DME-Fuel pump rly intermittent signal failure

[mr brightside]

DME fails to switch the fuel pump relay ('86 924s), i've got an indicator lamp on the outgoing fuse and this is definitely the cause. The relay seems to have two sets of separately switched contacts and manually linking the supply to the pump will not result in engine starting, the other contact has to be doing something and it's got too complicated for me to safely override. The only clue to the root cause is that it is a cold start related problem. Does anyone have any ideas or a tried and tested override procedure as when i leave work today at 5.15 it's going to strand me again. Thanks.

 

[alwigley]

Take a look on Clarks Garage http://www.clarks-garage.com/ at the DME relay page under Garage Shop Manuals - requires a 3 way jumper to bypass and I think the 3rd connection yopu mention supplies power to the DME/injectors, which is why she won't start without it...... Don't forget to take the lead off once switched off otherwise you will get a flat battery !

Hope you manage to get home []

 

[mr brightside]

Cheers alwigley, i'm pretty confident the relay is ok i tested the contacts, the diode and the resistors this morning but i might order a spare anyway. I'm still suspicious about it being a cold start only problem but at least i can override it now.

 

[mr brightside]

After studying the excellent drawing on the Clark's Garage guide i think i might be dealing with a missing ground to the slave relay coil from the DME, i've heard there can be a time lag on some cars to allow the build of oil pressure? I've soldered a ribbon cable onto the relay pins and connected up to a bank of LEDs so i can see what's happening when it next goes wrong.

 

[alwigley]

Sounds a little more impressive than my Heath-Robinson emergency jumper cable that lives in my glove box..... let us know how you get on. Clarks Garage tends to be a very good port of call for this sort of issue.

Alan

 

[mr brightside]

Cheers Alan, i've turned my DME relay into a combined indicator unit with LEDs and an override switch on the slave coil ground which is the problem signal. I'll modify the Clark's drawing and add my wiring and post it at some point, i still don't know what the fault is but i'll be able to diagnose and override it next time it happens.

 

[tref]

The indicator lamp on the fuse out is a brilliant idea - wish I had thought of it, but I for one would be interested to know what indicators you are rigging up now.

One to watch out for, I have an intermittent DME related "no start" scenario on my S2 - if the passenger door is opened/closed sometimes after it will refuse to start - the tacho needle doesn't "bounce" when turning over. Waggle the DME relay, and off it goes - a dodgy connection I am sure, but like all intermitent faults...

Just worth keeping an open mind to weird causes to some problems.

 

[mr brightside]

ORIGINAL: tref

The indicator lamp on the fuse out is a brilliant idea - wish I had thought of it, but I for one would be interested to know what indicators you are rigging up now.

One to watch out for, I have an intermittent DME related "no start" scenario on my S2 - if the passenger door is opened/closed sometimes after it will refuse to start - the tacho needle doesn't "bounce" when turning over. Waggle the DME relay, and off it goes - a dodgy connection I am sure, but like all intermitent faults...

Just worth keeping an open mind to weird causes to some problems.

I've finished the wiring now so i'll take some photos of it tonight hopefully

 

[mr brightside]

Here's what i've rigged up along with a modified Clark's Garage drawing of it. The red and amber LEDs are in parallel accross the primary and secondary coils and confirm the supplies to them, the green one is on the load side of the fuel pump fuse and confirms successful operation of the relay. The setup is geared more towards diagnosing faults with signals around the car and less concerned with problems that might arise inside the relay, although failure to light the green LED could loosely represent a relay fault. I've soldered four cores of a ribbon cable onto the relay pins for the coil LEDs and trapped a ring crimp under the fuse cap for the pump LED which is cleanly grounded to the car body via the fuse box mounting screw. The override switch is there to substitute my temperamental DME ground signal with the good one from the primary coil wiring. The box it's in is just a little flippy top box work buys their ICs in, which i've mounted with self adhesive clips which are trapped behind the bonnet pop surround. The resistors (6.4k) on the LEDs are to drop the voltage, i could only get hold of 3v LEDs from work.

