Stumped - No Start stock '87 duke

[Series8217]

Do you have a scan tool?

[Cheeze]

Yeah I bought an OTC Genisys with the OBD I and II cables and chips. Software through 2005 domestic ECM, ABS and AIRBAG modules.

[The Dark Side of Will]

The MAP and coolant temp should be ignored until closed loop and my scanner says air temp is pretty accurate ('87 first year for IAT).

So today's plan is verify correct fuel pressure. New gauge. I did key on with the supply line disconnected from the TB and it pushed out a good amount of fuel when it primed but I didn't measure the volume to compare to the factory spec. So I'm hoping the block is in the throttle body because I already have the rebuild kit and I don't want to drop the tank again to disconnect the fuel pump and blow out the supply and return line.

MAP is used when cold. Coolant temp is responsible for cold start enrichment and needs to be accurate. When the engine's cold, coolant temp should be basically the same as IAT. Is this what you see? An open circuit coolant temp sensor can cause flooding. Open circuit will show as -40 or so on the scanner.

Do the injectors start leaking as soon as you turn the key to run, or do you have to crank it first?
Do the injectors leak if you hotwire the pump to run with the key off?

[Cheeze]

Do the injectors start leaking as soon as you turn the key to run, or do you have to crank it first?
Do the injectors leak if you hotwire the pump to run with the key off?

Interesting, I did not know MAP was used in cold open loop operation. So I checked a couple things. Before starting the car:

Coolant Temp: 64F
Man Air Temp: 68F
Fuel Pressure: 12 lbs
Map Voltage: 4.75V
Man Pres: 29iHg
The engine must be cranked for fuel to come out the injector. There is not a drop of fuel if I hotwire the pump for constant on. I can hear the fuel flowing through the TB and gauge reads 12psi.

I cleaned the crap out of the throttle body to make sure there was no blockage. I don't believe there was one and there definitely isn't one now. All new gaskets, wire brushed, carb cleaner, hot water soaked then dried with compressed air. It's the cleanest thing in the engine bay. Soaked part of the injector in Techron while the rest was being worked on.



I did have a blocked gauge port as the new gauge reads fine and without delay. I also had a bad IAC valve (failed since this problem started apparently). Now the car is really iffy on start, sometimes it will start without throttle, sometimes I have to feather it and it gives me some really nasty backfires. It sounds like my starter is getting ripped to shreds. I am using a 3800 starter off some Buick in the junkyard. It's been working great the last 4 months and still cranks good. I got under it to inspect the flywheel and all the teeth are still there luckily. When it starts there is a fairly normal idle that wanders from 850 to 1000 rpm. It's always done that. Let it warm, closed loop readings -

Coolant Temp: 195F
Man Air Temp: 138F
Fuel Pressure: 12 lbs
Man Pres: 8iHG

So I figured I would drive it around a bit. Took it for about a 10 mile beating and it ran fine and doesn't choke out like it used to around 4k. Revs fine, runs up to 80 mph just as slow and normal as it always has. Pulled back in the lot, turned it off - won't fire back up. WOT clear it then it cranks and cranks and spits and bucks, backfires and starts. I still don't get it. Would like my car to not be living life inside.

[Series8217]

Make sure the TPS is working properly. Monitor TPS voltage and throttle percentage. The algorithm works like this:

* Record the output of the TPS with a 5V reference voltage applied.
* If the TPS voltage is less than the minimum threshold, set an error flag.
* If the TPS voltage is higher than the minimum threshold, set an error flag.
* If no error flag was set:
* If the TPS voltage is less than the previously recorded minimum, record the new minimum. That value is now 0% throttle.
* If the TPS voltage is greater than the previously recorded maximum, record the new maximum. That value is now 100% throttle.
* The current % throttle position is: (current value - minimum value)/(maximum value - minimum value)

If you get sporadic voltage dropouts or spikes it can cause weird behavior because you either can't reach 100% throttle (because you can't reach the highest recorded voltage from the spike), or can't reach 0% throttle (because you can't reach the lowest recorded voltage from the dropout). On the low end of throttle positions that can affect startup and idle.

[Shaun41178(2)]

4.5 volts on the map sensor tells me your computer thinks you are near 0 vacuum aka full throttle. Key on engine off I think map voltage should be around 2.5.

