84 Porsche 944 Cranks But Doesn't Start (No Spark)

[Gavan Parks]

A while back my 944 all of a sudden stopped wanting to start. It wouldn't even crank at all so we checked the battery which was alright but not great and replaced the starter solenoid. Now it cranks but won't start. We replaced the DME relay, battery and ignition coil but it still will not start. The coil is getting 12 volts as it should when the key is turned but we are getting no spark from any of the sparkplugs so we checked the ignition coil wire which is also getting no spark. There is a chip off one of the towers in the distributor cap (the one that connects the ignition coil wire) which could warrant replacement but if we're not getting spark from the wire itself, we don't think it will be the problem. I am thing of replacing the wire but I just wanted to see what anyone's thoughts on this was. Also update I checked the reference sensors which tested well. However, while the 944 wasn't starting I removed the power steering pump to replaced the seals and I took off the belt connected to it, I didn't see any warning to mark it and put it exactly where was but I thought I would add that.

[944 Ecology]

https://www.clarks-garage.com/shop-manual/ts-01.htm#fire

[Monkey Wrench]

So you said " The coil is getting 12 volts as it should when the key is turned but we are getting no spark from any of the sparkplugs so we checked the ignition coil wire which is also getting no spark."


yes the coil may see 12V with key turned but is it getting it's pulses? the primary coil should be seeing pulses of power, during cranking, not just constant DC, this creates the spark in the secondary coil, and on to the distributor/ spark plugs.

if you put a small watt bulb , like a dashboard light or similar across the coil there ( leave all else connected) and then crank, at cranking speed you should see some rapid flashing. or flickering..

if you start it and run it then the lamp may be flashing so fast it appears to be working and dimly lit because then the filament is basically not able to react fast enough to show much of a flicker..


I believe the missing piece here is that while the coil sees constant 12V, the ground is ( or should be) being interrupted by the DME electronics. if they fail to work due to...

( bad engine TDC or speed sensor or bad ground or bad 12 v hot to DME , wire break for example)

... the you may see no bulb flashing and hence no pulsed DC and as a result, no spark.
under normal operating circumstances the coil doesnt' "run on 12 V"

, rather, it depends upon the primary coil receiving pulsed DC..

it is the rapid DC pulses in the primary coil that in turn generate the higher voltage output in the secondary coil, ( high voltage pulses) by way of induction.

the secondary of the ignition col has many turns of finer wire so it is in fact functioning as a step-up transformer, increasing the voltage into the thousands of Volts to jump the air gap of the spark plug. ( higher voltages can jump larger air gaps)

I dont know why an "ignition coil" is referred to as a "coil" when in fact it ins not a coil but a transformer. You won't change that terminology, because the word "coil" is so ingrained..


between the MCE - electronics and the ignition coil there is a little amplifier, for want of better words.. a transistor..

it can fail but they seem pretty stable. I think in laymen's terms, it is using the low voltage pulsed signal from the MCE board and transforming it into a pulsed ground.. so the coil is fed a hot and a pulsed ground and the result is the primary of the coil is seeing a pulsed DC current, then the secondary sees a higher voltage pulsed DC, thats the output to the distributor..

the distributor isn't changing the pulses , it is directing them to the appropriate plug lead at the right time. the only time a spark can be transferred to a plug lead is when the distributor's rotor points to it. the distributor is a mechanical device.

on older cars with points, the distributor was also used to set up the advance curve. i affected the dwell timing and engine timing. it contained springs and weights, that functionality was mainly replaced by electronics

id determine if you have a pulse at the coil when cranking, if you do not then look back towards why the MCE board may not be generating the pulse,,,

If the coil has a pulse in the primary it is likely getting to the plugs.. yes you could have a damaged cap.. normally giving the electrodes inside the cap a scrape will get them working otherwise if it's made of junk materials just replace it..

always have a good look at the cap , clean it wiht alcohol look for cracks, any signs of tracking . look if the plug wire sockets are all green inside.. etc. check the plug wires, sometimes they sort of fall apart. try not to yank the cables out of the molded ends.. there are some real junk import plug wires out there.. OEM is nice, but at least inspect what you have there..

[Gavan Parks]

I tried your test and it seemed the bulb dimmed when I cranked it. This means its working well and the problem is most likely with the coil wire or dist cap?

