Help needed: Hard, Slow Cranking issue with my 951.
 

Having a hard start problem with my 1986 944T (2.5). Engine completely rebuilt, running a pre-Lindsey Rogue M-Tune (80# injectors, 3-BAR FPR, pump gas), early Rogue Tuning DME. Intermittent (90% of the time) very hard/slow cranking when starting up. I checked all of the battery cables, wires to starter and grounds and they are good (voltage drop, continuity). Starter/solenoid and battery are good/newer. Checked speed and reference sensors ohms at the DME and they are good (gapped with Dave Gunderson’s tool). Replaced the ignition switch a few years ago.

I removed the fuel pump fuse and pulled the coil wire while cranking and testing the wiring. Without fuel and spark, the hard cranking disappeared. So, I’m thinking it’s either an issue with my DME or fuel system. Despite the hard cranking, the engine starts, and it runs very smooth. Timing mark at TDC is spot-on and the AFM looks good when running according to my wide band.

Today, I ran more tests to confirm what I am seeing: I purchased a new battery and ran a few more voltage tests. My other battery was still in decent shape but was several years old. So far, I have tested both the battery and the positive cable while cranking. With removal of Spark and fuel, while cranking, I am only seeing about a one volt drop at the battery. No big deal. When I add just spark, I get the same result. When I test with fuel and no spark, same result. Drop is to about 11 volts while cranking and it turns over fast. When I add fuel and spark, I am getting a large drop at the battery (new) to 8 volts and very slow and hard cranking. But it doesn't happen every time I crank it. Probably 90% of the time or better it cranks hard. I did a drop test on the positive cable while cranking using the same parameters as above. No issue until I add fuel and spark. I am getting .90v drop at the starter positive wire and slow, hard cranking, which I believe is no good. No fuel and or spark and I am getting a .18 V drop at the starter with easy cranking. I tried different FQS positions, thinking that timing might be an issue but no change. Any help would be appreciated.

strange symptoms !
I'm thinking it might be trying to fire too far before TDC and maybe predetonation is causing the firing to put more strain on the starter? slowing the starter? that or there is some massive short, but if there is a short that big it must be producing heat and nothhing seems to be frying up.

does placing a jumper cable between the battery - and the engine block make any change? or (carefully ) between battery + and starter positive cable?

I was recently fixing an AC brush motor , it ran intermittently and I was looking for the cause. and realized it had issues due to some grease on the brushes. they looked ok but just as an experiment , I took the brushes to a grinder and took a touch off them , and it was solved.. they file easily as they are just graphite. ,,

a problem with cables or ground does not explain why you are getting different results when you have the firing conditions and not when you remove fuel or spark.

I'm mystified. maybe someone else can reason their way through this one..

i wonder what happens when you have fuel and spark and everything connected except unplug the spark plugs, or maybe as a second test, with plugs removed?

removing the plugs should make it crank fast as there will be no compression and if you can find it slowing under that condition then you have eliminated predetonation as a culrit.. so then it would need to be to do with a short or bad connection then..

if the fuel rail lost its charge by a leaky check valve it might make it need to crank a bit more to establish fuel pressure before it catches.

timing belts didnt slip right? I'd make sure.. just dont want to hear of a valve piston collision event.. ,,

could this modified ECU be picking the wrong map or something? can the OEM one be put back as a test? I was wondering it could be confused and outputting a strange pulse for the spark or the ignition timing. the electronics may be able to advance or retard the spark and perhaps doing so extremely incorrectly? I don't know how one could even check that.

Sounds like too much spark advance at cranking, if you can hook up a timing light maybe you can estimate the amount of spark advance you are seeing at cranking. More than maybe 15 degrees is probably bad news.

Try disconnecting the KLR and jumpering the connector:
https://rennlist.com/forums/944-turbo-and-turbo-s-forum/1066928-new-old-951-will-only-start-with-klr-jumped.html

This removes the KLR's functionality but if the KLR has lost it's marbles your problem should go away. If that does not help then maybe the Rogue DME/tuner is getting the spark advance wrong sometimes.