[Cheeze]

I looked at the TPS last night as well. Full closed pedal throttle is 0% and 0.49 volts. As I slowly step into the pedal the percentage rises to 80% max and voltage did not show any spikes (I had the signal on scrolling graph mode to catch spikes).

This was done ignition on, engine off. I have another TPS from the other engine I can try as well.

Talked to one of my buddies today that knows Fieros fairly well and he suggested Crank Position. He said it would be most erratic during crank before there is good engine speed. I had a bad OEM CPS about 2 years ago that was split down the side. I put a BWD piece of shit in there from Auto Zone and the first one did not work. I think they have really bad quality control because it was shorter than the OEM and I'm guessing wasn't picking up a good signal (which was a nightmare to conclude my new sensor was indeed bad). The second one worked. Interesting article on Hall Effect vehicle sensors here if anyone wants a read: http://www.google.com/url?sa=t&rct=j&q= ... 9499,d.aWw

Unfortunately I can't really tap into the sensor output being where it is. But I'll monitor RPM while cranking, then pull it out and inspect it (I did add the di-electric grease and rubber seal when installing), clean the grounds for the ignition module and probably order up a new AC Delco OEM sensor from Rock for $27. Perhaps I messed something up washing the engine. I live in Michigan and drive this thing year round. I wash it almost every day at the quarter car wash which is still less expensive than commuting with the F-150. I spray down the engine bay maybe once a week if we got lots of snow and salt being careful not to hit hot coils, ignition module, exhaust, etc. But that area probably gets hit with water more often than most and this no-start issue did begin on a rainy day.

And I'm done with these cheap-o sensors and car parts. I shouldn't be in such a rush to fix it with local parts store crap. I would order up a new TPS while I'm at it but the good Delco one is $52 which is a little pricey to buy just for the hell of it.

[Cheeze]

4.5 volts on the map sensor tells me your computer thinks you are near 0 vacuum aka full throttle. Key on engine off I think map voltage should be around 2.5.

Good idea, I had to think about that one for a while last night as well. Engine off the manifold should normalize to atmospheric pressure through idle circuit, head valves and throttle blade which would resemble a wide open throttle high manifold pressure (compared to running with a vacuum). Found this tidbit on obd-codes.com:

"A good way to diagnose if the MAP sensor is to blame would be to compare the KOEO (key on engine off) MAP sensor reading on a scan tool with the Barometric pressure reading. They should be the same because they both measure atmospheric pressure. If the MAP reading is greater than 0.5 volt off of the BARO reading, then replacing the MAP sensor would likely fix the problem. Otherwise, start the engine and observe the MAP reading at idle. It should normally be about 1.5 volts (varies according to altitude)."
http://www.obd-codes.com/p0108

I just looked up what atmospheric was here last night in the weather archive. It was 29.19 inches of mercury and my scan tool didn't give a decimal output so 29 is as close as I can get.

[The Dark Side of Will]

4.5 volts on the map sensor tells me your computer thinks you are near 0 vacuum aka full throttle. Key on engine off I think map voltage should be around 2.5.

What OP wrote is correct.
Key on, engine off, there is atmospheric pressure inside the manifold. MAP output is 4.75V and manifold pressure is 29inHg is correct.

If the engine's idling at 20 inches of vacuum (NOT the same as 20 inches of pressure), then MAP reading will be down at or under 2V.

[The Dark Side of Will]

The engine must be cranked for fuel to come out the injector. There is not a drop of fuel if I hotwire the pump for constant on. I can hear the fuel flowing through the TB and gauge reads 12psi.

I cleaned the crap out of the throttle body to make sure there was no blockage. I don't believe there was one and there definitely isn't one now. All new gaskets, wire brushed, carb cleaner, hot water soaked then dried with compressed air. It's the cleanest thing in the engine bay. Soaked part of the injector in Techron while the rest was being worked on.

Did the previous failure mode return? By that I mean does it still dribble/spray/pour fuel with KOEO?

If yes:
Next step: Hotwire the injector.

Put +12V on the +12V side of the injector. Connect a wire to the low side and tap the other end as briefly as you can against a ground. The injector should give you a spritz for every tap and then should STOP passing fuel. If it continues to pass fuel after the first tap, but when it's not grounded, there's a problem with the injector.

