Ischmitz,

Two questions concerning the DME. The other car has a SW chip in it, will this affect my 87. Should I put the stock chip back in? Also the other question I have is my car is a Euro engine vs the NA engine. Is the timing between these two engines the same? I just want to be 100% before going long ti

Dennis,

both the SW chip and the difference between Euro and N/A are so subtle that they won't cause any damage. I would swap your DME into the known good car and have the owner drive it to see if he can replicate the problem. That way you will know if the problem travels with the DME. If yours doesn't have the a chip your friend might notice a slightly different throttle response. Nothing to worry about though.

Also, if you can borrow the good DME for a longer period of time you could first drive your, reproduce the problem and then swap in the good one as soon as you have the no-start issue.

Ok I will give that a go and report back. The 88 is my dads car and the two of them sit side by side so this won't be an issue to get for a more lengthy trial.

This should make it really easy - swap the DME's and find out with car will have the issue. I am curious.

Hi Dennis, I have been keeping an eye on your progress. It looks like you are getting pretty close to finding where ever the fault is. Just a quick question, when you attempt to start your car, and when you first turn the key, can you feel a click from inside the DME relay, and then if you turn the key further, can you then feel the second click at the DME relay? I ask, because it will make the two inner contacts, connect the farther you turn the key to start the engine. But, you should still feel two distinct clicks from the two inner contacts closing inside this duplex style relay.

Another question, being I can't remember if you stated or not. Can you smell raw gas at your exhaust outlet, when your engine is displaying a no-start condition? If I can think up, or come across any other ideas that might help, I will post that info for you. Good luck!! Tony.

Thanks 3.2Cab.

I have just rewired the fuel pmp so it only starts up on key start only condition. Next time I try and start the car I will check the DME relay, however I have tried to swap it out with a good DME relay with no effect. There is no smell of fuel, and I now know that the injectors are not getting a voltage change from the DME.

Open to all suggestions, so keep them coming.

I am going to try imschmitz idea next.

OK latest update.

I took the DME out of the working 88 car and put it in my car. Drove it around for 10 minutes and shut it down. Tried to restart and would not as usual.

Just to be sure I also took my DME and put it in the 88 car, and took it for a 15 mintue drive. Car ran fine. Once it was warm I turned off the car and started it back up with no issue.

I would think this now conclusively rules out the DME being damaged internally or a solder issue.

OK, now you are getting somewhere. The next step is to isolate WHAT prevents your car from starting when the engine is warm. Fuel, spark or both. I have my money on fuel.

- If it is fuel-related you could have a bad fuel pressure regulator, bad fuel pump or check valve.
- If it is spark-related I really can't think of anything but the coil acting up when warm.
- If both fuel and spark are missing it's got to be one of the two flywheel sensors.

What confuses me is that you can not re-start when warm but you don't report issues when warm. Maybe vapor-lock in the fuel system due to a bad check valve in the fuel pump. Just a guess.

How long does it take after the no-start until you get it to run again. This could be indicative of what is going on. Something like hours or minutes.

As the next test, have an assistant with you and bring a spare spark plug and gloves. Once you get the no-start, get to the engine bay, disconnect one of the spark plug wire and hook up your spare spark plug. Have the assistant crank while you ground the spark plug to the engine metal and see if you get a spark. If yes you have ruled out your sensors and the coil and pretty much narrowed it down to fuel pressure or delivery. I like to wear gloves because stuff is hot and I don't like to get zapped by the spark.

ingo

Ischmitz,

I have already confirmed that I have no spark (did spark plug check even off coil wire) nor fuel injection signal nor smell of fuel. Which should point to the flywheel sensors, however I have tested both with a scope and both seem fine. I even replaced the reference sensor just to be sure.

In terms of time to operation, normally it takes 2-3 hours before the car will start up again.

The only thing that sort of concerned me when I replaced the reference sensor was that my cut I made in the bell housing of my transmission appears to be very tight if not touching the edge of the idle sensor. Could this cause an issue. I do get a good strong continuous sine wave from the sensor at the computer. I can try and take pictures later this week.

Thoughts?

Warmed up, dying, no spark are the classic bad coil symptoms. Happens to us old harley guys every few years. Try swapping w/ the other 911's.

Hold on, maybe I am being misunderstood. The car runs fine and doesn't die; it will run perfect. When you shut it off then try and start it up again 10 seconds later, or 10 minutes later... it just cranks over. You have to wait at least 2-3 hours before it will get spark and fuel again.

I would think that if the coil is bad then the car would die while it was running, that isn't the case here.

OK, so if you don't get spark nor fuel one of the sensors seems indead to be the issue here. I remember other threads where temperature-related issues where finally tracked down to one of the sensors.

