[urq] intermittent ignition cutout

[Buchholz, Steven]

... guess I'm going to have to get my car running again just to test
this one out!  I know that it has been discussed before, but I've
"reverse engineered" the circuit inside the ECU, and the Hall Sender
signal directly gates the TDC sender signal, and if the Hall signal goes
away the TDC signal should too.  Now that I'm thinking about it I may go
back to check my notes ... it may be possible that the signal will float
to the active level, which would mean that all TDC signals would be sent
to the processor, and in that case the engine might well continue to run
... but I'd hate to see what might happen if the computer just happened
to see the TDC at the top of the exhaust and resync itself!  In any
event, just because the engine runs with the Hall Sender disconnected
doesn't necessarily imply that the engine will continue to operate given
any particular malfunction of the Hall Sender.  If you have the car in a
failed state, test power and ground on the Hall Sender, and use a DVM or
test LED to verify that the Hall Sensor signal is toggling as the engine
is cranked.  

I'm not exactly clear on how you would expect the ECU to operate if you
replaced the reference or speed sensors with a resistor ... the ECU
depends on these signals transitioning ... in fact if you look at them
with a scope they look quite similar to the odd looking waveforms you
see in the IST doc.  The shield is only there to minimize noise getting
coupled into the very low level signals.  It wouldn't likely matter that
much either way if there were a problem with the shield, although you
might take an ohmmeter to the shield and make sure that one of the
non-shield leads isn't getting shorted to the shield when the problem
arises ... 

Ed, you are familiar with the V8's propensity to "fry" the same type of
flywheel sensors ... I know that they seem much more robust on the I5,
but since they are so key to the operation of the engine and given that
you've already tested for most of the other standard culprits, you
really should just try swapping one out (assuming that you don't find
the Hall Sender bad).  

Bon chance!
Steve B
San Jose, CA (USA)
>  
> Going from memory  I think the third wire you're referring to 
> is the shield wire for the 
> rpm  sensor. I believe it goes to the ecm and it taken to 
> ground internally by the ecm. As for the tdc sensor I think 
> that's the one in 
> the distributor. If that's the case once the engine starts 
> the computer doesn't look at that
> signal again. The car will still run perfectly with the 
> distributor unplugged. The rpm sensors are 
> known to fail seeing as they are of the variable reluctance 
> design. The windings inside the sensor  can and do go open 
> circuit based on temperature variances due to expansion of 
> the sensor windings. Once the sensor cools and contracts, the 
> sensor will function as intended.
> I would suggest that you get an ohm meter and measure the 
> resistance of the sensor to get a baseline # when the engine 
> is cold. Then immediately after the engine has stalled take a 
> second reading and compare the two readings.
> Keep on mind that the readings will vary slightly even with a 
> good sensor because temperature has a slight effect on the 
> windings. you are going to be looking for a large number 
> change usually in the thousands of ohms. Keep in mind that 
> the sensor has to be unplugged to take the readings.
> One last thing. Although this kind of failure is extremely 
> rare, I've seen a few vehicles (not urq's) where the tack 
> itself will short internally and take the primary side of the 
> coil to ground. When this happens the tack will read 0 rpm 
> even when  the engine is still in a coast down condition. If 
> it's not too much trouble disconnect the tach and try a road 
> test again.
> 
>  
> I hope this helps.
>  
> Cheers,
>  
> Craig Cook.
> Ed Kellock <ekellock at adelphia.net> wrote:
> Had hoped that my IC and burnt wire adventures would yield 
> some change to the cutout.
> Still chasing the wire, so won't know for sure until its resolved.
> 
> But, I've gotten a pretty clear fix on the behavior of the 
> cutout, purely subjectively of
> course. The cutout does appear to be temperature sensitive to 
> some degree, though not
> summer heat temp type sensitivity.
> 
> The car starts up beautifully and runs nicely as it warms, 
> but as it does, the cutout
> manifests. It's almost always momentary and as long as it's 
> not a idle (even sometimes if
> it is), the cutout is short enough that the engine will catch 
> and continue running.
> 
> I bought some resistors to put in place of a couple of 
> sensors so as to fake out the ecu
> and either eliminate or isolate the issue. In looking at the 
> tdc and rpm sensors, there
> is a 3rd wire for ground. I haven't quite sorted out how to 
> rig that to fake out the ecu.
> 
> I have swapped in a spare ignition module (thanks Dennis!), 
> however the behavior was unaffected.
> 
> The cutout is definitely electrical in nature. The tach drops 
> to zero and I think even
> the fuel pump quits, though I'm not positive about that. I 
> replaced the ignition switch
> and the fuel pump relay about a year or 2 ago while 
> diagnosing a no-start situation.
> 
> I've been wondering about the load reduction relay. Mine 
> appears to be original. I do
> hear a click when the ignition cuts out. Haven't been able to 
> isolate it yet though.