Starts but won't stay running HELP

[skittleman12]

Hello everyone,
I couple of months ago purchased a 1984 944 NA non-running from a mechanic who got tired of trying to fix it. The previous real owner had the clutch, water pump, oil pump, and timing belt replaced but never paid for it. Then it got passed around to a few mechanics, and now me. At the time of purchase, there were several things wrong:
1: The car had no tach bounce and only cranked. Upon inspection, I found that the car had new reference sensors, but they weren't gapped correctly. I gapped them using a 3d printed tool and then I had tach bounce!
2: The previous owner jumped the alarm incorrectly, I made a jumper to correctly bypass it.
3: The distributor cap was cracked - I replaced it
4: One of the spark plug wires had zero continuity - I replaced the set

After all this, the car would try to start but only get to around 300 rpm and would sometimes stumble along at 300 before still dying.

I removed the accessory belts and it would get up to 1100 rpm before immediately stalling. However sometimes if you got the gas just right, it would roar to life and run for as long as you changed the throttle position. If you stopped giving it gas OR gave it a consistent amount of gas, it would quickly die.

To try and troubleshoot this I tested and replaced several parts: (ended up adopting the "throw parts at it" method at one point)

Fuel related:
Fuel damper - replaced
Fuel pressure regulator - replaced
Fuel pressure - 30psi +
(the last guy replaced the fuel pump and fuel filters)
Gas - The last guy flushed the tank and put half a tank in
Throttle body - fully cleaned out the throttle body and can now hear the switch click

Electrical:
DME relay - replaced
DME - Tested good on another 944
Spark plugs - replaced
Battery - replaced by the last guy
Ignition switch - replaced
Tested MAF sensor via Clark's garage - tests good
Tested Throttle Switch via Clark's garage - tests good
Tested DME temp sensor - tested good but I broke it when removing so replaced it
Starter - replaced by the last guy
Ignition coil - replaced by the last guy
Alternator - tested good

Vacuum related:
ALL Vacuum hoses and intake gaskets - replaced
Tested for vacuum leaks by having a friend smoke into the sealed intake - no leaks
Tested for compression - 170-180 psi for all cylinders
Intake valve (the one for AC) - tested good
Idle air control valve (the one under the intake) - stuck closed so I bypassed it because it should be open when cold

Other:
Checked timing - All marks line up perfectly.

At this point (2 months later) I understand why the previous guy gave me such a good deal on the car. It's so frustratingly close to running, I've gone through everything I can think of to possibly fix this, read every forum I can find on no starts, and no luck.

If you have any ideas or questions I'll appreciate it!! Thank you all for your time!!!

[JoshB]

Injectors - are they all working? Are you absolutely positive you have the plug wires in the right order?

[kdjones2000]

You bypassed the AAV. It only has a small hole in it to allow more air in when hot. I'll bet you are lean b/c you bypassed it.

[skittleman12]

Thank you guys for responding so fast!
Yes, I tested the injectors using 4 clear glasses, they all had even spray patterns and equal amounts of fuel delivered.

Is the AAV supposed to be open when it's hot? Or closed? I was under the impression that it allows more air to cold start, then closes as the engine gets hot. The valve I took off was in the closed position when cold, only allowing a tiny amount of air could get through, I'll add a picture here.

[walfreyydo]

You didnt mention what year your car is (assuming its a 944).

This is important because the idle air control valve works differently on a late model 85.5+ (oval dash) vs an early model (square dash). Mainly, in late models, the IACV is used to moderate idle speed, and is variable (hence a non-working valve may not allow the car to idle properly). Early cars have a binary on/off valve.

So if the valve isnt working correctly, and you have a car with a variable valve, that may be your issue: https://www.clarks-garage.com/shop-manual/fuel-21.htm

Testing (some good info in this thread) https://forums.pelicanparts.com/porsche-924-944-968-technical-forum/481223-idle-control-valve-testing-procedure.html

ICV may need to be removed and clean if it fails the test (does not actuate/click open closed with voltage) - usually gunk buildup which can restore function after cleaning with some carb/brake cleaner.