When you say you tested them how exactly did you do that test? With an oscilloscope I assume? Can you test the signal while the sensor is hooked up to the DME once when cold and once when hot? Just have the cover off the DME and attach the scope probe to the pins of the 35 pin connector while cranking.

At this point you could be chasing a situation, one of the sensors gives a reasonable signal into the high-impedance (10 MOhm) scope input but its signal maybe breaks down when hooked up to the DME input stage with an impedance in the 30 - 100 kOhm range.

All this still is doesn't quite explain why cranking is so different from running but it is my best guess at this point.

Ingo

Ingo,

I agree! Most likely this is related to the Ref and Speed sensors.

When the car won't start do these tests:
1) AFM must have 5V
2) simply place the scope or LED test light across the '-' and '+' posts on the coil you should have coil pulses or the LED must flicker (Only a LED test light will work) you are testing the primary side of the coil here not the secondary.
3) unplug injector #1 and place scope or LED across the 2 pins in the unplugged harness, you should have pulses or LED flicker.

If you do not have coil or injector pulses then unplug the Speed and Ref sensors and place scope across them, my bet is one of them is dead or weak. Also do a resistance test of the 2 sensors with car cold and with the car warm under no start. The 2 sensors are identical and must measure 1000 ohms +/- 20% (most read about 900 ohms). The sensor(s) could simply be heat soaked and fail after shuting down a hot engine.

It is most important to verify that you do or do not have coil and/or injector pulses at failure time. Intermittent issues are a real pain to figure out, but your condition remains for more than 30 minutes so you should be able to figure this out.

Bottom line is that no coil or injector pulses are often from bad ref/speed sensors.

CylHeadTemp sensor will not result in missing pulses, it simply causes rich/lean conditions coupled with stalling. But you will have coil/inj pulses even if CHT sensor is shorted or open.

You only need a simple cheap 12V LED test light to test for coil and injector pulses. Don't waste your money on expensive Spark or Injector testers! Just get something like these:
http://www.autotoys.com/x/product.php?productid=686
http://www.audiooutfitter.com/ib27450/install-bay/p74410.html

Also Walmart sells one like this for about $5.00 in the automotive section, you must get an LED test light not a normal bulb one. The normal ones won't work for this purpose.

It looks more and more like the reference sensors are either bad or way out of adjustment. It may only be one of them bad, but when mine had the no-spark problem, there was only one actually defective, but I replaced both of them, being both of them had been there the same amount of time, subject to the same heat and other factors that are known to be damaging to them. Good luck!! Tony.

You can use the BMW Ref/Speed sensor at 1/2 the cost of the Porsche one. Our host sells both search for the following.

Flywheel Ref Sensors
911-606-215-01-M14 (OEM Style - Pelican)
12-14-1-708-619-M14 (BMW style - Pelican)

I just helped someone else out with no-start issue and he confirmed the BMW sensors work and look the same as the Porsche ones, direct plugin.

I would much rather have you test them with the scope when the car does not start just so we know what the signals look like when no start. And if you could test them with car cold as well. If you simply unplug one or both the car will not start and you will not have coil or injector pulses even with the engine cold.

If you want to see what the coil and Ref Signals look like on a scope in my car see this thread:
http://forums.pelicanparts.com/porsche-911-technical-forum/433152-diy-wasted-spark-84-89-carrera-2.html
post #66.
Sorry, I do not have pics of the speed sensor.
But you can see that my ref sensor pulses are about 3v peak to peak.

Pin 2 is wired to the idle stop microswitch and when this line goes to ground (switch closed when at idle stop) it tells the DME to start using the idle maps.

Pin 3 is wired to the Full Throttle microswitch switch (WOT Switch) and when throttle is fully opened this switch closes and grounds pin 3 which tells DME to start using the WOT maps.

If nethier switch is closed (not idle and not WOT) then the DME simply uses the PartThrottleMaps.

Hope this helps

In addition to what Sal said the inputs also decide how the DME handles fuel signals and the ICV operation:

With the idle contact closed the fuel signal will be cut off above a certain RPM. Somewhere at the 3000 RPM (don't quote me on the exact number) with a hysteresis window of about 500 RPM. This is used to decelerate the engine when you lift the throttle. In addition the ICV will try to maintain the programmed idle speed. For that the DME varies the duty cycle of two signals that move the ICV.

With both contacts closed the ICV will move to its center position (50% duty cylcle) and NOT control idle air. This is used when adjusting the air bypass screw on the AFM.

Ingo

Ingo, I see you also figured out that jumpering pins B&C at the diag port is the same as closing or shorting out the WOT switch. I figured this out some time ago and I have an old WOT switch that I soldered closed and simply unplug the harness from the WOT switch in the car and plug it into my modified one to set base idle. Kind of neat trick.

Also as you mentioned the ICV circuit is only activated when the idle switch is closed.