PerryB

I have been a reader of this forum since before I bought my 1987 951 in December of 2011. I have never had to ask for help, but I have gained much useful information from many posts and replies. I recently registered and hope that readers can help me.

Unfortunately, I have not been able to start my car since last December. I took the car off the road to begin a clutch job that took about one and a half months. After finishing, I went to start the car up for a test drive only to find that the car would crank, but not start.

I have tried for the last one and a half months to diagnose this problem but have had no success. I have also posted on the Clark's Garage forum, but am still unable to fix this problem. Please bear with me as I describe all the work that I did during the clutch job, and the subsequent work I did trying to cure the no start.

Removed exhaust from turbo discharge pipe back.
Removed transaxel.
Removed torque tube and rebuilt it.
Removed bell housing, replace clutch--including pilot bearing, T/O bearing, rear main seal, fork bearings, pressure plate.
Resurfaced flywheel.
Removed speed and reference sensors.
Removed and replaced the oxygen sensor.

When I removed the speed sensor I broke it. It was stuck in the mounting block and it broke off in my hand as I was twisting it to remove it. I replaced it and the reference sensor, as I was also pretty rough with it during removal. I slotted the bell housing opening to ease re-installing it. I set the clearence between the speed sensor and the ring gear with a feeler gauge before installing the bell housing.

I re-installed everything without incident.

I also replaced the timing and balance belts and the crank gear that drives the timing belt.

I also changed the engine oil and filter!

I cleaned the ground connections at the top of the bell housing and the back of the block (as well as the one on the firewall and the B- connection).

The car cranks strongly but does not start. It has neither spark nor fuel injection nor tach bounce.

I have tested the coil for primary and secondary resistance. It checked out.

I jumped the DME relay with a three wire jumper. Fuel pump runs, mid 30's psi when just running, low 30's when cranking.

I have tested the new speed and reference sensors for resistance, cranking voltage, and, after borrowing an oscilloscope from the local science teacher, peak to peak voltage. All values are correct.

I also replace the wiring harness for the speed and reference sensors. Still no start.

I tested the continuity between the DME computer hook up and grounds, the AFM, coil, ignition switch. When I tested the speed and reference sensors I took their values at the DME connection.

I tested for cranking voltage at the DME: 11.1 volts when cranking.

I removed the KLR and jumped the ignition pins (#16 and #9, if I remember correctly). Still, no spart, no fuel injection.


I concluded that the DME seemed to be getting all the correct information, but was not putting out the correct outputs. I bit the bullet and replaced the DME with a rebuilt one. Still, no spark, no fuel injection.

Am I missing something? What could I have done during the clutch job that would cause this condition? I am severely depressed and am looking to the gurus on the forum to point me in the right direction to fix this car.

I will greatly appreciate any help.

Thanks,

Perry

Van

These statements are conflicting. I'd tend to believe the first one is accurate - the DME is *not* seeing a signal from the speed and reference sensors.

Have you swapped the sensor plugs, just to see what happens?

PerryB

Van,

Thanks for the reply. I don't know what you mean here about conflicting. If you can explain, I will clarify.

I have swapped the sensors with their connections in the harness on the chance that I got them reversed. It didn't make a difference. I recently double checked to make sure that the speed sensor is connected to the speed sensor harness and the reference sensor is connected to the reference sensor harness.

Perry

konakat

No tach bounce means no reference sensor or a fundamental failure for the DME. I would have bet on having the reference and speed sensors swapped, but it sounds like you have that covered. Make sure to check the part number for the DME to ensure it is the right one for your car. Beyond that I would check power and ground for the DME.

Dal Heger

Did you replace the set screw on the flywheel after you had it resurfaced, sometimes these tend to get lost. If you did, did you set the height correctly so that the reference sensor has the correct gap? This would explain the no start (since the DME doesn't have the TDC reference) and the no tach bounce when cranking, since it doesn't have the signal from the reference sensor.

Just a thought.

william_b_noble

when you tested the sensors with the scope, did you see the proper pattern of pulses on both sensors? I would remove the plugs to reduce the load on the starter when you do this test. If you see the proper pattern (one of them will pulse four times as often as the other), then the next step is to repeat the test measuring at the pins where the sensors connect to the DME (down by the passenger's seat foot well). As others said, this really sounds a lot like a sensor issue. maybe the harness got pinched when you moved the engine? maybe a wire is broken? The diagnosis chart in the service manual is pretty helpful, do you have a copy of that?

Van

What I mean: if you're not getting a tach bounce, an injector pulse or an ignition pulse (and presumably the fuel pump isn't getting power) - and you've tried 2 DMEs, then my guess is the sensors' signals aren't getting to the DME(s).