 

[Northern924]

Looks good.
And for those don't have the skills for that kind of work just remember a Vauxhall Nova or Aastra GTE one will work []

 

[Eldavo]

In place of the Porsche DME? Could be a cheaper spare if so!

 

[peanut]

this subject has been covered comprehensively on this board very recently. There is no point in re-typing it all. Its worth doing a search .

The DME relay is grounded through the ECU.

The ECU will not ground the DME relay unless all the relevant sensor inputs are met ie crank speed (200rpm+ min) water temp, air temp , etc.
Terminal 30 on the DME relay is always live 12v+ so removing the DME relay and using the 3x way bypass will effectively over-ride the ECU and all the sensor inputs.

The fuel pump will not get a live feed normally until the ECU sees 200rpm plus all the sensor inputs at which point it will ground the DME relay coil and operate the fuel pump and pulse the coil to feed the ignition circuit..

I would recommend you google for Bosch Motronic fuel injection system and read about how the system works . There is usually a fault finding section with the faq sheet.

Re your diagram.
It is the relay coil terminals 85 and 86 which are grounded through the ECU on ignition switch on , sending the 12v+ sitting at terminal 30 through switch 1 to operate relay coil 2 which closes sending 12v+ to the pump.

When you switch the ignition on the ECU powers the fuel pump for approx 1 second then switches the DME relay off again. That is sufficient to pressurize the fuel system for starting. When the engine is cranked provided all the sensors have the correct input the ECU then operates the DME relay again and pulses both the fuel injectors and the coil . So your fuel pump won't get a continuous live feed until the ECU is satisfied under cranking and running

 

[mr brightside]

ORIGINAL: peanut

this subject has been covered comprehensively on this board very recently. There is no point in re-typing it all. Its worth doing a search .

The DME relay is grounded through the ECU.

The ECU will not ground the DME relay unless all the relevant sensor inputs are met ie crank speed (200rpm+ min) water temp, air temp , etc.
Terminal 30 on the DME relay is always live 12v+ so removing the DME relay and using the 3x way bypass will effectively over-ride the ECU and all the sensor inputs.

The fuel pump will not get a live feed normally until the ECU sees 200rpm plus all the sensor inputs at which point it will ground the DME relay coil and operate the fuel pump and pulse the coil to feed the ignition circuit..

I would recommend you google for Bosch Motronic fuel injection system and read about how the system works . There is usually a fault finding section with the faq sheet.

Re your diagram.
It is the relay coil terminals 85 and 86 which are grounded through the ECU on ignition switch on , sending the 12v+ sitting at terminal 30 through switch 1 to operate relay coil 2 which closes sending 12v+ to the pump.

When you switch the ignition on the ECU powers the fuel pump for approx 1 second then switches the DME relay off again. That is sufficient to pressurize the fuel system for starting. When the engine is cranked provided all the sensors have the correct input the ECU then operates the DME relay again and pulses both the fuel injectors and the coil . So your fuel pump won't get a continuous live feed until the ECU is satisfied under cranking and running

I haven't had time to go through the other threads yet, of which i'm sure there are plenty. I was in a pickle at the time and didn't want to be sleeping at work so just started a new thread for some quick info. Thanks for the heads up on the revs thing i'm still trying to figure out what conditions have to be satisfied to enable the ground signal at 85b, which is the one i think i am missing although since setting up my indicator box it's not failed to start on me! I've got the Haynes Fuel Injection manual so i'll have a look in there some time as the 944 Haynes manual electrical schematics are diabolical.

 

[peanut]

you might find the following links useful.
http://openpdf.com/ebook/porsche-944-motronic-1-2-pdf.html
This troubleshooting guide is brilliant and whilst not specifically for the 944 the operation and principles are exactly the same.
http://www.hiperformancestore.com/Ljetronic.htm