Good crank, no spark – Anything left but ECM?
 

Hello braintrust. I’m trying to diagnose a no spark condition on an ’88 Carrera and am hoping I’ve forgotten or overlooked something.

I was idling at a stoplight when the engine suddenly turned off. No cough, no sputter, just off. Once towed home, I found no spark at the plugs. So here’s what I’ve checked in order:

• Tried a brand new DME relay from our host (no change)
• Found plenty of fuel in the fuel rail (just in case)
• Removed distributor cap and cleaned the copper points and rotor head (no change)
• Checked for spark coming from distributor input wire (no spark)
• Tested the coil for incoming power and internal resistance (12v and 7 ohms)
• Removed and cleaned the 11th fuse in the frunk (no change)
• Tested spark plug wires for resistance (1 ohm)
• Checked resistance on crank reference sensor terminals (97ohms on 1&2, 9,600 ohms on 1&3)
• Had the engine rotated while I looked for the crank reference pin through the transmission porthole. I didn’t see it, nor did I see a hole for one. Chalk that up to user error
• Visually checked all three DME grounding locations and all were still tight and solid

Is there anything I’m forgetting before I try a loaner ECM? The Bentley manual says the crank reference sensor resistance should be >100,000 ohms, and I’m getting .96 on my 200k ohms voltmeter. But I get the same .96 when I try the speed sensor resistance, so I also think that’s user error.

If there is no fuel pulses and no spark it could be FW sensors. If you have fuel pulses (strong smell of fuel at the tail pipe after prolonged cranking) but no spark this points to the DME itself.

The best way to check for fuel pulses is with an LED test light or NOID kit on one of the injector plugs. Testing for spark with an old spark plug grounded to the engine while cranking.

The sensors both need to measure around 1000 Ohm (1kOhm). Assuming you reading is in kOhm 0.96 is a good value. Also check OL between shield and signal pin. And then some sensors measure OK and still turn out to be root cause. So ohming and finding a bad result is conclusive. But ohming them and finding the readings in spec is NOT conclusive. The best test is to check for fuel pulses and spark pulses while cranking. When both are absent 9 out of 10 times that's a bad sensor. Only in very rare cases the DME is at fault with that finding.

If that's the primary resistance, it's 10X to large. Hopefully it's not a silver coil.

Read here; http://forums.pelicanparts.com/porsche-911-technical-forum/621262-911-3-2-no-start-troubleshooting.html

If you have a silver coil, buying a replacement is just a 505-50 odds crap shoot that the new one is any good. The old 35 year old black coil is usually a far batter coil.

If the flywheel sensors are original, replace them and the head temp sensor even if they test out fine. They will not live much longer anyway. It can be a bad connector or just a bad head temp sensor itself.

The OP measured the resistance of sensors and the value seems to be OK. It’s not clear if this is a “silver coil issue”…

A quick test with an LED test light will reveal if the sensors are likely at fault (spark and fuel pulses absent) or not. Why not systematically diagnose rather than throw parts at it. And if the OP has access to a loaner ECU then by all means try that too rather than falling for the “check and rework the ECU“ scams that don’t fix anything and come back with “you must have had two problems and I only fixed one. Good luck”

No doubt just throwing parts at is is not a sane diagnostic method.

But if my 3.2 Carrera had the original 35+ year old flywheel sensors and I ever planned to drive it much (and I do) I would replace those sensors even if the car had just 20,000 miles and was a perfect concours car. Maybe that is just me. I have 185,000 miles on my 911 and those sensors have failed twice on me.

I will be doing another 3,500 mile road trip soon in my 85. I want it in ready to drive condition. The flywheel sensors live in a harsh environment. I prefer to KNOW they are fresh before driving 1,200 miles from home on a 10 day road trip.

Thanks all.

Both flywheel sensors are <2 years old, so they are certainly not original.

What IS original (or at least very old) is the coil. It's black. I can't believe I never replaced it during my full restoration, but here we are... Anyway, it seems to be testing fine still.

Sounds like I need to source some noid lights before borrowing a ECU. I'll get on that asap and keep you posted.

keep the coil -black- most likely not your problem.
Do you have any alarm or immobiliser in your car.is my Q

Ivan

To check if it's the power supply to the DME (aftermarket alarm or factory alarm triggered) simply verify the ICV is humming with the ignition key in ON position. If yes the DME is all powered up and waiting for flywheel signals to drive injectors, the coil, and the tachometer.

Well the noid light was big fat 0. No response at all. I also removed my flywheel sensors from their bracket, wrapped them in cloth to prevent any grounding, and tried again - no difference.

I'll check the ICV (I haven't heard anything, but I'm generally playing music), but it sounds like we're zeroed in on the flywheel sensor?

