Hi All,
I have an 88 2wd GMC that has recently been dying on me. The issue is extremely random, but becoming more often. When driving or idling, the injectors will start skipping and the engine will sputter and often die, but even after the engine is dead, the injectors will continue firing with no certain rhythm. Sometimes I can restart it and drive around the block fine before it acts up again, and sometimes it won't restart at all. Has anyone ever seen this? I have ruled out fuel being an issue as I have replaced the filter and the pump continues to run for a couple seconds after the engine dies. I can watch the injectors start missing just before the engine quits. I have a video of it happening, but not sure how to get it posted.
Thanks,
Dalton
Greetings Dalton,
Great video of the fuel delivery malfunction. (!) Watching your video made me realize that when attempting to remotely
troubleshoot an engine bay how much in the real world I actually rely on aural cues in addition to the visual ones.
Anyway, in order to avoid just firing the parts blunderbuss at your problem, I'd like to describe a little TBI control
theory. Come up with a list of parts involved. And then try to organize those parts in terms of probability.
****
A common sense list of the '88 model year TBI/ECM goals might be listed as follows:
* Protect the $$$ catalytic converter(s) from harm.
* Maintain
positive control over the fuel delivery, engine
running or
not. (ie: Avoid 'Burning down the house'.)
* Provide a positive driving experience, no matter the load, temps, altitude, & driver demand, literally anywhere in the world.
Given this, the symptoms in your engine bay are both unusual and very interesting. In order to achieve the goals
listed above, the ECM is only supposed to allow fuel delivery (via injector cycling) when there is engine rotation.
(And further, to support a combustion/power stroke every 90°)
So watching the injectors not only start to malfunction, but to see/hear the injectors chattering
after the engine
stalls & stops is pretty wild. But we have to decide whether the ECM is a victim of bogus inputs (faking the ECM
to believe that the engine is turning) vs the ECM as a perpetrator of spurious injector drive pulses when there
should be none?
****
In a computer controlled system, at first glance most folks just want to become familiar with all of the electrical signalling.
But in order to troubleshoot bad behavior that doesn't stimulate the onboard diagnostic routines to illuminate the SES light,
it helps to understand which signal has everything else built upon it, and also to understand just how much/how little authority
it has. In the TBI/ECM system, the name of the main control signal is the "REFERENCE" signal:
You must be registered for see images attach
When the 8-sided starwheel starts spinning past the pick up coil in the distributor, this generates a signal that represents another piston at TDC on
the compression stroke. (Assuming that the distributor housing is set right at 0°.) Disregarding what happens below ~400 rpm, this signal from
the pick up coil is converted into a square wave and sent out Pin C from the distributor over to the ECM as the 'Reference' signal.
And in turn the ECM uses this Reference signal to generate both ignition and injector drive pulses. (The exact timing and/or duration of each is
dependent upon all the processing after all the other inputs are sensed, mode of operation, etc.) It is also used to keep the fuel pump relay engaged.
(ie: no engine rotation = no reference pulses = no fuel pump...so that even if the injectors were to malfunction there would still be no uncontrolled
fuel delivery -- nobody wants the possibility of an entire fuel tank to be pumped/sprayed on top of a hot, stalled engine. (!)
****
So the paths involved are:
Spark: Distributor pick up > ICM > ECM > ICM > Coil.
Fuel: Distributor pick up > ICM > ECM > TBI injector.
Now, in an attempt to assign the probability to each possibility, I need more info on the failure mode.
Q: Can this happen at any time, hot or cold? Or does this failure only happen once everything is fully
up to temperature?
When it comes to electronics, when the components reach their end of life and become weak/flaky,
with everything else held to a constant, the
higher the heat more extreme the temperature variation,
the higher the probability that the part will fail to operate properly.
In English, IF this problem only presents with high temps, while the ECM internal temps may rise some
while operating under the dash, the ICM in the dizzy will go from ambient underhood air temps to
essentially the same temps as the engine itself once it is fully warmed up. So if the failure shows up with
heat, the
probability that it's the ICM instead of the ECM. (NOTE: This doesn't eliminate the ECM, it just
lowers the relative probability.)
On the other hand, if it can occur hot or cold, then it could be a loose connection somewhere in the wiring
path. You know those rumble strips on the highway to help prevent people from falling asleep at the wheel?
Those can be very helpful to stimulate an intermittent connection to show up. (Just pick one closer to home
in case you can't get it running again. :0)
Alright, here's another page from the '89 FSM, this time showing the TBI Injector wiring:
You must be registered for see images attach
I included this only because I'm a big proponent of the Big 3 of engine troubleshooting: Spark + Fuel + Compression.
And I thought that it was neat to see the FSM document how to check the TBI injector being driven by the mechanic
simulating a reference pulse using a test light on circuit #430. (REFERENCE)
****
So how do we best troubleshoot this? If you had a classic 2 channel oscilloscope, I'd put the A channel on the Reference line
from the ICM to the ECM, and then put either one of the injectors on the B channel. If the spurious signal was observed on
the A channel, then the 'ECM as
victim of a bad input from the ICM' would be my answer. On the other hand, if the A channel
reference signal from the ICM was steady with engine rpm, yet the B channel started showing the spurious activity, then
my diagnosis would be the ECM is the
perpetrator of the malfunction.
Since you don't have a scope, I'm trying to figure out how a OBD1-era Tech tool would tell us which direction to go?
NOTE: I did find a section in the FSM where they discussed how a faulty TPS sensor could cause injector misfueling,
but I don't think that this is relevant in your case since the injectors keep firing after the engine stops? But in terms
of troubleshooting, this would be easy to prove/disprove, since all you would have to do is disconnect the TPS sensor
and see if the problem stays/disappears. (See attached.)
****
While researching this a couple of other forum members have since made valid suggestions. And I agree that a flaky ground
not adequately sinking surrounding RFI could allow electrical noise to become 'phantom reference pulses', tricking the ECM into
pulsing the injectors when it shouldn't. Or you could even have a thermal failure where normal levels of background noise
take over once the weak electronics goes wonky. So I'm definitely in agreement that you want the engine & VCM to be sharing
a solid ground plane in order to eliminate this as a flaky behavior vector.
Again, I am against changing parts unless I have a defensible theory as to why. Therefore, I will wait until
you can respond with when the failure does/does not occur before I comment further.
Apologies for the length, but I wanted to give you some background troubleshooting theory. And if you haven't
already done so, be sure to download the Factory Service Manuals for your vehicle. (
88+ FSM manual links)
Hope this is helpful. I just checked the replies that came in, and I'm in agreement that the fuel delivery system to the
TBI injectors is not a suspect at this time. The failure is decidedly electrical in nature -- more specifically, there is a
malfunction in the injector drive pulses that continues after the engine comes to a halt. And that is an illegal condition.
Without a scope, you will have to rely upon cleaning connections & a good visual inspection of everything involved. And if
we can get a clue from any temp-related failure info that might provide some key insight.
If something above doesn't make sense then please don't hesitate to ask questions.
Best of luck, and please return & share the wealth when you discover the root cause of the failure.