[Monkey Wrench]

during cranking the bulb should flicker. if it starts it will appear dim because it is flashing rapidly enough and there is some warm up / cool down time for the filament so it will appear as a dim bulb.

the coil is actually a transformer, two coils, the low tension coil is 12 V and should be seeing pulsations. the pulsations come from the ( computer) and go through, basically a transistor it will be around the front left headlight area.. then to the coil.

the nearby high tension coil picks up the power by the wires being wound near to each other without contact, that's called induction.

each time the primary coil is closed ( turned off) there is a high voltage current induced intot he secondary coil , that current is used to spark the plugs.

the HV power is sent from the coil through the distributor and depending where the rotor is pointed, through the appropriate plug lead then it sparks as it jumps the plug electrode to the block which is ground.

if you hooked a small watt bulb up across the coil terminals ( the 12 V wires) what you have there is the coil is still connected as normal and the bulb is also connected, the two are in parallel, when the coil gets a momentary impulse of power, so does the bulb.

the reson I suggest doing that is to see if you have a flickering bulb when cranking, if you do then you know you do have a pulse to the coil. as long as you have that then it means that the DME computer is sending out those pulses and that proves that the engine sensors are working and the DME is doing that part of its job.. ok so knowing that you can stop troubleshooting engine sensors and the DME.. well sort of.. If you failed to see any pulse then we'd go back and loook at what the problem could be that caused it to not have a spark.

what we are doing here is breaking the problem into bite sized easier to understand chunks and by a process of elimination narrowing the possibilities of where the fault lies. or at least trying to.

often it is advised to pull off a plug wire hook it up to a spark plug and ground the body of the plug then crank the engine and observe the plug's spark, this also is a way of proving it has a spark. I find that awkward as the plug is a long way from the key and if you get someone to hold the plug for you they can receive a surprising jolt , its unlikely to hurt them unless maybe they have a pacemaker or some underlying condition but sometimes a surprise jolt like that might cause an injury due to the person suddenly pulling away..

a different way is to connect a timing light and tape the button down then you can see the light flash from the key switch.

[Monkey Wrench]

"I removed the power steering pump to replaced the seals and I took off the belt connected to it, I didn't see any warning to mark it and put it exactly where was"

no there is no "timing between the power steering pump belt and anything , all its doing is turning the pump. the timing belt and the balancer belt, yes they have timing. those are toothed belts. the teeth keep the belt in time, it can't slip, ( unless something is horribly wrong)
similar to a chain. V belts are not timing belts. you can recheck the tension, rule of thumb is you can turn the belt 1/4 turn without stressing your fingers. too tight loads up the bearings, might cause the belt to run crooked. when looking at belts youcna put a straight edge between the pulleys, try to "eyeball" and observe you don't have a situation where the belt and pulleys are not in line or that will cause noise, fast belt wear. sometimes , at leas twiht some cars, there is a rubber bushing or similar that goes egg shaped from wear and cause crooked belt to pulley geometry.
I've heard of that belt getting too close and damaging rad hoses, maybe have a look to make sure there is adequate space there. If its not near the rad hose I might just be talking about a different belt.

[Monkey Wrench]

the other job of the DME computeris to give impulses to the fuel injectors, the fuel injectors open when they see the very short duration impulse, releasing fuel. as the car needs or wants more fuel the DME increases the duration ( sometimes called dwell) of the impulse, that injects more fuel by holding them open just a microsecond (roughly speaking) or so more..

if you have no spark your car wont start. if you have no fuel your car wont start. something needs to be mechanically wrong to not have compression.

I suggest we work at trying to be a bot more certain if you have a pulse. yes the spark plug leads or a bad cap could possibly cause a no start condition... the spark can jump to ground so if plug wires are wet or dirty or worn then that could happen. sometimes people will crank thecar with the lights down low so they can more easily observe a problem like a spark plug lead being grounded..

sometimes a bad or dirty cap can have the spark following a carbon track and to ground within the distributor.. if you look inside the cap watch for any signs like black marks indicating arcing. If you thought maybe the spark was being grounded within the distributor you can test for that,,

connect a plug lead to the coil and then to the plug or timing light and do the same test for spark without the distributor even being connected.. then its removed from the picture.