Not saying thats your issue but it does seem to be a fueling/air-fuel ratio/vacuum leak issue and the IACV is certainly related.

Also consider that the IACV can develop leaks and bypass air, so another thing to check (a little about that is in this thread: https://forums.pelicanparts.com/porsche-944-turbo-turbo-s/330312-idle-valve.html)

[wwdwgs]

How about "an idle screw" on Throttle body? Did you play with it?



There's also an idle-speed adjuster (under the intake 930 606 161 00 or similar). Could it be faulty and let more air in? Or its connector?

[skittleman12]

It's a 1984, so it uses the slowly moving binary on/off valve for the iacv. That is what I currently have bypassed because it was closed when cold. (the car didn't have any changes in symptoms after bypassing it)

I will try the idle air screw as soon as the temp sensor arrives! I haven't tried that one yet!

[wwdwgs]

I know that this would sound stupid - are all ground points clean?

[Monkey Wrench]

I had similar problems once when I took a car all apart and had had the flex plate wrong. In that car it's what the TDC reference sensor reads. the symptom was it would start and not run higher than 800 RPM. the fix was very straightforward, all I needed to do was pull the engine or tranny back out and turn the flex plate one bolt hole :-)


on a 944 It might read the flywheel teeth or something in that area. Id doublescheck whatever it is the TDC and speed sensors reference

I bought a car with a similar issue , the former owner had taken the dstributor apart and reassembled it wrong, they basically had it looking correct but 180 degrees out of time.

try this :
put your finger over the spark plug hole and continue turning until it is "exhaling out that plug hole for # 1 and not far from there, keep turning slowly to put the engine to firing position with Cyl 1 at the top of it's stroke,, you can feel with a drinking straw for the piston rising or do similar..

by doing the above, you are putting the engine to near to it's firing position for cyl #1

look at the distributor cap, you should see the rotor pointing to the wire that leads to plug # 1 Beyond that it should turn anti clockwise 1432 ( I think) check that the other plug wires correspond to the firing order and of course you know that the rotor will turn in the normal direction. I think it's CCW but correct me if Im off on that.

On some cars they have places that the plates in the distributor bolt to the case and they may be moved. some cars ahve a pin that locates the gear to the distributor shaft, Ive seem people drive the pin in the shaft with the gear 180 degrees out, its hard to notice that's wrong.
some cars can be timed wrong by the incorrect mesh of the distributor gear to camshaft, that might be a bit different in a 944 but just making sure it's pointing to cyl 1 at "firing time" for cyl 1 I think that is the easiest way to think about it..
. this verifies if some guy made a goof up and reassembled the distributor incorrectly. I know porsche may have some unusual arrangement with a belt driving it , or similar.. the broken cap might be a clue to someone mucking about with the distributor.

if you have it running even if not running right, you cna pull one plug wire off at a time, If you do find one that isnt' causing any difference, thats a clue.

if you put a small low watt bulb across the primary of the distributor you will see it flickering. the primary charges and falls, so the test lamp will also, the reason I like this is you can keep an eye on it when having trouble starting , if something is interrupting the ignition pulse you will see it stop flashing, so long as the test bulb is flickering your coil is getting its pulses on the primary coil. from there you may assume it most likely does have a spark. If you do see problems maintaining a constant pulse to the primary coil then you cna dig back through the powerstage transistor cna check MCU grounding voltage etc. my theory is that is just connecting a test lamp and if you do have a pulating primary coil, then its sort of proves that's not the cause of the starting issue. then maybe you can avoid a bunch of other troubleshooting because you know the coil has it's pulses.

easy thing to overlook but make sure the battery and cable connections there are ok. you can have an issue simply because you slipped the cable on the battery but left one of the little wires there a bit loose. Ive seen it where I had the battery post tight on the battery but one of those small wires that also goes to the battery post was still loose, causing a no start.

