2004 3.4l Impala dies suddenly at idle or driving, no spark after

[johnwohl]

I have a problem with my 2004 Impala 3400 sfi that I need some help with.

When idling or driving for a length of time that seems to vary, but usually more than 45 minutes, the car will shut off like you just turned the key to the acc position. No check engine light, no stumbling, no drama. All the lights stay on and everything seems normal except for the car is off and won't start again. It won't start for anywhere from 3-10 to ten minutes, and then runs fine again for a while.

I have taken a video that shows that after it shuts off like this, there is no spark (the video is shaking but trust me). I've also seen twice now a good sized puff of white smoke coming from the throttle body area or towards the firewall from there, and a bit of a smell that I can't exactly locate. Also i've noticed a noise that I don't think I've heard before, it occurs about 19 seconds into the video.

https://youtu.be/d76DvmIt_C8

I am prepared to do troubleshooting. I have a DVOM and an OBDII interface with Torque Pro for android.

The only thing changed even recently was the fuel pump and relay. I tested the fuel pressure during the no start condition and I appear to be getting fuel, just no spark. I can't get a noid light on the injectors as it looks like the whole upper intake has to come off to get at them.

My wallet is very thin and I cannot afford to take this anyplace so I'm hoping you all can help.

[Padgett]

Sounds like a typical ICM overheat issue. I carry a spare in all my 3800s and need to get one for the GTP. 3.1/3.4 has the ICM and coils in the worst imaginable place, tucked under the front exhaust manifold. The Alternator gets extra cooling, why not the ICM ? That said a fuel pump going bad rarely heals & total ignition loss (does the tach drop to zero when it cuts out ?) with no codes is usually the ICM.

[johnwohl]

My coils and ICM sit on top of everything near the firewall, but that is a good point.  I will pull the icm and have it tested at autozone tomorrow.  I wasnt polling the rpm sensor last time but I will do that tomorrow as well and report back.  I did notice that the maf sensor gave wild out of range readings during the crank but no start event.  I was also told to check the crank position sensor which I'm assuming I'll need to check while im replicating the crank no start condition so that should be fun trying to whip the balancer off and crank the engine by hand enough to test the cps before it wants to start back up again haha.

[Crazy K]

The sensor that controls spark is on the rear of block.

 

Most likely ICM, but I have found the crank sensor and two color twisted purple/yellow wire between the sensor and the block to be at fault on other cars.

[Imp558]

Another option is to hit whatever local electronics supply you have and buy a can of "Cold Spray", drive the car until it dies then cool a couple things and try to start it again. If you spray the ICM, CKP and CMS one at a time you'll find something relatively quickly I would think. BTW welcome aboard.

[johnwohl]

The sensor that controls spark is on the rear of block.

 

Most likely ICM, but I have found the crank sensor and two color twisted purple/yellow wire between the sensor and the block to be at fault on other cars.

Oh there is a separate sensor?  What is that called so I can look it up?  I thought the flat plate that the coils plug into was the ICM.  Also, I can't for the life of me find the connector to test the crank sensor.  The info I can find tells me it should be in plain sight but I'll be buggered if I can find it.

 

Another option is to hit whatever local electronics supply you have and buy a can of "Cold Spray", drive the car until it dies then cool a couple things and try to start it again. If you spray the ICM, CKP and CMS one at a time you'll find something relatively quickly I would think. BTW welcome aboard.

Thanks for the welcome!  How can I spray the crank sensor with the harmonic balancer in the way?

[Padgett]

Crank sensor is just a hall effect receiver, again something that rarely heals.

[Galaxie500XL]

Had a similar issue with my LQ1 about 3 years ago.  Replaced the CPS on the backside of the block (near the pan, between the block and transaxle housing, passenger side of block), and haven't had a problem since.

[Padgett]

Well crank sensor is a good thing to keep a spare. Factory Service manual usually has a section called "cranks but does not run" that starts with a test for the crank sensor. Problem is you probably will not be able to test when failed (hot).

 

That said if the ICM and coils are easy to get to I'd swap that with a spare first, to check the crank sensor you have to pull the harmonic balancer.

Edited March 28, 2016 by Padgett

[Galaxie500XL]

It's a 60* V6 (3400).. I doubt the 24x sensor under the harmonic balancer is an issue. That one wouldn't prevent all spark, which is the symptom he's describing.

 

Would be true of the Buick motor, but that's not what he described.

[Padgett]

Got that but ICMs fail the same way, besides the offset crank on a 3800 makes each cyl  120 degrees apart just like a 60 degree V6. However the D1977A uses a single while the D1998 has several.

 

That said, the failing completely when hot sounds like an ICM. Try opening your hood and see if it restarts faster.

[johnwohl]

I didn't have a chance to do any work on the car today as my job took me out of town but tomorrow I am going to check the ICM, and the 7x crank position sensor on the back of the block that controls the ICM.  Luckily that can be tested from the plug at the ICM when I have the no start condition.   I'll probably check the 24x sensor after I get this problem sorted.  From what I've read, if the 24x sensor fails the car will still run, poorly.

[johnwohl]

Right now it looks like the problem is the 7x crankshaft position sensor.  

With the car cool the resistance is 996 ohm, and the range is 500-900.  So that is out of range.

During the no start condition I have 12v and good ground to the ignition control module.  The 7x sensor does not give an alternating AC voltage when cranking.  I get no resistance reading from the 7x sensor at all.

I went ahead and ordered the sensor, but I'll check the wiring for the sensor as well before I change it out.

Sound right?