Id like to be a bit more certain whether or not you have a spark, " the light is dim " isn't really proof .

lets say for example you have a bad spark plug lead, well there are 3 more so engine should fire up, might run rough but should fire, probably will start.

its not impossible but I dont think the coil or distributor or leads are the most likely issue. normally I would replace the cap and wires on any older car, keep the old ones for spares and if I wanted to test, then just swap back.

look inside the cap at the 4 electrodes. are they copper , brass or a white metal like aluminum.
they will get corroded from the ark jumping from the end of the rotor to the electrode. an OEM cap will be brass or copper, there is a lot of new junk out there where they are some sort of white metal.. they dont last very long.. all the same scrape , clean the electrodes, end of the rotor too.

take a long sharp knife, try to scratch the electrodes to clean them of any buildup if they re corroded. try not to scratch the cap itself. too much buildup will cause problems, probably weak performance and poor running.

are we sure we have a spark? or sure we don't have a spark?

if you think it does have a spark and it wont start then I'd try spraying some quick start in , remove air cleaner give it a shot, close up air cleaner... then try..

If you find that it will start by using quick start, but then runs and dies out, then Id suspect its not getting fuel. if it runs you also know it has a spark.

if you think it has no spark then the troubleshooting turns to why does it have no spark thats one path..

If quick start proves it starts and then dies then it probably has no fuel.. it doe have a spark, in that case we are going towards the road of ascertaining why it has no fuel or a lack of fuel.

At this point we need more concrete evidence of if it does or does not have a spark at the plugs.

Im sure all this could be worded better i know sometimes my posts are too long and too wordy, I'm trying not to lead you in too many directions at once and confuse things.


Clark's garage will have a troubleshooting sequence for a no start condition. Id also suggest reading that .. It may take a bit different sequence but you are basically trying to eliminate possibilities in order to get closer to proving the cause. throwing money into buying a new coil or things like that is likely to just make you more poor and not solve anything , what we want to do is prove which part is bad.

I know how tempting it can be to throw money into parts and hope, but really want you need is more proof of the exact issue, that's troubleshooting. The "shotgun method" of guessing and ordering and replacing random parts is often not the best way to solve issues. that can work a lot more economically if you have a spare car handy to borrow parts from..

at this point I'd try to make you think:

from what you know what do you think the problem might be?
what possibilities do you see?
what possibilities have we eliminated?
do you have a spark? does it come from the coil? does it reach the plugs?

[Gavan Parks]

I have tested all the spark plug leads with an inline spark tester and none have spark which I don't think they would all go out at the same time so I think the problem is father down. Just incase I disconnected one of the injector as sometimes they go bad and overload in which disconnecting one of the injectors would fix it and the car would start which was not the case for me. I ordered a new coil lead and distributor cap just to eliminate anything in front of the coil since the distributor cap had a chip off one of the towers which would warrant replacement. I also wanted to check for grounding issues so I went to the ground locations and brushed them off although none were very corroded or bad.

[walfreyydo]

Just follow the testing procedures that George posted above:
https://www.clarks-garage.com/shop-manual/ign-04.htm

Read it thoroughly, and follow the procedures. Dont deviate or go off guessing (injectors have nothing to do with lack of spark). Be systematic by following the troubleshooting steps. Test FIRST before ordering parts or you are just going to waste your money and time...

[Jfrahm]

Tyler has a good point but I am not wild about the coil testing procedure in Clark's.
The test lamp dimming doesn't sound right to me, you should see distinct negative pulses to the coil. I recommend firing the coil by hand.

In this guide you can see how to access the coil and related wiring. If you remove the signal side (green) and hook up your spark checker to the high voltage lead, then with the ignition on tap the signal-side terminal on the coil you should get a spark. Be careful as you might get a zap if you do this bare-handed, high voltage sometimes does whatever it wants.
https://www.pelicanparts.com/techarticles/Porsche_944_Turbo/31-ENGINE-Coil_Testing_and_Replacement/31-ENGINE-Coil_Testing_and_Replacement.htm

Just tap the terminal with a grounded tool like you are hitting a drum. Wear a glove.
If you do not get a spark, verify +12v on the other coil terminal and test the coil with an ohmmeter as per the above-linked guide.

If the coil produces a nice fat blue spark that is 10mm or more in length, we move on to checking the negative pulses to the coil.

-Joel.