pull your meter out, black probe to ground check each fuse, mark down which is hot, check n=both sides. if one fuse has a bad contact or is blown you ll find it. If things look corroded, fuse are cheap, you can pull the battery post and clean all the fuse holders up really well , you might pull the entire fusetray assembly away a little if it helps access. Id make a map of all my fuses before I get confused. pics are good.
do a ittle chart showing each of the fuses, which is hot with key out, key in, and each key position. You can put yoru little "chart" in the back of your shop manual. for quick reference later.

when it is running OK you can try again.
the thing is that if you have reference written down, you can always pull out your meter and expect just the same readings on each end of each fuse in each key position. if things differ from that point it's a quick way to find out and then look deeper. what I'm saying is it's nice to " KNOW what is Normal" and have a quick reference to that, so you can easily find it when it turns " not normal." without consulting manuals, reading schematics , getting confused etc.


also ifyou have power at one end o f afuse, then the other end of the same fuse should be hot and the wires connected, ( nearby spade trminals) If not, you just found a bad contact by using your meter.

there is a reference to the voltage drop across a fuse, and it's load in amps , but that's getting more complex.
for now you are just checking for intermittent, bad spade connection , bad fuse, fuse not making good contact etc.

It sounds like a situation where someone cornered themself, If you go troubleshooting and don't solve the issue and then create more issues you can have a car that wont start for 3 different reasons, at that point digging your way back out of it becomes a more elaborate challenge.


pull the wires out of your cap and check in there, sometimes I find a drop of water got in where the wire enters the cap and was reinstalled wet..
and then you have a green connection, that is a bad connection.

It happens. some of the caps have aluminum electrodes, generally the garbage Chinese ones. if it's OEM probably copper or brass.. aluminum ones oxidize fast. whatever it is you can often just clean the cap and rotor. Yours was cracked, why cracked? poor installation? I like to wash caps in alcohol to degrease them. scraping them with a pointy knife gets more life until they are too pitted.

[wwdwgs]

I second Monkey Wrench regarding the materials of Dirst. Cap and the rotor - exclude alum.
Also, just to be sure that the parts is not damaged, the rotor is held be a driver, which is keyed and goes over the Camshaft.

Its hexagonal sides can mislead into thinking that it's a threaded part and a mechanic could be tempted to "turn" it. I was, and I stripped internal keyway and had to buy a new part with a key. Fortunately, the keyway on the shaft was in perfect condition (harder material).
The keyway is shown at "4 o-clock" position.
https://cdn4.pelicanparts.com/catalog/images/SuperStock/944-105-142-05-OEM_1.jpg

Another thing I learned is to test wiring with a test light coupled with a small 12V (from electric scooter) battery. Multimeter may show a continuity, but wiring may be corroded to the point that the real load won't be "passed along" the wires. This tactic helped me to resolve a wiring issue on some other car - the multimeter showed a continuity, but test light won't and won't activate a solenoid (in my case).

I know that your situation is frustrating, I spent about TWO years to troubleshoot some problems (other car), which indicated that my problems were with grounding and wiring (corrosion). I had to "fix" 3 separate systems (wiring) in the process.

Good thing that these "old" Porsches are not that complicated, compared to the newer cars.

Your favorite beverages would help, too!

[wwdwgs]

Watch this video, if you haven't yet:
https://www.youtube.com/watch?v=1WuZuPcVuqM

[skittleman12]

Hello again, I'm back from my holiday trip and just got done trying to fix the 944 again.

Today I:
- Installed the new temp sensor (I broke the last one when inspecting)
- adjusted the idle screw on the throttle body (I didn't seem to change much, however, it did have a harder time running when turned more than a few times in either direction.)
- Adjusted the MAF tension (it also didn't seem to make any real difference to the no-start)
- Sanded down all the grounds

The car is seemingly no different than before, at this point, I feel like I've checked just about every probable system on the car and I'm awfully close to taking it to the shop (AKA giving up after 2 months)

What do you guys think? What else could it possibly be...

[kdjones2000]

Get the DME Test Plan and actually do some trouble shooting by measuring values at the DME connector, rather than throwing parts at it.

If everything checks out, take it to a pro...

