got this 87 944 turbo in the shop that's testing me. it originally went to a shop near by for a no start issue. the techs there replaced the rpm and reference sensors and did a repair on the ref sensor wiring connector and advised the client that the dme was faulty and should be replaced so he towed it home and sent the dme to 911chips for a overhaul. they stated that there was no issues found but went through their normal rebuild test and returned it. the client installed it and found no change , no start no tach bounce. this is where I enter the picture. after checking a few basics and charging the battery i verified no pulse at the coil . i suspected the ref sensor and upon inspection(oscilloscope) found a poor ref signal at the dme. physically checking the sensor found it was loose in its holder! so I grabbed a 10mm socket and snugged up the bolt . with the bolts tight I attempted a new scope pattern and upon cranking I heard the flywheel smash the sensor (oh joy). of course the sensor tip got jammed on the flywheel and could not be retrieved . after consulting with the client and explaining the problem it was decided that the clutch needed replacement anyway and some deferred maintenance (oil leaks, water pump ,t belt,rms oil pan gskt etc) were needed anyway. when i removed the sensors I noticed that instead of adjusting the sensor bracket to obtain the .8 mm clearance they filed down the sensor on its mating surface where it meets the bracket!!. I ordered up the parts, bead blasted the sensor mounting bracket and replace all of the stripped fasteners. I reset the clearance to spec and was rewarded with a clean ref signal 5.6 volts) and clean rpm sine wave (4.34 volts) at the dme . i didnt trust the previous wiring repair so I stripped the connector and verified the repair and cross checked for continuity to make sure the wires were all separated properly . the result was a no start no tach bounce issue (still). i performed the dme/klr test plan from porsche and found it passed . checked clarks procedures as well and found no problems. I pin checked the dme free style for powers and grounds and found all okay. I had replaced the dme relay with my known good test unit to no avail when the client showed up with a new focus 9 diagnostic relay. f9t states all leds should light up when the ignition is on but only 2 lights do, the fuel injection pwr and ignition switch. the other lights fuel pump and fuel pump control only light under cranking .fuel pump runs fine when bridged at the relay. I thought I had found the problem when I discovered a open circuit from pin 85b to #3 pin on the 14 way connector but when I jumpered with a test wire no change occurred (only lighting on cranking or if I ground the terminal . I bypassed the alarm module # 911 637 104 02 by bridging 7 to 8 and 1 to4 or as my wiring diagram colors are not the same as clarks description pin 15 to pin 87a and pin61 to pin 61 all with no change. the tps test revealed an intermittent open on the closed throttle switch so I temporally grounded that pin to continue testing while we wait for a new tps. WHAT am I missing here ?? .at this point I am suspecting the dme but hesitant to replace it as i am unsure of the missing lights on the dme relay. the coil is powered as are the injectors which operate fine when grounding at the dme. where is the trigger? any and all help would be very much appreciated .

 

Well going down different rabbit holes that dont yeild results can be frustrating. Periodically check basics while doing it.. something simple as moving the 36 year old wire harness around can cause more issues. Just keep in mind that the turbos are batch fired. Be mindful of the injector harness. if they short together it brings the whole system down.. im working on a racecar right now and deep into the same issue but im working with an na harness and 968 engine in a turbo chassis… ugh..

 

Is it a pump/prime relay? If so do all four lights go on during the 3-4 second prime? The pump only operates when you are cranking or engine running (unless you have a pump/prime relay) so I do not understand why they say all four lights have to be on to start the car. I bought my ftech relay before the led feature was added. I would call them and ask for clarification.
As GPA951s says check the basics. I know you say the coil is "powered" but do you get spark from the spark plugs? If you are still getting no tach bounce this is usually caused by three things - speed/ref sensors, bad tach or bad coil driver in the DME.
What do you mean by the injectors operate fine?. Did you remove them and watch them spray? If not how do you know they are working?

