84' 944-No start until pull/replace DME Relay
 

I have a 84' 944 that will not start until I pull the DME relay and then replace it. I can do this with the original DME Relay and a fresh one I just purchased. It will start and run fine, then if I turn it off it will crank over, sound like it's firing, then does not follow through with a start.

I followed the Clarks Garage test procedures for testing the wiring where the relay plugs in and even checked it at the computer.

Any thoughts on next thing to troubleshoot?
Does this sound like a bad wiring connection?

Prior to this the car ran fine. We replaced the fuel pump recently because one of the fittings started leaking. Then did the pump and the in fuel tank filter/screen. We do get fuel to the rail.

One post suggested testing/changing out the speed and position sensors. Upon trying to test, the connectors disintegrated, so we just replaced those.

It will run fine once it does start. Just not sure the best procedure is to pull the DME relay each time.

Sounds like a connection/wiring issue. You can likely just wiggle the relay instead of removing it for the same effect.

Having an Ftech9 diagnostic DME relay would help nail it down quickly. It has LEDs on the inputs and outputs of the relay, so you will be able to tell which wire is having an issue. Or, if you are handy in that way, you could add LEDs to one of your existing DME relays.

Why not test with a jumper wire and see if you can restart then?

Do you get tach bounce when it won't start?

It might be a DME issue with the driver circuit for one part of the relay. You could also try disconnecting the battery for a minute or two and then restarting to see if de-powering the DME that way also affects starting.

i would be careful about pulling the relay out too many times, it could damage the relay board. At the least put some lube on it

Like Jfrahm mentioned, create a jumper as is clearly outlined in Clarks Garage and test. It might be wiring to/from the relay loose, so if you can replicate it with the relay bypassed, next step is to start chasing the wiring, especially the connections at the fusebox/relay

small indicator bulbs might help here because you can wire them in and then see when it fails, if the DME has power and the fuel pump etc. a lamp is easier to see an intermittent issue with than a meter display.

if there is ever high power, like a short, sometimes it can get the spade connectors hot enough to loose their spring tension. that may also burn the contacts of a relay , even a new one. so you could have two bad relays if you had a pump that failed and stuck or similar and drew a higher than normal load.

if you heat brass and quench it it goes soft. if you heat it and cool slowly that may help restore the temper or "springyness" to the terminals.

you might try with a hunk of wire and a spade , press it into where the relay seats to feel the connectors on an individual basis. See if any seem to have gone soft or maybe just need a squeeze up to be tight again. hard to tell with the relay itself, as it will have more than one that seats in unison. easier to get a feel with just one wire at a time.

Thanks for the feedback. I've been out of town for a week. Getting back to it now.

The remove/replace option stopped working after a few more attempts. I didn't see the concerns early enough. But also wasn't thinking it would be a permanent solution.

I suspect some wiring is loose, but will also see about jumping the circuit.

The Tac has been jumping since we got the car. Our thought was that something behind the dash was loose as we could tap the gage and often it would come back to life. (not sure how scientific the approach is)
We did replace both the speed and position sensors going into the bell housing. Still no-start. The cable connector on the sensor side just crumbled when we touched it. We did not test the continuity of the other half of the circuit. Would that be a good place to check first? or just pull down the wiring loom behind the fuses to check for loose or damaged wires (under driver side dash)?
We did get the battery tested. It tested fine and continues to crank.

If you're still stumped, I would replace your current DME relay with this one. Now you can get an indication which of the four electrical paths to concentrate on.http://forums.pelicanparts.com/uploa...1682404400.jpg

Hi, please bypass the relay as we mentioned earlier so you can confirm whether the DME relay is bad, or if something else is the cause. Do not associate other peculiar behaviors to the no start issue until you go through the proper testing procedure.

DME relay bypass:
DME Relay Information and Testing

If the bypass doesnt work, the next thing to check is the basics - do you lack spark or do you lack fuel. This guide will lead you through the steps to diagnosing the no start issue:
https://www.clarks-garage.com/shop-manual/ts-01.htm#fire

Latest update - continued with the theme that no spark or no fuel. Reading through the Clarks procedure came across one step that stated that an over-pressure of the fuel system could cause a no start as the injectors draw too much and the computer shuts them down.