-Joel.

First thing I would do is change the battery out, then the starter.
or pull the plugs and see what the voltage drops to when cranking with the plugs out.

it seems like the electronics might be able to change fuel injection, but what I dont get is how they could change Ignition timing, because isn't that set mechanically by way of the rotor and cap relationship in the distributor? maybe it is the case that the spark timing can still be advanced or retarted electronically.. by a few degrees maybe?


ok so lets assume that the spark can shift, but as it is passing the correct distributor "tab" for the corresponding plug, there must be some limitation to that, because if it were shifted too far, then the spark may happen when the rotor isn't pointing to the correct spot in the cap, so what would that do? cause no spark to get to the plug? it may try to spark but if the rotor isn't lined up well enough then it may not have the ability to jump the air gap.

what if we pull the cap and look for any signs of carbon tracking inside the cap, maybe that could cause odd effects? spark jumping to some ground within the cap , rather than to the plug lead ?


in my 88 ford van with an L6 , there is a step where you unplug the "spud connector" this disables the spark advance/retard ability from the ECU, then I set the timing, then I plug it back. this procedure is normal and a way to eliminate the adjustments made by the ECU during normal operation, It's done that way so that the timing setting can be constant and adjusted separately.

with this it also means that if a person just connects a timing light , he may "think he" set the timing correctly, but since the spud connector is still connected , what that ( error in sequence) will do is cause the timing to be incorrectly adjusted.

i don't think the Porsche has anything that is unplugged in this way so when you are setting the timing, could the internal electronic adjustments it is making , be interfering and causing one to set the timing up incorrectly?

normally if you have the timing light on and you are setting that up there will be a RPM, at near idle,, and if you were to increase to 2000 RPM, then you will see the timing advance.. ( or is it retard) easy to get the two confused but if you make it happen sooner in relation to TDC that's an advance.. i think.. so basically the timing light and the mark would move towards the right as you speed up the engine, facing the engine from the front.. ( I think)

anyway I ws thinking maybe with timing light connected you might be able to view if the timing is shifting, as would probably be expected with RPM changes. I believe this may show that the electronics are actively changing the timing, based upon RPM..

the old school way was to change timing by way of a mechanical advance in the distributor, but that moved to being electronically controlled prior to the 944..


if there is an E-prom involved, they can be damaged form static.. if you touch it you can wear an anti static wrist strap to make sure your own body doesn't discharge a static shock , that can blow them.

you dont need an "official one", a couple of wraps of copper wire to a ground point will make sure you are grounded. I often just reach for a grounded place and make sure to contact that prior to touching a chip.. that discharges any static in my own body prior to contact with a chip. something like that can wipe out the chip's memory and ruin your day..

Did you do the bell housing cut out around the speed/ref sensors? That can cause extra EMI that can cause slow cranking.

odd question but were you charging the battery? any chance a short happened or a cable mixup? I I blew up a Volvo ECU once with a momentary little goof up.. ( wasn't thinking right and got my cables or my charger backwards) it happens.. I could hear a little pop and then that was it , no start.

Thanks for the mountain of advice. I'll give your ideas a try and respond back. kdjones2000, I did cut the bell housing around the sensors years ago. JFRAHM, I have another KLR, I can try also. Thanks again!

Good thought KD, it's probably the cut bellhousing. There might be a way to get some metal tape in there or something to reduce the interference from the starter.

the available high torque starter might produce less electrical noise also, at least based on some postings.
-Joel.

often the way to prevent EMI is to construct a "faraday cage" I dont understand the reason to cut the bellhousing but maybe a grounded cage can contain or prevent this stray interference.. ? it might not be so hard to construct a rough but workable cage and ground it, as an experiment.