[commdiver]

It sounds like your cylinder head has a bent valve or two unfortunately. I would do a compression check on all four cylinders to confirm as if there are bent valves from previous mechanics or owners then it’s never going to have compression in all 4 holes and never run and could be why it was brought in for a timing belt and water pump.

[walfreyydo]

I still wonder about vacuum leak. You can get a vacuum gauge and see how much vacuum it pulls when running and report back.

You say you adjusted MAF tension. Not sure what that means (these cars dont have a MAF, they have an AFM). The AFM is very very sensitive and you shouldnt be "adjusting" any tension in the AFM unless its out of calibration. Was that checked? Or did you just start messing with it without testing? Typically the only adjustment needed on the AFM are the contact pickup points slightly moved

I would recommend testing and making sure you are hitting the max voltage. If not, then your tension adjustments need to be reverted/recalibrated.

Here is the Clarks guide on voltage testing and retracking the contact points (unfortunately Clarks doesnt say what the specific max/min voltages are so youll need to dig that up somewhere else online I think):
https://www.clarks-garage.com/shop-manual/elect-22.htm

Changing the tension on the barn door in the AFM means you changed the calibration. These things sense position of the barn door as air moves past and its calibrated to send a voltage to the DME which tells it precisely how much air is coming into the engine. Changing that tension means that the voltage curve is now misaligned to the amount of air flowing past that barn door - ie: it now takes more/less air to reach a certain barn door position. The DME thinks a certain amount of air is coming into the engine, and providing fueling based on that. Now you changed that (not sure why) and probably created compounding issues (ie: hypothetically vacuum leak plus AFM mis calibration now). Again, you typically should never need to do this, only retrack the arms/contact points (as outlined clearly on Clarks). And if you ARE going to change the tension, it must be done in careful coordination with a multimeter making sure you hit the max/min voltage values as outlined on clarks - so you must now recalibrate.

[Monkey Wrench]

that's a lot of great info !
Tyler would you recommend using a spray, like deoxit, on the wiper?
It sounds like this voltage output from the AFM is being provided to the DME and that may be a sensitive circuit..

in fixing old radios I come across old test gear, and the item I'm suggesting is a vacuum tube voltmeter. the way this differentiates from other analogue or digital meters is that they don't load down the circuit being tested. Using a scope might produce a similar result. the ones I see appear like a analogue meter ( with a needle) and are bench style equipment..

I'm wondering if it's true than that this variable voltage which is being supplied to the DME may be measured with a scope or a VTVM, when the car is running, and it may be the case that if you go and connect your typical digital voltmeter the load of the meter itself may upset the circuit resulting in bad readings and an unrealistic mixture while it is connected.

I dont know if this last point is true but it may be the case?

someone smarter than me with a better electronics background might have a way to do this with a digital multimeter, perhaps by using a 1 meg resistor to prevent loading of the circuit through the meter..

a 1 meg resistor presents such resistance that the load ( through the meter) may be reduced, this is because the 1 meg resistor will conduct the voltage and a very small amount of current but it's hard to "drain" the circuit through such a high value resistor as it is basically blocking most of the amperage from passing..

It MAY BE TRUE that adding a 1meg resistor to your meter probe prevents the meter from taking a large amount of current and upsetting the operation of the sensitive DME circuit..

if this is true maybe we can establish a voltage which is reasonable or compare vehicle to vehicle to know what a normal voltage upon startup should be, or during idling ? maybe at Wide open throttle with the car standing still?

[Bulldog9]

Looking at all you have done, I would say you have a major vacuum leak or your AFM is fubar.

[walfreyydo]

Originally Posted by Monkey Wrench

that's a lot of great info !
Tyler would you recommend using a spray, like deoxit, on the wiper?

I would recommend what clarks garage recommends - which is to slightly bend the pickup so the contact points are re-positioned slightly, but only after a voltage test, OR experienced problems with the car - such as stuttering and hiccups (a common symptom when the AFM contacts go bad).

But since the AFM tension/calibration has been changed, this may not fix the OP's issue since now the voltage to airflow readings are going to be off

[Monkey Wrench]

the logical solution is to buy a second 944 so you can swap known good parts ;-)