The test was to disconnect one of the injectors to see if it started. All be it on 3-cyl. It took a few tries but did start, and was repeatable again with a few tries. I did have to use the jumper to get to those results. Getting it to start there points to a fuel pressure regulator issue.

Is there anything else I should be verifying?
My one concern is that I recently replaced the fuel pump. Should a new regulator be able to handle the higher pressures? Or did I get too strong of a pump by mistake?
(I am working on getting a fuel pressure gage.)

So if you had to jump the relay in order to get it to start (even on 3 injectors), how do you know its the injector issue and not an issue with the DME relay?

Did you try starting with just the DME relay jumpered and leaving all injectors connected? If you havent done that you should. Test each system independently, dont add multiple variables.

Get a pressure gauge on the car before making any assumptions about fuel pump or FPR. A bad FPR will sometimes (but not always) have fuel leaking into the vacuum line, something you can check before getting the fuel pressure gauge.

It's been a while. Following the string of looking for spark and fuel. I pulled the injectors and did a full cleaning including replacing the cap, O-rings, and "screen". I did replace the fuel pressure regulator as well (I had bought before checking the last note in this string.)
On 2nd try, after letting the fuel pump run, it fired right up. Then, as typical of late, it would not restart.

I continued back through all the checks. Confirmed the DME relay was good. Checked voltage at 18 & 35 on DME computer connector - good. And a few others.

One of the checks in the Clarks Garage Fuel-16 procedures is to check the voltage at the injector plug terminals. It says that 1 should be approx 12vdc and the other should be at 0vdc. It does not say what to do if both are reading 12vdc with the ignition on. I checked 3 of 4 injectors, all had the same problem. Thoughts? or did I read the procedure wrong.
Is there a place on the wiring harness that may be prone to issues?

Is there another test to try?
When cleaning the injectors, they all clicked when voltage was applied 12v on one side and ground on the other.

Appreciate the continued assistance.

Sorry to hear you are having this problem. when it won't start try jumpering the dme relay. as you plug in the jumper listen for fuel running in the fuel rail to confirm fuel pump. now try to start it . if it won't start then check for spark by removing #1 wire, plug in an extra spark plug, with a aligator lead from the spark plug threads to ground, and then crank and observe if here is blue/white spark. remove the dme jumper afterwards so as to not leave the fuel pump running.

if there is spark this means the dme is probably ok however there have been cases where there is spark but no injector pulse due to a dme problem.

edit: constant +12V is normal on the injectors, the dme will provide the negative 12V when it pulses the injector.

The dme jumper is simply a 3-way Y connection from pin 87 (12v power) to 87b (fuel pump) and pin 30 (dme power)

More info on the Dme relay and jumper here: http://www.clarks-garage.com/shop-manual/fuel-05.htm

The reason we are so insistent on this is because a bad DME relay is highly correllated with hard warm startups, because the heat causes them to not work. Once they cool down they work. The fact your thread title states that the car starts when you remove and re-plug in the DME relay, makes this the #1 theory on your issue.


YET... you still refuse to jumper the relay. The next time the car acts up, jumper the relay. Report back. Im not going to respond to this thread any further until this is done.

That is not how you confirm the DME relay is good.

I have used the jumper. Currently I can get it to start with all 4 injectors plugged in on the 2nd-5th try relatively consistently. With the jumper engaged, the fuel pump is clearly audible (fuel pump new @ 2weeks prior to issues beginning).

I used the Clark's FUEL-16 procedure for the following -
DME Relay check - good
Confirm ignition switch voltage at DME relay plug (on fuse panel)- good
Confirmed voltage to the DME Computer plug - 18 &35 - good

I have replaced both the speed and reference sensors after attempting to confirm resistance caused the original plugs to disintegrate.

I confirmed spark with a spark tester.