The NOID light has to be plugged in with the correct orientation to flash. Did you try both orientations? Not sure what you mean by “wrapping the sensors in cloth”. The sensor works perfectly fine when grounded to the transmission in its clamp. A short can happen inside the COAX cable. GND comes from the three pin connector.

Another useful test is to use a DVM in AC setting or better yet an oscilloscope and measure the output of the speed sensor and reference sensor while cranking. Best to measure inside the 35-pin connector (DME removed).

Plug and unplug the 14 pin connector check grounds.

I was concerned the noid light might have a polarity, so yes I tried it both orientations.
I wrapped the sensor plugs (in the engine bay) away from their metal bracket because someone here suspected them might be grounding out. I don't have an oscilloscope, but I can try the voltmeter while cranking.

I've checked all ecu grounds, but haven't unplugged/replugged.

Is the ICV humming with the ignition on?

Let's try again. Read here & follow the troubleshooting procedure;
http://forums.pelicanparts.com/porsche-911-technical-forum/621262-911-3-2-no-start-troubleshooting.html

Had a similar problem with my '88 this Spring. Initial start after winter storage went well and test drive went perfectly. I then removed the Motronic ECU to install a new performance chip. Got it all back together and no-start. Could smell fuel but obviously no ignition. Removed ECU and swapped chips - still no-start. Swapped DME relay, still no go. Returned entire ECU with new chip to Steve Wong who did a diagnostic test and repair. He found problems on the digital circuit board which required re-soldering. These components are 34 years old and problems can occur just due to the age.

Not really! Most likely the circuit board was damaged (weakened solder connections) during the chip exchange, or during the logic board separation.

I can confirm the ICV is humming when the ignition is on!

@mysocal911 - I read the post you're repeatedly suggesting before ever posting here (I know the search feature exists). I do not find it helpful.

Adam, with that you know the DME gets power and only needs speed and reference pulses to release spark and fuel pulses.

Try to measure the outputs of the speed and reference sensors at the 35-pin connector while cranking. Set the DVM to V-AC and report back.

Further, do you have access to a loaner DME. That’s a quick way to either identify or rule out the DME itself.

I’m still wondering if maybe your NOID light test wasn’t conclusive. First have you confirmed +12V at the injectors and at the coil primaries with ignition on? Next, the pulses the NOID light emits are very faint and short, best do this after dark. And lastly, you can also use the NOID light across the coil primaries to check for ignition pulses.

Assuming your current findings are correct in that both fuel and spark signals are not issued by the DME the potential faults are:

(1) one of the flywheel sensor signals isn’t present
(2) flywheel decoder or logic gate controlling the flywheel decoder inside the DME are damaged.

9 out of 10 times it’s (1) but then you said the sensors are new. Still, it could be a quality issue with the replacement sensors. So don’t discount it. If it’s (2) a simple resolder will not fix that.

It doesn't get any simpler than that, i.e. it covers all the key troubleshooting areas. Sounds like you need to take it to a Porsche repair shop.
They'll find the problem is less than 15 minutes, or send the DME ECM for testing.

As an example of your not understanding, in post #4 you incorrectly measured the primary coil resistance, or the coil is bad. The primary resistance should be 1/10 the value you measured.
You should have resolved that before continuing!

Confirm you've checked all your fuses and your earth straps.

Priority for me would be to get a scope on the low tension lead on the coil while cranking. A multimeter set to measure frequency might work.

Sent from my SM-G988B using Tapatalk

I do not have a spare ECU laying around, so I'll check the signal at the connector asap.

I also just ordered a new MSD Coil from our host here. Took off my black coil (to remove any test interference from the wiring/battery) and tested it off the car. Primary is still .7-.8 ohms, but the secondary is now measuring Open.

I've checked all fuses and grounds repeatedly.

I'll try the coil and check the ECU connector while cranking.

I just replaced my speed sensor yesterday to fix a no start issue. Only a couple years old and gave the same resistance readings as you were getting and was no good.

Here's my ohm findings on my original black coil (88 3.2):

- primary is .4 ohm. Allow for some error due to contact points and cable own resistance, so your own .7-.8 ohm is still a good measurement. In your original post you mentioned 7 ohms. It must have been a typo then.

- secondary is 5720 ohms. Again, that can vary from one coil to another, but you should not measure Open, unless you chose the wrong setting on your ohmmeter. Make sure you use a proper range for high resistance (at least 10k) or use an autorange DMM.

Tried out a new coil from our host last night, and no change. Still no spark. Given the diagnosis so far and how the suggestions here are leaning, that comes as no big surprise. Tonight/this week I'll take out the crank sensor and bench test it. I'll also check the pins on the ECU while I crank to see if it's outputting signal (before I take out the sensor of course...).

Honest money says the crank sensor...

