Intermittent No-start condition

Hi All,

Have a problem here. 1994 Chrysler T&C. I think it has either a bad
crank sensor or a bad cam
sensor, but I can't figure out which one it is.

The car runs fine until whicher sensor it is decides to stop working. Then
the entire engine just stops.
Attempting to start gets no attempt to fire. I know it's getting fuel
because a guage on the fuel rail
connected when it's not starting shows around 41/42psi. I know it's not a
jumped timing chain
because when it is running it's smooth at all speed ranges. I'm pretty sure
it's not the coil because
all plugs are affected. When it won't start, it's not getting spark, I
checked that.

I think it's heat related because if I let the engine sit for 10-20 minutes
it will start up again.

It just started doing this today. Fortunately it didn't do it when
driving down the highway or anything
like that. This evening I let it idle after getting home and after about 15
minutes of idling it stopped.
Trying to restart it, no-go. Temp guage for coolant does not show
overheating. The CEL is -not- set.
I haven't pulled engine codes yet to see if there are any.

There is a procedure in the FSM for determining which sensor it is, but that
would mean letting it
idle tomorrow probably for about a half hour, and hope that it dies. But I
never know when it is
going to start so the test results could be screwed. And I could let it
idle for a lot longer but it
could possibly not die.

Ted

 

Dont you have a scan tool?

Glenn Beasley
Chrysler Tech

 

I do, and the problem was when I posted this last night it was late and I
wasn't thinking straight enough to think of checking the sensors with it.
Of course, today, cold, it starts up fine and both sensors are fine
according
to the scan tool. So I'll just have to let it idle for a while and cross my
fingers
and hope it dies again. I'll post a followup.


Ted

 

Were you thinking straight enough to check for trouble codes?

 

No, I hadn't pulled engine codes as I mentioned.

Today I spent some time with it. There were no engine codes. The scantool
that I
am using is an OTC Monitor 4000E with Pathfinder 96 and it has a sensor test
that will test
both the crank position sensor and the cam position sensor.

I idled the van in the driveway and ran the 2 sensor tests. The response is
either
-yes- or -no- I got yes's on both while the engine was running.

I idled the engine until it got hot enough and died. I ran the scantool
tests while
cranking. The crank position sensor result was 'yes' the cam position
sensor result
was 'no' with an occasional 'yes' Ah ha, I chortled, bad cam sensor.

I replaced the cam sensor with a new one (that was a chore) and repeated the
idling test. The engine died again! I then waited till it cooled and
repeated the
scantool test, with identical results (yes and no) However the second time
I
gave it a lot more time on the cam sensor and the result on the tool kept
flipping
between 'yes' and 'no' In short, the old and new cam sensors give identical
test results with the scantool.

Poorer but wiser my thought now is either the coil pack or the pcm is the
problem.
I really doubt the pcm because why would it work fine while the engine is
warming up?
The pcm isn't even in the engine compartment and shouldn't be subject to
heat buildup
problems. By contrast, coils sometimes fail in temp-dependent ways.

I'm going to knock off on it today, tomorrow I'll swap the coil pack between
my
95 T&C and the 94 T&C and see if the problem goes away.

Ted

 

My bet would be the crank sensor. On my wife's car, a bad cam sensor
(or disconnecting the cam sensor) will ALWAYS set a code, but the crank
sensor doesn't. Apparently the engine managment software (its a 93
Vision) doesn't check the cam signal for pulses when its not getting
crank signal pulses- it just ASSumes that the engine isn't rotating
rather than saying, "Hmm, cam signal, but no crank- must have a bad
crank sensor!"

That's just my observation based on... well, observation. If anyone
knows a better explanation of why that model car doesn't set a crank
sensor code, I'd love to hear it.

 

Your problem is with the crank sensor, If im not mistaken we talked about
this years ago at the Training center when these vehicles were new, it
seems to me that the DRB/ scan tool would fail the opposite sensor that was
causing the problem the problem,
Replace it and make sure you get the proper air gap setting.

Glenn Beasley
Chrysler Tech

 

Thanks! I had went ahead and bought both sensors when I bought the
cam sensor. I should have done the put-test-light-on-B+-of-coil test in
the FSM instead of depending on the scantool. Sometimes the simpler
diagnostics are better. At least the crank sensor is easier to get at.

The aftermarket sensors come with a small piece of thick paper glued to the
end of the sensor, you push in the sensor until the paper touches the
flexplate (or cam), that sets the gap. Presumably, any scuffing on the
piece of paper is of no consequence.

Ted

 

Absolutely correct on all counts, it was the CRANK sensor. As soon as I
replaced
it I ran the CAM sensor test again, and this time the scan tool showed the
cam
sensor no longer flipping between yes/no, now it was steady on yes as soon
as the
crankshaft started spinning. I test drove the car a good long time and no
problems.