I'm conjecturing that a worn out injector may hold pressure until it's first opened, then not seat properly and leak after the first time it's opened each start cycle.

If the injector checks out ok, but fuel still dribbles/sprays/pours with KOEO after first crank, then there's a problem with wiring or ECM.

Now the car is really iffy on start, sometimes it will start without throttle, sometimes I have to feather it and it gives me some really nasty backfires.

Pulled back in the lot, turned it off - won't fire back up. WOT clear it then it cranks and cranks and spits and bucks, backfires and starts. I still don't get it. Would like my car to not be living life inside.

CLASSIC symptoms of flooding from excessive fuel delivery.

[Cheeze]

CLASSIC symptoms of flooding from excessive fuel delivery.

God I hope you are right on the sticking injector theory. My dumb ideas haven't worked so far. I pulled ignition module, the crank sensor just looks so cheap. I'll have to snap a pic of it. There are pieces of metal or different plastic sticking out of the side like it was molded with random junk. I wanted to replace it anyways with something quality. I would really like to drive the car to the Smokey mountains this summer and Florida at Christmas time and I don't want to be second guessing all the stuff I've done along the way. Probably a bad idea but I've driven it to New Jersey and Boston before and got 40 mpg ALL highway doing 65-75. I just like the car and want to keep it going. Won't have time for it next couple days but I'll let you know what happens. Thanks to everyone for the help.

[Shaun41178(2)]

4.5 volts on the map sensor tells me your computer thinks you are near 0 vacuum aka full throttle. Key on engine off I think map voltage should be around 2.5.

What OP wrote is correct.
Key on, engine off, there is atmospheric pressure inside the manifold. MAP output is 4.75V and manifold pressure is 29inHg is correct.

If the engine's idling at 20 inches of vacuum (NOT the same as 20 inches of pressure), then MAP reading will be down at or under 2V.

that seems logical. I got confused somehow with 2 bar map voltages.

[Cheeze]

The injector will not dribble/spray/pour fuel with KOEO. It will still flood during cranking however. I performed the hot-wire injector test:

Jumped fuel pump for constant pressure. Used jumper wires with alligator clips, the +12V right to the battery and the - neg to chassis ground. Everytime I closed the circuit very quickly I would get a short spritz of fuel but not for longer than the brief closed circuit contact. Injector appears to be operating well.

I then installed the new AC Delco Professional crank position sensor, reading rpm during crank - no change.

Installed new Bosch Premium spark plug wires - no change.

I can still hold the pedal down and achieve engine start and it will come down and idle. Fuel spray pattern looks normal compared to what I've seen in the past. This is spray pattern at 1000 rpm idle:





Not sure if this is all true but I read this on http://www.gmtuners.com/tech/modes.htm -

STARTING MODE

When you first key on the ignition, the ECM loads the information off the PROM chip and performs a quick, self-diagnostic test. Some later ECMs also pulsed the injectors during this period to aid in what is known as “quick start” (obviously helping the engine startup more quickly). All of this happens in a very short amount of time, usually much less than 1 second after the key is turned on. Also during this period the fuel pump is activated for 2 seconds to “prime” the system. Next, the ECM waits until it receives reference pulses from the distributor (or crank sensor). Once it receives reference pulses (indicating the engine is turning over) it will use air/fuel ratio values stored in the cranking fuel pulse tables (in the PROM programming) which are based on the coolant temperature to determine how much pulse width (amount of injector ON time) it needs to command the fuel injectors open. The lower the coolant temperature, the richer the Air/Fuel ratio. The higher the coolant temperature, the leaner the Air/Fuel ratio. Ignition spark advance and IAC valve position output functions use similar tables.



CLEAR FLOOD MODE

If for some reason the engine should become flooded, provisions have been built into the computer to help clear this out. If during cranking of the engine you depress the throttle more than 80%, the ECM will enter what is called “clear flood mode”. In this mode the ECM commands a delivered Air/Fuel ratio to very lean (usually 20:1). The ECM will stay in clear flood mode as long as the throttle is 80% or higher and the RPMs are less than about 600. As soon as throttle position falls below the 80% threshold or RPMs go above 600, the ECM disables clear flood mode and calculates fuel delivery based on coolant temperature and other factors it normally uses.

