Intermittent(?) No Start... Have Spark And Fuel?

[gogmorgo]

1991 4.0HO, aw4 4x4.

So this morning I hopped into my MJ and it wouldn't fire up. It cranks and sometimes fires, but it just won't catch and go. I'm stuck outside of the city and don't have much more than basic tools with me.

 

I have spark at the coil. Checked it with a screwdriver in the wire from the coil against the valve cover. Plug wires, cap, and rotor are less than a year old and look to be in good shape.

 

I have fuel. Don't have a pressure gauge with me, but I unplugged the quick-connector at the rail and there was fuel pumping out during cranking. It occured to me after I did that that I should have just poked the schrader valve on the rail and that there's now air in the rail. My bad.

 

All fuses under the hood are visibly OK (pulled them all, don't have a multi-meter).

 

My truck was running fine when I parked it last night. This may be an intermittent issue, as similar things have happened twice in the past.

One time, I couldn't start it in the evening, came back the next morning and it fired up on the first turn of the key.

A couple weeks ago, it did it again in the evening. My low fuel light was on when I parked it, so I just assumed I'd run out. I put about 20 litres of 94 octane into in (an old jerry can of what we run in our FSAE racecars) and then it took a long time to start it. I wore the battery down and had to get a boost from a buddy. But then when I went to fill it after driving directly to the gas station, I only put another 50 litres of fuel into the 90 litre tank. I've definitely run the tank much lower than that on a few occasions and not had an issue.

 

A buddy's coming to pick me up for something, and I can probably grab tools at our shop cause I'll be going there anyhow.

[HOrnbrod]

Try unplugging the replugging the CPS 3-pin connector at the back of the engine. I had this similar intermittent problem on my 91 and that usually fixed it temporarily. The permanent fix was a new Mopar CPS. Does your truck ever cut off when underway? Mine was doing that too.

[gogmorgo]

I've had it cut out very briefly, really just a cough, under heavy acceleration once or twice, but the last time I remember it doing that was back in October, iirc. I'll definitely check that connector though. I'll grab a multimeter and test the CPS as well.

 

While looking into the CPS testing procedure, I noticed that one of the possible symptoms of a bad CPS is the fuel gauge not displaying properly. As of about a month and a half ago, even if I fill it to right up at the top of the filler neck, the gauge only ever reads 3/4 full, and it wanders around by about a 1/4 tank in either direction but with a general downward trend. Just wondering how the two could possibly be connected?

 

Source: http://www.jeepforum.com/forum/f11/no-start-I'm-assuming-its-crank-position-sensor-803967/ quoted section in post#2

 

Checked the codes on my way out the door after posting this, and none other than 55. About a month ago I did have a 41 (Alternator field switch.
Switch not operating correctly and check charging system.?). It even threw a Check Engine light, but I cleaned some corrosion off the battery terminals and it hasn't come back.

[HOrnbrod]

I have seen that fuel gauge / CPS interaction writeup before, but it wasn't the case with me. Also I did the ohms test and it read okay. There's no real way we can test the CPS like there is on the Renix. My distributor sync sensor went out about the same time now that I think back when I had the random starting / cutout problem, and I just swapped the 91 distributor with one from a 96. Didn't you recently replace your distributor? Or is my memory corroded?  :hmm:

[gogmorgo]

Your memory's fine. That was back in March/April. Got a code 54 and swapped the dizzy for a cheap '95 from a Rock Auto closeout sale because it was cheaper and less work than a new sync sensor. It was the end of the semester and time and money were more scarce than they typically are. It seemed to be holding up. Haven't seen another code 54 yet at least. I put a new Napa/Echline brand cap/rotor to replace the cheap ones that came with the dizzy at some point in the summer after noticing the terminals on the top of the cap were rusting. I guess that doesn't really speak to the quality of the dizzy and that sync sensor, then, does it...

At any rate, yeah, I had some pretty serious cutting out issues and a few no-starts before I swapped the dizzy, and they went away.