I now have a fuel pressure test gage that I'll be installing this weekend.
I also ordered one of the F9 solid state DMEs with the LEDs to see if that helps. That should arrive today.

But, at present, neither DME will get it to start more than 1 time. The jumper will get it to start multiple times.
I'll provide an update this weekend - for sure on pressure check, and hopefully with a different DME.

A bad DME is really kind of the last resort option once youve gone through anything else, regardless sounds like thats not the issue as youve already tested with a known working DME.

The fuel pressure test will hopefully reveal some data. The fact that the car starts easier (but not perfectly) with the jumper in place is interesting and perhaps points to multiple issues. One of them potentially the DME relay, so its good you are getting the F9tech relay. I have one and they are apparently 100x more reliable than analog. It will never be an issue again (hopefully), assuming thats one of the issues.

A couple other things to check (apologies if this has already been mentioned):
-DME temp sensor
-AFM barn door voltage test
-Fuel Pressure Regulator (this will be revealed during the fuel pressure check if too high)
-Fuel Pump Check Valve
-Fuel Filter or fuel tank strainer clogged (this would show up as a drop in fuel pressure during cranking/running as the fuel cant push past the filters quickly enough to sustain pressure)
-Injectors clogged
-TPS (does it click when you first open the throttle?)
-IACV (should snap open/shut cleanly with 12V, if its sluggish or only open/closes partially it need to be cleaned)

All of these are related to fueling and air/fuel balance, especially during startup, which requires a richer mixture than when running. The DME needs to know when the engine is cold (DME temp sensor) and at idle (TPS switch). It also need to measure the amount of air coming into the engine (AFM) and have the correct amount of air bypassing the throttle plate (IACV). And of course needs the appropriate amount of fuel to balance those (injectors, fuel filter, pump check valve).

The clarks reference is only so useful as it is for a 85.5 or later car. here is the wiring diagrams for 83 and up, should you need it, e.g. the dme relay is not getting the correct signals... I have never had to chase down this angle, and I am only familiar with the 85.5 and later cars. I don't know if the pin numbers are even the same?

https://web.archive.org/web/20070314120555/http://cannell.co.uk/944_workshop_Manual/Vol4-electrical.pdf

2nd try at posting. Apparently the forum doesn't like me...
Weekend update.

I did receive the new F9 DME Relay on Friday. We installed and it fired off on 2nd try. I has a fuel prime component in the solid sate logic, so the fuel pump was easily noticed when I put the ignition to Run (prior to start). I was able to get it to restart with the new DME, however, it still took a couple of tries. And it seems that it will either start in the first few seconds, or not. Prolonged use of starter didn't help it kick over.

Observation -
the red - Fuel Injector led - was "on" when I plugged it in, and remains on after the ignition is turned off.
the green - Ctl Fuel Pump led - turned off after start when at idle. But the Blue Fuel Pump led remained on.

As for the other tests -
1. Fuel Pressure - I need to find a connector for the fuel pressure gage I have. Any suggestions are appreciated. I'm not sure I'm up to the task mentioned in the Clarks procedures.
2. Suggested items to check:
-DME temp sensor - added to list. I see the procedure.
-AFM barn door voltage test - added to list. (We have noticed a rich condition in the exhaust at start...)
-Fuel Pressure Regulator (this will be revealed during the fuel pressure check if too high) - noted for when I can gage hooked up.
-Fuel Pump Check Valve - recently replaced with the fuel pump.
-Fuel Filter or fuel tank strainer clogged (this would show up as a drop in fuel pressure during cranking/running as the fuel cant push past the filters quickly enough to sustain pressure) - Fuel tank strainer was just replaced at same time as fuel pump. I have a replacement fuel filter available and am going to change that out.
-Injectors clogged - just cleaned the injectors and replaced the consumables - O-rings, spacers, and injector screens.
-TPS (does it click when you first open the throttle?) - will check
-IACV (should snap open/shut cleanly with 12V, if its sluggish or only open/closes partially it need to be cleaned) - will check

Thanks all for the additional guidance and reference info (electrical diagrams).