[Monkey Wrench]

If the coil has a pulse and no output then yes it could be a bad coil but I would put my money on the problem being elsewhere.
I'd clean the threaded posts if thre coil has threaded posts, are they rusty? Ive seen rust there cause a connection issue personally and observations in troubleshooting

If the coil has no pulse to the primary coil then you need to go troubleshooting back towards the ECU. ( DME)

there were some previous threads showing how to do some checks with a scope.. the trick with a little low watt test lamp across the coil seemed to give you an indication of a dull lamp and not a pulsation during cranking. that may happen if it's a larger bulb , maybe the filament takes longer to heat and cool and I dont know how large of a bulb you used..
If you do not have pulsating power there then need to find out why but I'm unclear of the result. You can check that with an O-scope and then with that, you can also check back further towards the DME.. Teh electronics gets info from the speed sensor and TDC sensor you can go into analizing that trace with a scope.. there are some older threads showing this with photos of the screen of the scope, they may help if you get into all that. you might check resistance through the sensors to make sure they aren't open circuit. bad/ broken wire, bad plug etc..

those sensors need to read the flywheel, or a pin on it,, the scope can verify, some will swap out those sensors for reliability. they may be both the same part, you can check .. the air gap between the sensor and what it's sensing can be out of range..

oil and vibration may get to that wiring to the sensors so you might have wire fatigue.

when you crank do you get a tach pulse? this indicates a impulse from those timing sensors. look if the needle on the tach is moving when cranking. it's a sign.

some issues have been caused by things like bad ground or bad hot wire to the DME , If you wanted to check that you could perhaps run a ( fused) jumper to quickly prove , bad ground, bad hot, etc.. do not get the polarity mixed up. blowing that up will be expensive so be careful there.

a thing you can sometimes do is search for threads with similar issues then try to understand the resolution.

If you follow Tyler's advice above the you are at least using a more methodical approach by following the steps already laid out in a logical order.. That's probably good advice.. If you have particular questions you can ask along the way. If I wrote too much and confused the thread or your sequence of troubleshooting , then my apologies.

I could go on and maybe not help by just making the post have too many words to be helpful. ...

[Gavan Parks]

When I first had this problem I replaced the coil as I had no idea what to check and test so I don't believe the problem lies with the coil itself but rather something in front or behind it. The lamp tester I have I dont think is the highest quality perhaps why it only dimmed instead of flickering I think its just a 12 volt circuit tester. Perhaps I was a little hasty buying a new coil and dist cap but its already on the way so I will wait to install those and see if it works or not. By the way we had tried spraying quick start in and once it came quite close to turning over however never did. The coil lead was pretty beat and the dist cap was damaged so they would be needing replacement soon either way so I don't mind replacing them. The speed and reference sensors test well and were replaced less than a year ago I'm not sure if I mentioned that. If the lead and dist cap don't work just to clear any confusion you think I should check the DME? From what I've seen on clark garage there no real way to test it except having another one. I was also told by an old porsche mechanic to check to make sure the rotor is actually turning and you said to clean the end of that as well so I will try to do that today.

[Jfrahm]

You should see distinct flickering of the test light if you have it hooked up correctly:
https://youtu.be/mDKj19Hnhd0?t=104

[Monkey Wrench]

if you have the plugs out you can turn it until it "breathes out " from #1 some put a wadded up ball of paper towel inthe plug hole so they see it pop out.. or just put your thumb or something over the plug hole..

when #1 is expelling air it is on it's compression stroke, keep going (normal rotation direction) until the crank is at TDC, now you are arrive at the firing position for cyl 1
pop the cap , the rotor should point to plug lead # i, next it will arrive at , 3 , 4. and 2 as per firing order, each 1/2 turn of the crank so you can make sure the plugs wires arent mixed up. if you turn the engine, in its normal direction, you'll see which direction your rotor is going in.

you can unplug those sensors and check the resistance through them, there should be some , not shorted , not open, Ill guess a thousand or 1500 ohms but that's only just a guess.

an oscilloscope can be used to check these pulsations and that will give a graphic represntation of the pulse, keep in mind you could have a pulse that is present ( detectable) but not strong enough.

someone here had the idea of using a radio instead of a scope, If you have a common ghetto blaster that has inputs for a mic and you feed into that then you'd hear the pulse, the radio's input is pretty sensitive. It may be easier to know what is normal , if you have another car to compare to.

those sensors are merely a coil of wire with an iron core, when they pass by something made of metal the metal induces a pulse into the coil, which runs throuhg the wires and they are interpreted by the computer. If you have one out of the car you can try to connect the two leads to a meter, you'll see a short pulse of maybe a volt or two if you wag a screwdriver near the end. ..


the air gap can be important, Id guess 2 mm but someone may have the spec..

