Rough idle and misfire through intake, please help

[Rhutchinson35]

Hello everyone,
I have a 1983 944 with almost 69,000 miles on it. I was driving it home from work one evening when all the sudden the vehicle began to idle real rough, output hardly any power when given throttle, smelled really rich and backfired through the intake. That all started sometime early 10/2023 and since then I’ve replaced a handful of parts trying to chase the cause of my problems. This list of new parts/repairs are as follows:

-Adapt Motorsports fuel dampener and regulator
-Fuel Injectors Clean and Rebuild Service from Advanced Injector
-Rebuilt AFM via Only944
-Throttle Position Switch
-Distributor Cap and Rotor
-Spark Plugs
-Engine Coolant Temp Sensor

What’s keeping me stumped when cold, the car cranks right up and idles fine. But once it’s warmed up closer to operating temps, the problem reveals itself and the 944 falls flat on its face. I’m at the point now where I just want to cut ties and sell it if I can’t find the remedy but that’s proven to be difficult since not many people want to take on the challenge themselves nor can I find a suitable replacement car for myself. I hope someone with more 944 knowledge will be able to provide some assistance. I’d post videos but this is my first post and I don’t know how to post them. Anyways, thanks in advance!

[Eskozki]

Have you checked that your timing is OK? What is the condition of the plug wires? Vacuum leaks? Fuel pump is OK and delivers the correct pressure and flow at the rail? I've owned a 944 for only a couple of years though, but these would be the things I'd check at first.

[Jfrahm]

I've been reading that some of the DMEs (not the DME relay) are dying and causing odd running problems. And some used replacements have been bad also. Not saying that is your issue but it's possible. Also if your issue is when you go into closed-loop (if you have an O2 sensor) it could be the O2 sensor or related wiring.

I would love it if we could compile some testing data to see if the injector pulse looks typical. With a cheap DMM these days you can read Hz and Duty Cycle (and AC volts) at an injector and I wonder if that could be compared to a properly running car to determine if the DME is trying to fuel correctly or over/under fueling. Maybe someone who knows more than I do can help.

Maybe that's too difficult and people jsut need to send a spare DME for testing/repair or buy a Focus 9 DME or alternative, if there is one. Maybe the people who rebuild them know what to replace to make them very reliable going forward also, that seems to be the case with the 928.

In my 928s I have had a few ECUs die and it's just not worth it to try to drive around on the old ECU hoping it does not die. The 944 DME has been more reliable but maybe that's changing.

-Joel.

[wwdwgs]

Spark plug wire(s) could be bad. The other day my #2 wire started to short - loss power, rough idle etc.

[walfreyydo]

This guide is the best starting point- go through the diagnosis steps, dont just replace parts and guess:
https://www.clarks-garage.com/shop-manual/ts-01.htm

First step - whats your fuel pressure? Lets make sure the aftermarket FPR is putting out the correct pressure.

Another thing to check - vacuum leaks, idle air control valve (plug wires also a good suggestion). Could be DME computer but you should eliminate all other possibilities first before jumping to that conclusion, since its hard to test/diagnose without testing every other option first, and DME computers arent cheap. You would be surprised how a simple vacuum leak can affect these cars.

[Monkey Wrench]

Id pull the plugs and check the color, see if one is different from the rest. You might have a stuck open injector, even though they were renewed. symptoms of a stuck injector might be white smoke on startup since it can then empty a lot of the fuel rail into the one cylinder with each shutdown. If you pull the plugs under that condition you might find a wet plug or a plug that looks different from the others. If one injector sticks closed it'll basically run on 3 cylinders and run very poorly overall. I know they were renewed, i'd check the spray pattern visually with one removed and pressurized with a cleaner , if you suspect one is bad, that might not be the culprit or even the most likely. pulling the plugs to see what they look like , wont hurt anything or add cost. Id clean or swap them anyway just to be able to monitor and know its not old carbon from before. If they all look black from carbon ( rich) or white from running too lean it may give you more clues so I'd monitor them as you check other stuff out. if they look tan and equal then that would be pretty normal.
its probably not that, but . even a new injector might get plugged or stuck open if there is dirty gas debris in the fuel system.

i suspect if the ECUs ( or DME) fail the culprit may be bad capacitors, they look like tiny black tin cans. You might be able to buy new ones and solder them in if you are so inclined. Its a experiment, they cost about 4 bucks. Just watch that your replacements have the same working voltage and value in microfarads and don't turn them around they are polarity sensitive. I've found I can snip them on the topside of the board , up about 1/8 from the board leaving tails to solder to. a lot of modern as well as older electronics fail due to the capacitors themselves aging out and failing and you don't necessarily need to understand how it works down on the board level to just shotgun replace them. what Ive done is wrap the little lead from the new cap around a pin so it looks like a spring then you can crimp that over the stub you left, then solder it.

things to look for are bulging the cases (aluminum cans) due to building up pressure, also look at the color of the board. when the electrolytic caps leak they may cause a tell tail stain near them, of you are perceptive about that. on the others Ive seen there is a heat putty and so if you do remove the board from the heatsink you have to renew that but it might not be necessary to do that to replace the caps. I suspect all other parts , resistors diodes, chips etc are fairly reliable but caps do fail over time.. the boards are sensitive to static , you can wear an anti static wrist strap because your body can energize a static discharge and that can cause damage. the wrist strap is grounded preventing you body from building a static charge and you can improvise. or yea you can just buy a new ECU and they may want yours as a core. If you try this and it fails don't blame me ;-)

any electronics tech with a bench and a solder station can do this for you easily. troubleshooting the circuit itself is probably too overwhelming for most people., is for me. you can re-wet the solder joints to try to eliminate a cold solder joint issue. If you do any of this try to find a soldering iron that's appropriate in size heat etc, for this. the tip must be tinned properly and dont go holding the soldering iron on a long time or you'll heat stuff up too much.

I repair some old radios as a hobby , mostly from the 30s and 40's they are always full of bad capacitors, modern electronics have the same issues, the caps do commonly fail after 35 years and they are very cheap. the old radios i work on have a lot of old and failed paper / wax caps, these are usually bad, but in modern electronics they use different caps , mylar etc, they are reliable over time, just not the little tin cans. the tin cans have a arrow like stripe on one side that indicates negative. If you put them backwards they wont work and are likely to explode their guts.

generally speaking if they are little tin cans these are electrolytic capacitors and these are polarity specific, there are other capacitors that may look like orange cubes or little jelly beans , there are different types.. I' d leave those ones alone as they are more reliable. I'd only concern myself with the electrolytic capacitors.

you are probably looking at something like 47 micro-farads ( it will look like a U with a tail) and 16 WV ( means maximum working voltage) for example, replace with the same value you have in there.. you can test them with an ESR meter I'd just shotgun them all if you even choose to go there. this may or may not save your old ECU it is just an experiment so if you do any of this, prepare to loose.