One other note - with it started, we did take her for a couple of runs. No hesitation was noted as we drove around.

You need to get an adapter from M16 to whatever the thread size is on the pressure gauge

I have not been paying close attention but fuel pressure seems unlikely.
Have you looked at the plugs after it cranks but does not start?
Did you try a shot of starting fluid when it cranks but won't start?

Feels like a typical wet plugs, blown FPR or damper diaphragm issue to me (or wonky temp reading skewing the mixture rich).

Id hook a small 12 V test bulb across the primary coil so you can see if it's blinking/ flickering when you have a condition of no start. If the light bulb stops flickering then it's a spark issue. work back through checking the spark path through the powerstage transistor and to the DME.

Id also pull the fuel rail and watch the fuel flow, see if it looks like the injectors are always spraying normally by trying ot start a few times, but do that outside and take precautions of the gasoline explosion or fire risks. cold engine , no sparks dont collect any substantial amount of gasoline. gasoline rated fire extinguisher at the ready.

maybe buy a noid light so you can see if the injectors are getting their pulse during this time when you have a no start condition. a fuel pressure gauge may help show the pressure if the raw gas makes you nervous.

i think either your fuel delivery or the spark is being interrupted intermittently. Knowing which will tell you which direction to go in. no point confusing the issue without making that differentiation
a fuel delivery issue could be either no pulse at the injectors or no fuel pressure to them.

I don't think you have bad injectors because they wouldnt all fail so it would still start even if rough. you drove it so the fuel pump can operate, that does not mean you cant; have intermittent power to the pump. Id put a small test lamp in parallel to the pump, permanently, so you can easily watch that lamp to make sure it is on when the problem occurs. it will always be on when the pump is then. Just use a little bulb like a dashboard illumination lamp. the power draw is insignificant then.

sometimes pumps come with new wires and those crimp connectors, id make sure your connections there are really good and solder them. a bad crimp there can stall you in a intersection or when pulling out in traffic , That can cause a car crash.

the lamps on the new relay might do that but can you see them when you are cranking? thats why I'd put them in view of the drivers seat. - so you can note if it happens but intermittently. I think yours , like mine is down under the drivers side and out of view when cranking unless you are a contortionist. I suppose you could extend the relay to a better position if you wanted but using an old relay as the "plug in" and then spade connectors on the end of your extension cord. thus move that relay in view of the drivers seat.



if you remove the fuel rail, also check any related ground wires, a bad ground could cause this and that's as likely as a no power to it situation.

you can wiggle wires when it's running , if you found a bad wire or connection and wiggled it, it might cause it to die or sputter.

maybe the immobilizer is preventing spark, my car has a key-switch just to the right of the drivers door key.. you can jump it, but I'd find out first if it actually has an intermittent no spark or no fuel condition, or not.

when you find the cause it will then be safe in traffic and this will probably be an easy repair, the trick is isolating the actual problem and then you can relax and feel better, until this is resolved it's a danger. You really don't want it to quit suddenly so until it's resolved refrain from depending on the power as it might stall. no pulling into traffic suddenly , etc.

if you swap a part and then it starts up ok and you think you have it resolved, Id recheck by putting the bad part back to verify, the reason is that, for your safety, you want to know what it was so you can be confident that it's fixed and not just hidden for a while, only to return at a bad time.

Good thing you have patience.

I don't have one of these relays, so I cannot answer if this is correct. You should read whatever instructions came with it to interpret.

More importantly...

What do the LEDs indicate when you are cranking and it does not start?