MW,
Faraday cage helps with visual hallucinations, too.

in a lot of the old radios I work on tubes and coils and sensitive balanced circuits are often shielded.

I have a collection of ( Canadian) rogers spray shield tubes.. they sprayed a metallic coating over the glass envelope to prevent interference..

Its also not uncommon to see shields in place on electronic boards of today.. a faraday cage is just a grounded cage to suppress interference, nothing to do with bondage ;-)

I dont think I'm hallucinating but I dont know why the bell housing was cut away..

i remember a case where we were trying to solve an ongoing problem in a very sophisticated machine. It baffled everyone that the machine would stop following it's machine code and become confused..

eventually a engineer hooked up a scope and watched and found an electrical wire that conducted a fair amount of current was intermittently emitting a spike via the wires being close.. he found a certain stage of the machines repeating cycle there was a blip on the scope..
the electronics got confused because the low voltage wire was getting power by induction.. When you have a current carrying wire near to a sensitive electronics wire there can be enough induction to create basically dirty power spike in the sensitive electronics wiring.. It was a cause that baffled people until it was found and proven..

the starter is DC but it does get turned on and off and there may be some hash by way of the motor's armature and sparking there..
it could be possible even as an experiment , to wrap some wiring in copper foil and ground it to the body or use a shielded wire or similar. even just try to separate any wiring that may be grouped?


i remember having a strange problem with my 89 volvo. it would start if I turned the key but once I let go it died.. I forget how i solved it but I found that my coil had good wiring rusty terminal posts,, and that caused interruption in the spark.. at the time I was trying to sort out a TDC sensor that had gotten damaged, lady friend hopped in, started car, gave full throttle with the key still turned, starter gear exploded, got carried up around bellhousing and bashed the TDC sensor and the plate full of holes that it reads,, called a flex pate.

instead of pulling the engine I drilled a 1" hole in the bottom of the bellhousing so I could see to use screwdrivers and vice grips to straighten the mangled flex plate. I turned it round and round bending it until I got it untangled enough to run.. of course frustrated and internally cursing the ladyfriend.. like couldn't you kill any car if you did that?

it worked, and prevented me needing to pull the engine or trans for access..

when I rebuilt the same engine I got it all back together and it would start and run but not over 800 RPM.. I called the volvo mechanic guru and over the phone he said, " I know what you did , you put the flex plate on wrong" .. at that point I had been renting a car, pretty much blueprinted the engine so my wife would have a good car,, I was frustrated,, tired of paying the rental,,

I had it towed to his shop and he had the tranny in and out in about 2 hours.. I had marked it, yes, but somehow it got on the wrong bolt hole.. I was pretty impressed he could diagnose that over the phone..
Everything was right except for my TDC position being confused.. I'm still driving it,, some 20 years later lol, paint all worn off, still runs like a champ.



I remembered then,, back to the same issue with an old points distributor, some 70' car I had... it had a "ballast resistor" involved.. at the time it was explained to me that this had some effect where it could give a good spark during cranking.. I replaced the resistor and that fixed it and I recently heard something in 944 discussion about a ballast resistor in the porsche wiring.. is there such a circuit involved? o we know about the 944 having a ballast resistor and if so , what is it's purpose?

I looked up the definition and it says :
A ballast resistor is a resistor that resistance increases as current flows through it is increases and its resistance reduces as current flows through it drops. As a result, a device that helps to preserve the electrical circuit's stability.


I never fully understood how that circuit worked or why it was necessary but I wondered if such might have to do with enabling the car to spark and fire before the engine electronics located the TDC position. I think it was involved in the start ( key) position wiring and not the run position or something similar.. but Im thinking back some 40 years or so..


Ill go sit in the corner with my tin foil hat now, so you can't read my thoughts ;-)

OT: Monkey, the need for a ballast resistor is because the coil has to work at 9-10v so it can fire the plugs while the battery voltage is low due to the cranking load. And then it has to work at 13-14v while running/charging.