Interesting clear flood mode only activates above 80% throttle and I can only achieve 80% throttle. I hate coincidences.

Like you said Will - this pretty much rules out everything except a ECM or wiring issue.

I'm going back through and following the TBI troubleshooting guide that I had posted earlier and they suggest verifying the basics first being the following 8 steps:

Pink Ignition wire MUST be connected to 12 volt switched ignition that receives power during crank and key on.

1. Check to make sure your check engine light is not on, or that it is on with the key on but the engine is not running.
2. Make sure that the red battery wire is connected to a battery source (It is highly recommended that this wire is connected directly to the battery) and the pink wire is connected to an ignition source. If your ignition wire is not connected to an ignition source your ECM will not be powered while cranking the engine.
3. Check that the ground wire is securely fastened to the block and that the interface between the block and the terminal are clean.
4. Ensure that there are NO vacuum leaks.
5. Ensure that your MAP sensor is connected to a full manifold vacuum source and not a ported source.
6. Set the ignition timing correctly making sure that you disconnect the set timing connector to set it. In some cases you cannot set the timing with the connector disconnected and keep the engine running. If this happens set timing to 15 degrees, allow the engine to fully warm up, then disconnect the set timing connector to set the base timing to the correct specification.
7. Ensure that you have full manifold vacuum routed to your fuel pressure regulator (if equipped on MPFI systems) and there are no vacuum leaks with this connection.
8. Check your fuel pressure to ensure that you are providing the proper pressure to the system.
Fuel Pressure

99% of all issues are usually taken care of with one or more of these 8 steps of diagnosis.

Perhaps a ground issue? I did replace the engine less than a year ago so many connections were loosened and re-tightened. The vehicle also gets heavily road salted in the winter. I was thinking about this the other day and the fact this was not a catastrophic failure might indicate something. The engine had previously died at idle and was requiring more and more throttle to launch. Maybe I did have a bad fuel pump coupled with a ground that is getting worse (but still allows the car to run?). I don't know. I am going to check and clean every ground I can find in the engine bay then begin the step-by-step.

[Series8217]

I looked at the TPS last night as well. Full closed pedal throttle is 0% and 0.49 volts. As I slowly step into the pedal the percentage rises to 80% max and voltage did not show any spikes (I had the signal on scrolling graph mode to catch spikes).

You would likely need an oscilloscope to catch any spikes.

It sounds like you have a problem with the TPS or related wiring.

[The Dark Side of Will]

Interesting clear flood mode only activates above 80% throttle and I can only achieve 80% throttle. I hate coincidences.

Like you said Will - this pretty much rules out everything except a ECM or wiring issue.

I'm going back through and following the TBI troubleshooting guide that I had posted earlier and they suggest verifying the basics first being the following 8 steps:

Pink Ignition wire MUST be connected to 12 volt switched ignition that receives power during crank and key on.

1. Check to make sure your check engine light is not on, or that it is on with the key on but the engine is not running.
2. Make sure that the red battery wire is connected to a battery source (It is highly recommended that this wire is connected directly to the battery) and the pink wire is connected to an ignition source. If your ignition wire is not connected to an ignition source your ECM will not be powered while cranking the engine.
3. Check that the ground wire is securely fastened to the block and that the interface between the block and the terminal are clean.
4. Ensure that there are NO vacuum leaks.
5. Ensure that your MAP sensor is connected to a full manifold vacuum source and not a ported source.
6. Set the ignition timing correctly making sure that you disconnect the set timing connector to set it. In some cases you cannot set the timing with the connector disconnected and keep the engine running. If this happens set timing to 15 degrees, allow the engine to fully warm up, then disconnect the set timing connector to set the base timing to the correct specification.
7. Ensure that you have full manifold vacuum routed to your fuel pressure regulator (if equipped on MPFI systems) and there are no vacuum leaks with this connection.
8. Check your fuel pressure to ensure that you are providing the proper pressure to the system.
Fuel Pressure

99% of all issues are usually taken care of with one or more of these 8 steps of diagnosis.