How do you test for a blown FPR or FPD? It is discussed extensively here: http://forums.pelicanparts.com/porsche-924-944-968-technical-forum/1006476-fuel-pressure-regulator-faulty.html

one problem with the shotgun approach ( replacing multiple parts to try to solve a no start issue) is that you can introduce a problem , like if that speed sensor or one of the other parts you replaced is bad. If I do that I'll often try the part note the difference and then if it doesn't; solve anything , remove it or the duration of the problem solving, because there is probably just one reason and you really dont want it to not start for more than one reason at the same time.

when it does start, does it fun perfectly well or is it running a bit rough? if it runs perfectly when it starts you might rule out a vacuum leak and look for a bad ground or similar wiring issue. a bad wire connection somewhere can easily go from running fine to not at all, a bad adjustment or vacuum leak etc might cause it to start only sometimes but not run quite right. That's why how well it runs when it does start might be a clue.

last night I had a problem starting my volvo 240 after replacing the head gasket. turned out I had a unplugged temperature sensor wire and that was causing it to start about half the time and switch back and forth between starting sometimes ( along with a check engine light) and not starting sometimes.. . I think the ECU saw that there was no temp reading, it got all confused and said no way I'm shutting down..

once plugged in right, problem solved. Maybe the porsche is similar where a bad temperature sensor or bad wire to it might cause a no start issue too. If I took a more analitical approach I may have found that he spark was not there sometimes and went down that path to find it was because the ECU was unhappy. on the porsche thats the DME, but similar ..

someone else here might know if a bad temperature sensor could cause a no start issue . I dont think the temp sensor would have been the first thing I'd suspect to cause this sort of issue. My volvo has two , I think one for the ECU and one for the dash gauge. I bet it would run with out the dash gauge sensor, bit not without the one that reports to the ECU

> someone else here might know if a bad temperature sensor could cause a no start issue

Yes it can cause the ECU to over-fuel and wet the plugs. They tend to fail suggesting it's very very cold.

JD, can you recheck this and make sure it is correct? Also what lights are on when it is running?

http://forums.pelicanparts.com/uploa...1682404400.jpg

Apparently I haven't responded in a while. Life got in the way - graduation and other family items all on the good side.
I've been trying to get the right fitting for the fuel rail. the M16 is too big for what I have.
I didn't notice that this thread got to a 2nd page. Thanks for all the input here.
I've got time this weekend to hit a few of the recommendations. I'll run through them and let you know what I get. I think the near term focus is on the fuel and spark suggestions. Tried a start from cold this morning and took 5 or more attempts then a rough idle. It did restart, but only after a few tries. So, I'll keep troubleshooting. More later...

I did continue troubleshooting. Pulled fuel rail and set up a safe environment to test the injectors. They all appeared to spray about the same, and no drips after we stopped.

I pulled the plugs and all looked about the same - half black the other half clear. Thinking they're slightly fouled I replaced them. Still takes 3-4 tries to get a start.

When it does start, it will run a little rough in the 800-900 rpm range. Smooths out the closer to 1,000 rpm you get.

The solid state DME relay with the 4 colored diodes - top to bottom
Blue - Ctl. Fuel Pump - on in run and start position
Green - Fuel Pump - off in run and start position (didn't rev or put under load)
Red - Fuel Inj. Power - on in run and start position (also remains on constantly)
Yellow - Ignition SW - on in run and start position

The Blue - Ctl. Fuel Pump relay will be on when you turn to the run position. You audibly hear the fuel pump run for 3-4 seconds. It then cuts off as does the light. Upon turning to the Start position the Blue light comes back on (can't hear fuel pump over starter). It remains on when you let the key go back to the run position. The Green "fuel pump" diode does not come on at all during the starting process.

Going to pull up the temp sensor check and also try some of the vacuum system tests. There is a new FPR on the return side of the fuel rail. I need to find the correct adapters for a fuel pressure gage. the M16 mentioned earlier was too big.

You dont "need" the adapter (although its best that you do) to do a fuel pressure test as shown in this video:
https://www.youtube.com/watch?v=Dq6anWRYlGA&t=150s

Have you run through the tests on Clarks Garage?
https://www.clarks-garage.com/shop-manual/ts-01.htm

A quick google search reveals the fitting is a M12x1.5...
https://www.google.com/search?q=porsche+944+fuel+rail+fitting&rlz=1C1OPNX _enUS940US940&oq=porsche+944+fuel+rail+fitting&aqs =chrome..69i57j33i160j33i299.4176j0j7&sourceid=chr ome&ie=UTF-8