battery voltage.

Though this sounds basic, battery trouble is often overlooked.  You need 12.4V minimum for an electronic fuel-and-spark management system to function properly.  When your defective battery gets hot, which is guaranteed if the alternator continues throwing high amps at a "dead" battery, you have an open.  Poor battery connections can also cause this trouble.

A battery needs to be load tested.  The simplest test with a voltmeter is to read battery voltage with the engine off.  If your meter is capable of generating a spark, check battery voltage away from the suspected battery.  (Hydrogen gases are emitted from a defective battery;  an explosion can occur if a spark occurs nearby from attaching jumper cables, connecting the battery cables, scratching with meter pins that create a spark, etc.)  Pull on the headlights without the engine running.  Leave the headlights on for just a few minutes.  If the battery voltage plummets from 12.6V-12.9V down to 11V or less, the battery is shot.

Unless there was a drain on your battery, the readout says the battery was shot.  Normal static voltage for a new or quality battery is 12.8-12.9 volts.  You want 12.8V or better after charging from a run.  12.6V is the minimum acceptable static voltage for a 12V automotive battery.

If the battery is shot, the alternator is continually charging the system at a higher amperage rate.  This could be killing your ignition coils over time.  The wild card here is how long your battery had been demanding a steady, high amperage charge.  Check for a battery drain that may have started this cycle.  If there are no drains and the charge circuit works normally, you simply had a defective battery.  Replacing the battery with a quality, higher CCA rated battery for your specific Jeep application should solve the problem.

Seven years is a long life for a modern lead-acid and other type batteries.  Most lead-acid types never get topped off with distilled water, we're told they are "permanent", zero maintenance.  On lead-acid batteries with caps, I periodically do check the solution level and top off with distilled water just to the fill rings.  Do not overfill.

To extend battery life, I use a Battery Tender on our 4.0L Cherokee and Ram/Cummins batteries year round when parked.  The Ram parks outside in hot and cold weather but always has 12.8-12.9 volts at startup.  What kills a battery is a lower static voltage when the engine cranks.  This causes the alternator to flood the battery with higher amperage.  That high rate of charge, over and over again, reduces battery life substantially.  Our Ram 3500 OEM batteries lasted 13 years, which I'm sure is a record for any diesel powered truck.

Moses

[Tom in NH]

Hi Moses,

Thank you for your full and complete responses.  And yes, you mentioned the battery right up front.  Had it tested, thought all was good....

Update.  New battery shows 11.9 volts this am.  Given your explanation, believe we have found the smoking gun.  Now have to find the actual culprit.  So, that means battery drain when Jeep is "off", correct?  To confirm, with Jeep in off position I disconnected ground and ran a test light from ground cable to neg post and it light up.  So, now I disconnect accessory cables at battery (we have 2 led light sets and an oversized positive lead for audio amp - I think I should check those first) one at a time until light goes out?  If light stays lit, I move on to fuse box and disconnect one at a time?

Any shortcuts or am I on the right track?

Again, many thanks - as I mentioned electrical is always something I avoided.  So I've got a learning curve here.

[Moses Ludel]

Tom...A lamp test works and also indicates the approximate amperage draw in terms of how bright the lamp gets.  A lamp load test on the ground circuit is very valuable.  Your follow-up tests sound practical as well.

As a rule, there are few if any mini-amperage draws:  the radio's "clock", sometimes the ECU/PCM or others.  This level of amperage draw would be evident by a limited, possibly no response in a load test.  You're looking for something more than this minuscule amperage draw when the battery depletes this rapidly.

The audio amp and LED light sets should be run through a typical 5-pin (Bosch style) relay that is ignition switch ON activated—unless you want these circuits to operate with the key OFF as well.  (If you run these circuits from straight, fused battery current, make sure these devices do not draw amperage with the devices shut off.)  Using a relay is simple, and it takes very little amperage from the ignition switch to activate the relay.  The circuit itself can be higher amperage since the ignition switch is simply activating the relay to close and open a higher amperage circuit.

Another simple test for amperage draw is to have the battery (+) cable connected and the negative (-) ground cable disconnected.  All devices off and the key OFF, scratch the negative cable terminal to the negative post at the battery.  If there is a perceptible spark, you have a draw.  The degree of spark indicates the amperage draw.  Caution:  Do not create a spark around a defective or possibly defective, hot battery that may be emitting explosive hydrogen gases!  This same test can be performed with the battery cables both connected and the heavy cable detached at the starter motor.  The scratch test at that end of the cable will be far enough away from the battery to avoid a hazard.

Moses

[Tom in NH]

Hi Moses,

Spent the last couple days tracking and tracing everything I could.  Narrowed the major draw down to what appears to be the ignition switch.  I got a draw (bright light)at the alternator, but not the thick red wire going to the single post.  That tests fine, but rather at the brown/tan resistance wire which according to diagrams I have seen leads to yellow ignition wire and what I have read suggests a faulty ignition switch.  Also got a draw at the green lead running from the hot post on the starter relay which again, diagrams say leads to ignition.  Have just replaced ignition switch, cleaned up as many grounds and wiring issues as I can find including all major grounds - some of which definitely needed cleaning.  I still get a dimly lit test lamp.  Using your scratch test, I get a very faint hard to detect spark.  Used to be a rather pronounced "pop".  I am guessing zero draw is not possible as I have a working original analog clock and the ecm needs some level of current to hold sensor settings, so I am hopeful the much improved and smaller draw is roughly within spec.

Will connect everything back up and I guess ultimate test is let Jeep sit overnight or two and read battery voltage in the am?

Thank you

 

[Moses Ludel]

Sounds good, Tom...You covered a lot of ground.  Yes, an overnight or extended park should be revealing...Moses