Perhaps a ground issue? I did replace the engine less than a year ago so many connections were loosened and re-tightened. The vehicle also gets heavily road salted in the winter. I was thinking about this the other day and the fact this was not a catastrophic failure might indicate something. The engine had previously died at idle and was requiring more and more throttle to launch. Maybe I did have a bad fuel pump coupled with a ground that is getting worse (but still allows the car to run?). I don't know. I am going to check and clean every ground I can find in the engine bay then begin the step-by-step.

Still the original PROM, right? It's rare for a PROM to go bad, but it does happen. Do you have a DVM with Hz or Duty Cycle functions? You could back probe the injector while the ECM is flooding the engine just to verify that the injector's being pulsed and not just commanded to constant on.

Check and maintain your ground connections first. On the V6 cars, there's a ground strap from the forward head to the left deck lid hinge box. I @$$ume there's something similar on 4 cyl cars. There's at least one more ground strap, but I don't remember where it is right off.

If you can't get a reading of more than 80% on your TPS, you need to either adjust or replace until you can get over 90%. Not only does that enter flood clear mode, that enables power enrichment so the engine won't ping when you have it flat on the floor trying to keep up with 80 mph traffic.

[ericjon262]

won't fix the main issue, but when you checked the TPS, were you going by the gas pedal, or by opening the throttle by hand? IIRC, the duke throttle linkages sometimes don't allow the throttle to fully open.

[ericjon262]

One of the most frustrating cars I've ever worked on was a TBI 350. Seems like a simple concept, but all the carb technology mixed with fuel injection made for some migraine inducing problems. Vent away man, we all understand.

The F-body guys have flooding when their regulators go out because the regulator lowers the pressure during start and idle. I have no idea if the duke TBI works the same way but the units look identical minus an injector. But I just looked up the stock idle pressure and it says 9-13 lbs. I have no idea.

TBI fuel pressure regulator is atmosphere referenced, fuel pressure should remain relatively flat at all times. only reason it would dip on startup is because the pump isn't running, but I can't think of a car that runs the fuel pump while cranking.

I'm not familiar enough with the TBI system to know what would cause 0 gauge pressure other than a blocked gauge port.

My BMW had a SOLID blockage in the return line inside the tank... it would run, but had about 80 psi of fuel pressure against a dead-headed regulator.

just a thought, Jumper the fuel pump relay and hook up your pressure tester. if pressure levels out at 9-13 PSI, no clog, if it goes higher. clogged.

[Cheeze]

Cleaned all the grounds I could find. Two on the bellhousing studs, one at the head next to the battery and one below the battery tray. No change.

Tried the backup ignition module/crank sensor / coil pack assembly from another running duke. No change.

Hooked up a noid light. Flashes during cranking.

Reinstalled the new ECM and old PROM. Buddy watching while I crank, LOTS of fuel. No change.

Had buddy watch rocker arms through oil fill cap while I cranked. The ones he could see all moved.

Re-probed injector circuit. No shorts to ground either wire. Full continuity between injector connector and respective pin on ECM connector.

Checked plug wires and oddly they were not on the correct posts as described in the service manual. Being a waste spark system they were obviously on there in a way that allowed the car to run. I never questioned it since the config worked but I set them to factory dictated location.

Reinstalled "good" ignition module / crank sensor / coil pack assembly. No change.

The engine was not started today. I unhooked the injector wiring to see if I could clear the flooded fuel at which point a stutter step backfire killed the starter. Just makes a weeeeeeeeeeeee sound to let me know it's pretty fucked up.

I'm now turning to hard liquor. But don't worry, if I go ballistic on this thing I'll at least get it on video.

[The Dark Side of Will]

Have an alternate PROM?

[Cheeze]

I do not have another PROM. Not even sure where to get one.

So get this. I just went back to work late tonight to load my car on the dolly and put it on the back yard for a while. I need to work on the Lemons car instead of the Fiero so away it goes for now. I grabbed the original starter since I never throw anything out hoping I could WOT start the car and drive it on the dolly. Easy change. I checked throttle and it goes to 100% actuating by hand, only 80% by pedal, 81% if I really mash it. All other parameters looked good as before. Clutch in, normal start. It sputters, almost dies, keeps running. Wtf, is this happening? I shut it off. Normal start #2. I do it 5 more times. Theories? F me. I'm loading it on the dolly now, more later.