Ted

 

I agree completely. My guess is that the software in the engine computer
is written to key everything off the crank sensor. When it starts getting
crank
pulses as I recall it gets 3 pulses in rapid succession, then there is a gap
then
the crank is 1/2 way round and it gets another set of 2, then more gap and
repeat. The pulses are generated by the holes in the flywheel flexplate.

The program uses the first 3 to ascertain the RPM then times all injectors
and airflow sensing and such to take place at specific times in that
sequence.
The pulses from the cam sensor are handled in that sequence and if they
aren't
there the software sets a code.

If the crank sensor isn't pulsing the computer assumes the engine is at rest
and
isn't executing the instruction sequence that fires injectors and such and
among
other things checks the operation of the cam sensor.

What they would have had to have done to check the crank sensor is to have
a probe wire going to the pcm that came from the starter. If starter is
spinning
and crank sensor isn't pulsing, then set a code. Otherwise the only other
sensor
the pcm can check that looks at engine rotation is the cam sensor, and it's
only being checked in the loop that isn't triggered yet because of the
cam sensor

Ted

 

Our 92 Caravan 3.3L had a similar problem the dealers shop had a tough
time to trace down. It would not start after sitting for a cpl of
days. I could start it with a small shot of eather. It was a poor
wire from the crank sesor. Seems that with the shot of eather it gave
it enough boost to lift the voltage from the alternator to register
the sensor as closed. Replaced the wire and all has been fine
since.

cheers
Marv

 

I dont think so!!!! Let us know when you put the fuel pump in it after it
wont start again.

Glenn Beasley
Chrysler Tech

 

Hmmmm very interesting.. what your saying actually makes more sense.
However the above is the story the service shop at the Dodge dealer I
bought the car from gave me. This was over 2 yearsand over 50K ago
now. I should go look for the seperate wire they say they put in.

cheers
Marv

 

It had nothing to do with a boost lift from the alternator that this
started, if you had to spray eather to get it started.........that would
mean there was no fuel at the injectors, nothing to do with a crank sensor.

 

I remember reading YEARS ago, I suppose it was for a
MPI 4 cyl, that for added reliability the computer can
pulse all injectors to emulate TBI injection if the cam
sensor fails, keeping it running in limp-in mode
with the light on.
In a similar way, maybe it can approximate the crank
signal using the cam sensor, if the crank sensor
quits.

I am thinking this may well apply to newer vehicles also.

?

 

you are thinking of the 4 cyl but it isn't the crank sensor that can quit
and still run. Its the cam sensor, if it goes out the vehicle will still run
but not to efficient.

Glenn Beasley
Chrysler Tech

 

But if the crank sensor wasn't working then how did it generate the spark
to ignite the ether to get the crank sensor working? ;-)

Here's my $0.02 on this, and an explanation on your crank sensor problem.

I think that when the engine is hot that the transmission bellhousing, being
aluminum, expands some. I suspect it's expansion increases the air gap
slightly between the sensor and the flexplate.

This is a hall-effect sensor (magnetic) and distance is important - too far
away
and the sensor won't pick up the sense holes on the flexplate.

In my case I think my crank sensor was weak. Probably from age and being
hot - heat is no friend to electronics. When my engine was cold the gap was
narrower and the sensor was just barely triggering enough to allow the pcm
to sense it. When my engine was warm the sensor was moved just a hair
further
away, just enough to make it no longer register to the pcm.

In your case I think your sensor may be weak/borderline as well. So that
sometimes it would start, sometimes not. Perhaps your fuel pump is weak
also, and the times you were testing, your crank sensor was working all
the time so the ether got it going.

In your case I think when the dealership pulled the sensor and 'rewired it'
and reinstalled it, they put it in slightly further than before. It is
rather difficult
to gap these things. When I installed mine, even with the paper spacer that
is supposed to set the gap properly, when I started the engine there was a
'tick tick tick' that wasn't there before, coming from the sensor, quite
obviously
the sensor was in too far. I shut down and repositioned the sensor just a
hair
out further, and it was OK then. In your case your sensor was too far out,
so
that when the car was new it was sensitive enough so that this didn't
matter,
but when your van got older the extra gap did matter as the sensor lost
sensitivity.
Probably the dealership didn't think that the gap might have been too wide
before, so when they got weak readings on the scope (or however they arrived
at the conclusion) they didn't think about sensor gap and just invented an
explanation about a weak wire to try to explain it.

Ted

 

Acually that is the most common failure of the crank sensor, when it gets
warm due to expansion it shuts down, you can see that with a lab scope

 

Thanks!