I think if you remove the starter and look in around the ring gear you' ll find a pin that the sensor reads. it' ll be near the ring gear.

I'm not sure on the porsche but in my volvo I just took a vernier caliper and used that to measure the depth of the hole then measured the length of the sensor to it's "land" and subtracted to give me the gap..

ive heard that one problem is with the clutches on the 944 being rubber centered rather than having the normal clutch plate springs. after 35 years the rubber degrades so if you remove the starter you could find rubber debris from that.

there is some means of adjusting that gap, I'm not sure if you physically move the sensor or if you add a spacer collar under it but that gap could be off, it is possible.

between the computer and the distributor in the wiring there is a little "amplifier" for want of better words.. it will be stuck to the body of the car and have some of that heat putty to transfer its heat away. It could fail but Ive never seen a bad one, a Volvo uses the same part. you can check and clean those connections.. I think 3 wires run to that and it will be connected with the coil.

the computer or "dme" is running at 12 V it is doing the sensing and it has an output, this output is not carrying or involving much current. It's a signal but not a strong pulse..


I initially figured this little part was amplifying the signal but I believe it is actually on the negative side of the circuit of the coil.. so I believe it is sort of making and breaking the ground, not the hot side.. basically it is using that signal , output from the MCE and transfering it to quickly connect and disconnect the coil ground.. the coil is taking a higher level of amps, so this part is necessary.

I would not order a new computer. i have an older scope myself, I belong to a club of old radio restorers and they often show up at our swap meets at not much money of you can probably find some cheapo ones. if you have an old windows computer with a soundcard there are oscilloscope programs you can load and use the soundcard as an Input. Ive tried that and it worked really well.

some of the more sophisticated meters have some cope - like abilties. If you started probing with an older analog meter you may see a pulse by way of the needle wagging in time wiht the sensor passing by its trigger, every once around. some digital meters have a bar graph function to mimic a an analog meter ( one with a needle)

with the scope you can see the shape of the waveform and it isn't exactly just a sine wave.. it can also "trigger" so that you can sort of zone in on a much shorter lived pulse than you can detect with a meter. at first a scope seems a but daunting with all its buttos and switches, they are not really very ard to use. You might get someone local to help a bit with that to help learn or watch some videos on usina an oscilloscope..

I never minded buying a few tools if I thought I could use them as I'm getting older I buy a lot less but Ive never regretted expanding my capabilities so personally I'd rather "waste money" on tools than parts that I don't need..


more modern scopes used for electronic analysis may be more than you need... I'd say pretty much any old working electronic ocilloscope is ok for this sort of testing of the sensors and for looking at the pulse.. so I would not go breaking the bank to buy the best, id look used myself or see how much you can get a relatively cheap one for on amazon or similar. first one I bought was 40 bucks ran on tubes not IC"S and it was so huge I could barely lift it,, cool but pretty outdated, still worked ok.

I think you areon the path of a good learnign experience and would say if you have the time and are not stuck just be patient and try to check as much as you can from the clark's garage site, try to refrain from going broke trying to solve it..

maybe a high school or similar has some electronics program nearby, often people will help if you ask nicely and just say you need a bit of help and try to print out any info that is helpful..maybe bargain a case of beer for an hour's help or similar, if people feel you are trying your honest best they will usually help more

Having someone lean over your shoulder can really help a lot. the way I see it you are trying , dont know everything, like most of us, so some assistance can really help lots and that's how we learn and next time you are that much further ahead. Many will give up and not learn anything and then some can't even check their oil , they pay a dealer to pull the plug and screw on a filter, for life.. i really do think that the determination and patience pays off, because then you gain knowledge.. that knowledge is transferable and that's exactly how you loose the fear of taking things on that you don' t quite understand and then master them.

YOU tube is a huge help nowadays I bet if you spend time on things like "how to troubleshoot automotive sensors with an oscilloscope" you'll find stuff on that that will help .