You might *think* the sensors' signals are getting there, but it sounds to me like that's not the case.

I agree, use an oscilloscope at the DME plug to see if you're getting the proper signals.

I also think it's inaccurate that one sensor will pulse 4 times as frequently as the other... It should be *a lot* more. One sensor will pulse every time the set screw on the flywheel goes past (make sure it's a steel set screw... I once knew of a car that had a problem because an aluminum set screw was used); and the other sensor will pulse every time one of the starter ring gear teeth goes past.

Starter ring gear = speed sensor (this tells the DME RPM)
Set screw = ref sensor (this tells the DME where TDC is)

PerryB

Thank you to everyone that posted. I hope that this brain trust will be able to solve my problem.

I used the exchange service offered by Specialized ECU repair in Florida. I have to trust that they have given me a correct ECU, I gave them the correct information about the year and model of my car.

The machine shop that resurfaced and balanced my flywheel did not remove or alter the height of the set screws as far as I could tell--although I didn't check with any degree of precision. They just looked the same as when I gave it over to them for their work.

About the oscilloscope: I borrowed this unit from the science teacher at the local high school. It must be 60 years old. It was made in Brooklyn, NY, uses vacuum tubes, and takes about 2 minutes to warm up before it displays any pattern. I found a users manual for it on the internet and have set it up to read peak to peak voltage. How accurate it is or whether I am using it correctly are unknowns, but I think that it's readings are correct in that it displays what the DME/KLR test manual says it should display for these sensors.

About the DME/KLR test plan manual: I downloaded the official Porsche document (thank you Arnworx) and have followed it as best as I can. During my testing of the speed/reference sensors I took readings at the harness to DME connector in the passenger's footwell. The speed sensor gave a peak to peak voltage reading of about 12 volts during cranking. I say, "about 12 volts," because the oscilloscope only has an analog display of a sine wave on this green screen. I have to quickly read the amplitude of the sine wave against these cross-hairs that are stenciled on the scope's green screen as I crank the motor. As the manual says that the minimum voltage for the speed sensor is 2.5 volts, I am confident that the speed sensor is putting out enough voltage to activate the DME.

Similarly, the reference sensor also puts out a single spike that has an amplitude of about 12 volts, with an initial positive spike, as called for in the DME/KLR test plan. This was harder to read that the speed sensor's output, but I believe that the reference sensor is sending the correct input to the DME.

I also have tried to vary the gap of the sensors by moving their mounting block up and down. By doing this I was able to dramatically decrease the voltage put our by the sensors. Moving the sensors closer to the ring gear increased their voltage output. They are correctly set at a gap of .8 mm, which generates the above mentioned voltages.

Some specific questions...
How does the tachometer read the engine speed? Could a problem with the tachometer prevent the car from starting?

Could a bad speedometer sensor back at the transaxel prevent the car from starting? (I ask this because I wonder if there is device that prevents people from unhooking the speedo sensor and thus driving without recording any mileage.)

Is there an inertia switch that shuts off the motor in case of a serious impact? I know that Alfa Romeos with Bosch injection have such a switch and that when the switch malfunctions the cars will not start.

Part of my problem is that I am really unable to take the car to a Porsche shop. I live in a remote part of Colorado. The nearest shop that would work on a Porsche is 130 miles away from my house.

I am really at my wit's end with this problem. Thank you for any help that you can give.

Perry

konakat

This is really perplexing. It could possibly be a connector issue, one of the contacts in the DME plug could have been pushed back or damaged so that it doesn't make contact with the DME pin. To test that you would have to remove the DME case and use the scope to get sensor traces from the DME connectors. The fact that you have no tach bounce or ignition in tandem indicates that the problem is on the input side of things, so at least that somewhat limits what you are looking at.

I would also recommend doing all your testing with the DME relay shorted. In case you have multiple issues going on at once, that will remove quite a few variables while you debug.

Van

The DME sends a pulse signal to the tach simultaneously with sending the pulse signal to the ignition coil. Car will start just find without a tach.

Car will also start just find without a sppedometer.

There is no inertia switch in the DME or engine electrical system.


Make a 3-way jumper to connect all the large large spade terminals from the DME relay. See if that makes a difference.

PerryB

Van, Konakat,

Thanks for the replies, thanks for your help.

I have made the three wire DME jumper and used it for some of my testing. I also installed a new DME relay. Neither seemed to any difference. I will, however, use the jumper whenever I do any future testing.

I agree that this sounds like a sensor malfunction but I've replaced both, gapped them correctly, and replaced the wiring harness that connects them to the DME. Additionally, I have tested both of them, through the harness, with an ohm test, a cranking voltage test, and the oscilloscope. They both have passed these tests.