I've got a fuel pressure gauge and multimeter from the shop for a bit of testing, and I'll pull some wires and check for spark at the plugs tomorrow if the CPS doesn't seem to be the issue. It's been dark for a few hours now and I've got some other stuff to deal with that is slightly higher priority.

Fuel pressure gauge is because I've got a friend who had a Honda show very similar symptoms when his fuel pump was dying. It's almost more for his benefit than mine, but you never know till you know. He was around both other times this happened.

[carnuck]

I recently replaced my fuel filter and discovered the old one was slowing the fuel pressure buildup and causing longer cranking to start.

[HOrnbrod]

I recently replaced my fuel filter and discovered the old one was slowing the fuel pressure buildup and causing longer cranking to start.

 

:clapping:

[gogmorgo]

Put a new filter on it in May cause it looked old enough to be original. Poured some black mud that had the consistency of ridiculously thick hot chocolate out of the old one. New filter didn't really change the way it ran.

Unless it didn't start at all, it's always fired up within two seconds of cranking. Stretch that to five for -20°F with the block heater unplugged. I typically turn it to on and wait the couple seconds for the fuel pump to prime before turning it over.

[gogmorgo]

Well, finally got around to checking stuff out this morning before school. (Don't have class till 2:30). I wasn't very optimistic about it starting when I realized that I'd forgotten to plug it back in when it didn't start and it being -20ºF out, and having run the battery down a fair bit trying unsuccessfully to start it on Saturday. And after doing absolutely nothing to it, what do you know, it cranked over once, coughed a little, and fired up.  :dunno:

 

So since it was now running, and it was cold out, and the engine was cold, I didn't really want to shut it off right away. So I figured I'd check the fuel pressure, being the least invasive to engine operation. Oops. For anyone reading this with similar lack of foresight, don't try to connect a fuel pressure gauge while the engine or fuel pump is running. You'll spray fuel everywhere, right above the exhaust manifold, and more importantly to me this morning, cover your hand with a liquid at -20ºF. Not fun.

But anyway, fuel pressure at idle was ~32 psi, revving saw it hop to ~38, and with the regulator vacuum line off, 41psi. So I think that checks out. Pulling the fuel pressure gauge off a running engine also sprays fuel everywhere, btw. :doh:

So I drove to school, and once I parked, I tested the cps through the above procedure and had infinite resistance on all of the scales (including the 2000kΩ scale). I'm assuming that's a good sign, but it could still be meaningless. The connector wasn't corroded at all.

 

If it happens again that I can't start, I'll unplug the CPS connector and see what happens.

[gogmorgo]

So it happened again, although more drastically this time. I was driving, and the truck shut off, after backfiring a bunch as I slowed down to a red light. I was able to start again, but only made it a short distance down the road before it backfired a whole bunch (massive amount of clattering) and died.

Couldn't get it to restart this time. I unplugged and then plugged back in the CPS connector, and was able to get it going with some throttle input, but it died when I tried to let it idle.

I checked the codes, and had a 51 and a 53, so lean condition and pcm fault.

I managed to get it started in neutral, and neutral dropped to get going without dropping to an idle. It made it maybe another half mile before it started coughing and backfiring, at which point I shifted back into neutral and brought the revs up to 2000rpm. It smoothed out, and I was able to go back into gear. Drove another 1/2 mile, hit a light, and put it in neutral so I could keep the revs up. Neutral dropped again to get going, and the wheels spun on ice and I slid sideways, took my foot off the gas, and the engine backfired and died. Couldn't get it to restart. Got a tow off the highway to a safer place from a passerby.

I unplugged the cps and plugged it back in, but it didn't work. Unplugged a bunch of other stuff, but at this point my battery was basically dead and could barely crank the engine over, so I called a buddy. As soon as he showed up, for $#!&s, I tried to start it, and it fired up like nothing was wrong.  :fs1:

 

So I drove it over to a better (read legal) parking spot and left it, cause I didn't want to deal with it not making it all the way to school or home and I was running late for a test. So it's still sitting out there.

 

 

But questions:

Would a bad cps lead to a lean condition? I've been wondering this for a while, as the engine's been running hot if I leave it idling for extended time.