 

Any chance that the reference sensors are FAE brand? I ask because the new ones I bought once were defective: the shield would have continuity with one of the leads.. this gave what looked like good signal and sine wave (except a little less reach on bottom half of wave) but DME would short when trying to start

 

thank you to all that responded. to gpa951 ,yes the turbo heat wrecks havoc on the harness . i did wiggle the harness on my continuity checks and found no issues (not a 100% test i know). to Break away, the only info I have on this relay is all 4 leds must light key on. all 4 lights do come on under cranking so maybe a call to verify is in order. as far as the basic tests the dme /klr test plan should have the basic circuits covered . the coil is indeed powered but not switching (the job for the dme and the reference sensor). there is no switching so of course no spark to the plugs (no primary so no secondary). the injector scratch test is just to test the wiring to the injectors and back to the dme and if the injectors click by simply grounding the control wires( ground as they are powered and the ground controls opening) at the dme. without the ignition pulse the injectors are not going to fire anyway so other than power and ground I am not concerned at this point whether they spray any fuel until I find the actual issue with the reference pulse as the dme wont latch until it sees an ignition pulse (collision safety). I agree with you on the causes stated and i will attempt to disconnect the tach from the circuit to see if it is pulling the circuit low. to Dan I am unsure of the manufacture of the sensors and i did read that fae brand was crappy (although the wave form looks clean under cranking) i will remove them and inspect for internal shorting.. I truly appreciate your input gentleman and will continue testing.......

 

I don't see the KLR mentioned. I would bypass the KLR and the DME relay.

 

yes both klr and alarm module are bypassed. i have 2 klr modules as the previous techs took a guess.

 

The information on ftech9 website is very poorly written and misleading. In one breath they say all four led's must be lighted in order for the car to start/run. In the next breath they say led #3 only lights when the DME commands the fuel pump on. Then they say led #4 indicates the fuel pump is powered up. #3 and #4 only happen during cranking/running conditions - both statements made by ftech9 can not be true at the same time the way a normal person would interpret the statements. They make it sound like all four must be on before you crank the engine with their wording.
Your problem is that you have good signals going into the DME but not coming out properly. Could you have the two signal wires plugged in incorrectly? This would still light both led's but would not run the car or give a fast enough signal to the coil driver to provide an output that would move the tach. You mentioned the ref sensor hit the flywheel and chewed it up. Did you mean the teeth on the flywheel or the stud sticking up? If the teeth then it is a speed sensor, not reference. The only other thing I can think of for the signal not coming out of the DME is the #1 problem the DME has - the three solder joints on the coil driver. This will not allow the signal to "come out" of the DME and go to the tach and coil. This connection can come and go with temperature changes, so could pass a bench test at on temp but not work on another temp. Can you get your hands on another DME to test? Just because a part is "repaired" or "new" does not mean it works.
Good luck

 

I agree that the web page is misleading and this is my first encounter with this relay. I have sent an email to them and have received no response ( makes it difficult to purchase a new solid state dme from them). my testing dme relay has pigtails I fabricated so i can see what circuits are active. definitely dealing with the ref sensor signal here , the trigger pin on the flywheel was inspected when I replaced the clutch (5mm protrusion). I have double,triple checked pin orientation on both sensors to the dme as a positive slope is required for the dme. totally with you on signal in none out dme as this was where my testing led me. the dme has a security seal or it would be already opened up. the client seems to have more faith in the rebuild than myself ($) but you can only test so much until you gotta try something different. small community here so no real testing dme available to borrow. many thanks for your input, working as a one man shop I don't have anyone to bounce theory's off of so i find this community very valuable. I will post my findings here as I detest threads with no conclusion......

 

Can you scope the coil output pin on the DME from the backside while cranking to verify the signal is truly coming back out of the DME? Breakaway might be on to something with an intermittent cold solder joint on the drive transistor or somewhere else in that one circuit. I don't know what the 'tests' look like that a rebuilder puts the DME through when they rebuild, but the solder joint could have given up the last little bit when it was in return shipping or something like that.

-Scott

 

yes, thanks Scott, there is no wave form coming from the dme to the klr . i am contacting the dme rebuilder for any information as i feel the problem is the dme.

 

problem found , vehicle runs! thanks to all for their input. I 1st repaired the open circuit from the dme relay to the dme. after contacting the dme rebuilder and finding out the repair was not under warranty I convinced the client that breaking the security seal to inspect the boards was the next logical move. upon disassembly to check for cold joints i noticed that the prom(chip) was poorly installed with many pins bent over and not in place in their respective terminals. what's funny is the rebuilder stated to the client that he had a performance chip installed, that he did not recommend that chip and he would replace it with a factory chip. the performance chip remains. this poor repair has cost the client many $ and me a lot of time. Early in my career I misdiagnosed a no start on a 928 as a faulty dme instead the open circuit that was actually causing the no start. since that embarrassing mistake I never blame the computer until I have totally inspected all inputs. thanks again.