I will redo these tests tonight, but I don't believe that there will be any change from the last time that I did them.

What is really confusing is that this car NEVER had any starting or running problems prior to my clutch work.

Again, I greatly appreciate any help.

Perry

Van

Tell us more about this... What other changes have you done to the wiring harness?

Does the car have an alarm? And, there is a pin in the DME plug that gets a 12 volt signal from the ignition switch when you start the engine... is that still working?

You say there's fuel pressure in the fuel rail? You should be able to use your state-of-the-art oscilloscope to check for an injector pulse at an injector plug, too.

PerryB

Van,

Thanks for sticking with me!

I had emailed Josh Cunningham at Rogue Tuning with my problem. He very generously gave his time and mentioned that one car that he had worked on passed all the normal diagnostic tests but was still a crank, but no start problem. He replaced the harness between the sensors and the DME and that cured the problem.

On the off chance that I was suffering a similar fate, and recognizing that during the clutch job I had pushed the sensor harness around to get to fasteners etc. at the back of the motor, I thought that there might be a intermittent discontinuity between the sensors and the DME. I purshcased a replacement harness from Ian Taylor at 944online and installed it. It didn't cure the problem. After that I was even more convinced that the problem was with the DME--everything else had been checked and/or replaced. I ordered the DME from Specialized ECU and installed it. When that failed to solve the problem I started this thread. Otherwise, the car is completely stock, including the wiring harness.

The car does NOT have an alarm, neither factory nor aftermarket.

I have tested both the pins on the DME (I believe that it's either #35 or #18 or both) that are the power to the DME. They both have continuity to the B+ terminal and get battery voltage during cranking and with the key in the ON position.

I have not used my state-of-the-art (funny!) oscilloscope (or a noid light) to check for injector pulse. Since there is fuel pressure and the plugs are not wet after extended cranking and I get no smell of gas, I concluded (perhaps wrongly) that the injectors are not firing. I will check that out after I let the scope warm up for a few days.

I will check for voltage at the sensors and injector pulse ASAP, tonight if possible.

Please, keep asking questions. Hopefully we will encounter the "smoking gun" that will allow me to fix this car.

Perry

william_b_noble

if you are measuring 12V at the sensor pins, that just doesn't sound right - some thing is wrong either with the measurements or with the scope or with the way you are connecting it. It sounds like an old Dumont scope - they didn't have a floating ground - if that is what you were using, all bets are off. I don't see a location - is there anyone near you who can loan you a more modern scope? can you use your ohm meter to check continuity through the pins to the sensors? can you borrow a dual trace scope and look at both at once? If you are a bit familiar with electronics, you can follow the pulses into the DME, on the off chance that you are able to do that, I've attached the DME schematics I have

PerryB

Mr. Noble,

Thanks for your reply.

Thank you also for your attached files of the schematics for the DME. The label at the bottom says that they are for a 911. Are the 951s the same?

I had some trouble reading the brand/model name on the oscilloscope--it was in Cuneiform! I used Google translate and it is an EICO Model 460. I will try to locate a more modern scope, but that might prove to be impossible.

I do have a high quality, modern, digital volt/ohm meter. Taking an ohm reading of both sensors through the DME connector and harness they both have about 875 ohms. The specification I have for them is between 600 and 1200 ohms, so they both seem to be right in the middle of that range.

I calibrated the scope per the instruction manual's set up for peak to peak voltage readings. When cranking, the speed sensor registers a sine wave with an amplitude of about 12 volts. The reference sensor has a single spike with a positive initial slope, that is also about 12 volts in amplitude. I'm not sure what you mean by "floating ground," but I hooked up the scope as per the manual's instructions to get these readings. The readings are repeatable, and, I was able to decrease the voltage output of both sensors by increasing the gap between the speed sensor and the ring gear. Correctly gapped at .8mm they produce the 12 volt readings I mentioned earlier.

I am in beautiful but remote Creede, CO. The nearest shop that would work on a Porsche is in Durango, CO, which is about 130 miles from my home. The nearest Porsche dealership is in Colorado Springs, CO, about 240 miles from my home.

The trouble seems to be that the DME gets the correct input, but then doesn't ground the coil or fire the injectors. All the grounds and all the sensors and all the positive connections to the DME are correct, but it still doesn't produce the correct output.

I recently read a no start thread on this forum by "specsalot" describing a weak spark problem that would not allow his car to start. He eventually discovered that a problem in the KLR was actually damaging the DME and that both units needed to be rebuilt. I want some more concrete proof before I send off my KLR for another $300 rebuild.

I can't help but think that this problem stems from the work I did replacing the clutch--that I left something detached, or I broke something while wrestling with the bell housing, etc.

Please keep asking me questions and sharing your thoughts.

Perry