Adam, you are aware there are two crank sensors, right. The speed sensor signal should be readily visible with a DMV in AC mode. The reference sensor only puts out one pulse per flywheel rotation and it's a bit harder to get conclusive data. If this was me here is what I would do:

- Testing the sensor while in the car with a DVM in AC mode gives you some more information. Fine but as others mentioned this is by no means conclusive and difficult for the reference sensor

- Alternatively, get a cheap handled O-scope for real measurements. For less than 40$ you'll see plenty of offers on evil bay and while these might not be extreme precision instruments they'll be more than adequate to test the sensors in the car. Something like the DSO150 or FNIRSI-138. And it'll be a nice gadget you can use for future diagnostics.

- If you remove the sensors for bench-testing I'd say you've crossed the threshold where putting these back in doesn't make sense regardless of your findings. I say that given the costs for new replacement sensors (BMW type) and the hassle of un-installing and installing these.

- After above that it's time to focus on the DME itself. At some point you'll have to bite the bullet and test the DME or put a known good loaner to see if that is your issue

Haven't picked up an oscilloscope, but today I played with diagnosing the ECM

All grounds have continuity to the frame

Both power supply wires show 12v

Ignition coil wire shows constant 12v at On, constant 8v while cranking. Bentley just says "pulse voltage signal" while cranking so this doesn't seem right

Continuity on both crankshaft speed sensor and reference sensor are in-tolerance at .9k. When I check the voltage on the + inputs for both sensors, I get 2.3v at On, 2.28 while cranking. Bentley says this should be variable while cranking.

So two hiccups while testing, but the last test send to point towards the sensor (at least to me). Beyond begging for a loaner ECM, a new sensor seems to be the next step.

Adam, it sounds like you did the measurements with the DME connected? For the coil the 8V during cranking is low but as long as the engine turns over it would start if everything works. After lots of tries it’s not a bad idea to charge the car battery a bit.

The pulsing in the coil is something you need an oscilloscope for. Depending on the DVM used it might get picked up or not. So inconclusive. Use the NOID light, an LED test light or an oscilloscope.

The FW sensors seem to check out in resistance mode. Measuring DC bias while connected to the DME has little to no value. You need to disconnect the DME, set the meter to AC and measure. It will read zero with the engine stationary. While cranking the speed sensor should show a healthy AC voltage. At least 1V AC. The reference sensor it a bit more tricky and it depends on the quality of your meter what you’ll read. But you should see the meter react when cranking.

You need to use a test light (ice pick incandescent bulb type) to check for a spark signal (flashing bulb) at the coil while cranking.
Also, using that type of a test light provides a good way to check for a good 12V supply voltage for pins 18 & 35 on the DME ECM connector.

Furthermore, post 4 (over a week ago) provided a link; http://forums.pelicanparts.com/porsche-911-technical-forum/621262-911-3-2-no-start-troubleshooting.html

That link describes the ACV (AC voltages) you should measure while checking the sensors;

"B. If neither noid nor spark then; check sensors (1k ohms, speed > 1.0VAC,
ref >.10VAC), DME ECM powers (1/18/35) & grds (5/16/17), or DME relay"

On 5/13 (post 19) you indicated that you didn't find the link "helpful". What didn't you find "helpful"?

Okay @ischmitz, that was very helpful. Just disconnected the harness and tested the speed sensor inputs (8 & 27) on the harness and got 0 voltage while cranking. So the sensor isn't outputting anything into the ECM. I'll pick up a sensor today from our host and let you all know when it arrives.

Hopefully, you were on volts AC and not volts DC, right? Did you check for a reference sensor signal on pins 25 &26, again AC volts?
You are aware that the DME ECM needs both signals to start?

Again, this was posted over a week ago in post 4.

Also, if you need to replace one sensor, better to replace both, as minimal additional effort required.

Left on a vacation, but just came back and got a result I did not want to see:
Tested the new FW / Speed sensor and got no change. Still no start. Tried the new sensor in both positions to see if it was reference or speed, but no love either way. This looks like it might be the ECU after all.

So now I need to rustle up a loaner ECU to swap and try.

If you trust your data and findings then the answer is probably "no" to your question in the title and your ECU needs help

***Solved***

It's the ECM. Tried my ECM in a working car - no start. Used that proven ECM in my car and it fired right up. Something in my computer went.

So now it's off to service.

Thank you everyone for your suggestions, patience, and help with this problem. There's finally light at the end of the tunnel!

Got the ECM back from repair and the car started right up! Thank you to all who helped on this thread, especially @ischmitz!

Already did a canyon day, so happy to have the car back!

After fighting problems, it is a really sweet sound to hear the engine fire up!

BTDT, I know the feeling. The best part, is your DME is all freshly fixed, and ready to go.