To facilitate this in Ye Olde Times they used a ballast resistor to lower the voltage and limit coil current during normal operation, and simply ran a bypass wire from the Start terminal directly to the coil. Thus the ballast resistor is bypassed during cranking. The 944 does not need one as a regulated voltage can be sent from the DME directly to the coil which is unaffected by the cranking-related voltage drop.

On-topic: The bellhousings on 944s get notched by people who want to pull things apart without first removing the speed/reference sensor bracket. The notch opens up the sensors to the blasts of RF from the starter brushes, like when you hear lightning strikes on an AM radio.

An intact bellhousing is needed to keep out "The Signals". It's like a tinfoil hat for your fuel injection system.
-Joel.

Joel, Thanks for the clarification..
it all makes sense , what you said, except I don't get why the DME wouldn't see a voltage drop anyway.. with starter motor running, it draws the voltage down throughout all the wiring.. right?

maybe a capacitor on the DME makes up the difference for short burst , or similar.. to transform the voltage to a higher voltage you'd need to make AC for an old school transformer to function unless its' like the modern digital "transformerless" power supplies that can..
bit of an electronic mystery there for me.. I guess. it doesn't really relate to the problem at hand.

OT: Regarding the DME and voltage drop, it probably runs at +/- 5v and maybe has a 9v or so regulated voltage for the coil so 12v doesn't mean anything to it. As long as the input voltage stays above 9v or so everything works the same, the voltage regulators (internal to the DME) just don't work as hard.

I am replacing my early sensor bracket with the updated version to determine if that will help my issue. The shield on the updated reference sensor is supposed to reduce the electromagnetic interference from the starter. In the link below, a diagram of the updated bracket, part #32 is the speed sensor 2mm spacer. Is this part needed? Using that spacer puts the reference sensor 2mm closer to the flywheel contact. A measurement on the bottom of my old bracket had both sensors sticking out 26.5mm each. The updated bracket measures the same 26.5mm for each sensor without the spacer.
https://forums.pelicanparts.com/porsche-924-944-968-technical-forum/796112-speed-sensor-shim-when-why.html

what should the air gap (clearance) be ?

I noted that in post 12 that the spacer was omitted..
https://forums.pelicanparts.com/7905988-post12.html

I think a thin sheet of copper or brass would be easy to form into the shape you like, you can make a box like shape , or what you like and solder edges or fold any over so they aren't sharp.

brass shimstock shouldn't be too hard to get.. I doubt thickness matters, so if its not too thick it might be easy to form into a cover of some sort. how about making up a "prototype" of card stock and masking tape , then use that as a pattern? easy enough to solder a wire to or maybe just catch it under one of the bolts that exist?

maybe its reason to take the starter apart, put the armature in a lathe or even a drill press and recondition the commutator.. polish it up and check the brushes, if its all dirty it may spark more and thus make more EMF.
Ive done this on an upright drill press, mount a center point on the table so it can spin as if it's in a lathe, other end in the chuck.. then polish with some wet-or-dry paper and clean with alcohol and toothbrush after.

maybe some shielding could be put, sort of surrounding parts of the starter ?

I was jus working on a 110VAC brush motor used in a pipe threading machine.. the brushes got greasy, that made it spark like crazy. cleaning the commutator didnt fix it until I took the brushes to the grinder and took a touch off to clean them up then way less sparking, so don't over-grease or over oil the starter..

It might have bushings that could use a drop or two. or they can be replaced if worn.

Problem solved! I replaced the reference/speed sensor bracket with the sleeved updated version, and this has solved my hard cranking problem. It is a relief to hear the engine turn over easily 100% of the time. I really appreciate all of your suggestions. Thank you! One of the main questions I had about the installation of the updated bracket was whether or not the 2mm spacer on the rear sensor was required. It is. This is because the starter ring on the turbo is larger than the NA version. Failure to set sensor height without the 2mm spacer can result on contact with the flywheel. Thanks again!