Would unplugging the CPS lead to the computer throwing a PCM Fault code (53)?

And could a known cracked flexplate lead to faulty CPS readings, or damage the CPS?

 

I'm planning on replacing the flexplate ASAP. I've got reading week next week, so I should have lots of time. Also planning on replacing the harmonic balancer, as my buddy noticed that the serp belt was riding a little too close to the engine on the fan pulley, and the HB has a pretty good wobble to it.

[HOrnbrod]

Never had those codes. How old is your O2 sensor? Check for 12V (heater voltage) on pin D (BLK/TAN Wire) of the four pin O2 connector and verify the harness hasn't fried on the exhaust. If all okay I'd think about replacing the O2.

 

Make sure the vacuum line is connected to the MAP, and if the above doesn't help, I'd go ahead and throw on a new CPS. Disconnect the battery after each to clear the codes and see how it runs.

[gogmorgo]

O2 sensor was on the truck when I got it, don't know how old it is. Wouldn't surprise me at all if it's bad.

I've got a pretty full day tomorrow but I should be able to do a couple small tests. The MJ's parked in my spot on campus for now. It even got there mostly under it's own power. I went back to where I left it this afternoon, and it started right up. I ran some errands, dicked around in the fresh snow on campus for a bit, then went to park. It died once and wouldn't restart, but I left it for twenty minutes and came back and it fired right up.

 

The codes cleared themselves... pretty sure I need new battery cables, so that's another thing on the list of stuff to do. Le sigh.

[gogmorgo]

Well, it took me four days over the last weekend to get the job done (first time I've ever dropped a transmission so I was going slow so I didn't @#$% anything up), but the new flexplate is in as of Monday and since then the truck's been running beautifully. The crack in the old flexplate ran all the way around the crankshaft bolts, splitting the flexplate in two separate pieces with a couple small chunks missing, so it's frankly a miracle that it ran and drove at all. There was about 10° of play between the two pieces of the flexplate, so God only knows how the ecu was able to figure out any kind of timing.

 

But it runs a $#!& ton better. It's awesome. Hasn't stalled out or refused to start yet.

[HOrnbrod]

Well, it took me four days over the last weekend to get the job done (first time I've ever dropped a transmission so I was going slow so I didn't @#$% anything up), but the new flexplate is in as of Monday and since then the truck's been running beautifully. The crack in the old flexplate ran all the way around the crankshaft bolts, splitting the flexplate in two separate pieces with a couple small chunks missing, so it's frankly a miracle that it ran and drove at all. There was about 10° of play between the two pieces of the flexplate, so God only knows how the ecu was able to figure out any kind of timing.

 

But it runs a $#!& ton better. It's awesome. Hasn't stalled out or refused to start yet.

 

Were you able to discover the cracked plate by removal of the inspection plate? Or was it just making a lot of clatter?

[gogmorgo]

Well, both.

 

It's been making a racket on and off since the summer, which I traced back to the bellhousing area sometime in September. It was either in late September or early October (before the snow at least) when I pulled the inspection cover off and saw what I thought was a barely visible hairline crack right up next to the crankshaft flange. At which point I ordered a new flexplate. I was planning on putting it in in December, but wasn't able to make time. As a full-time student I don't really have many opportunities to take my MJ out of commission for more than a day or two, which is why it took me so long to get around to it. I wanted a couple extra days in case something happened. As it turned out, I needed the couple extra days.

[carnuck]

Was the thin small plate under the crank bolts there? If not, you found the cause of the cracking.

[gogmorgo]

The cause of the crack was the bellhousing bolts falling out and/or shearing many moons ago. As was stated above, if not in so many words.

 

I followed the fsm putting it back together, and everything is there that should be there.

[cruiser54]

I recently replaced my fuel filter and discovered the old one was slowing the fuel pressure buildup and causing longer cranking to start.

The “this worked for me” advice.

 

Try some diagnostics on your own to determine the cause. Just because my ride does ____________ and I did __________ to fix it, should actually mean nothing to you, unless of course you just like wasting money.