DME/DME relay troubleshooting help

[dstudeba]

Hopefully this is something simple that I am missing but I could use another brain on this.

1983 Porsche 944 hasn't run for a while due to lack of time and changing the fuel tank.

Car turns over but doesn't catch
Can't hear the fuel pump come on
Two DME relays were bench tested as working and installed but didn't help
Jumpered the 30, 87, 87b connectors, fuel pump runs and car starts with foot on the gas.
Car will run rough with foot on the gas but not idle with foot off the gas.
Tested speed and reference sensors in the engine bay and at the DME connector. Both resistance and AC voltage with ignition on were within specs

Tested the current at the DME relay slot for 86, and 85 by putting the leads in 86 and 85 when the ignition was turned which showed DC current which should close the switch. I did the same thing with 86 and 85b to mimic the closed ignition switch in the relay and also got DC current. (note, I am an electrical novice so I could be testing the DME output completely incorrectly)

So I am at a loss as the relay tested fine, the sensors tested fine, and it appears that the output of the DME/ECU should get the relay to switch and the fuel pump to go.

I also realize that there is likely more than one problem, the fuel pump/starting problem and the not idling problem.

Thanks for any help, it is much appreciated.

[wwdwgs]

Corrosion of the socket where this DME relay plugs in?

[dstudeba]

Originally Posted by wwdwgs

Corrosion of the socket where this DME relay plugs in?

Thanks for the answer. I don't think the socket is the problem since the jumper I used to make it work was connecting the prongs on the DME relay and putting it back in. (The paper clip method from Clark's Garage)

[944 Ecology]

If you make a two wire / three male spade connector jumper wire and connect the three pins marked with circles, you will eliminate the dme relay as a possible culprit. If the car starts, it was the relay. If the car does not start, it is something else.

[dstudeba]

Thanks for the reply. I followed the jumper directions from Clarks Garage and the car started. However I bench tested both relays and they worked as expected.

That is why I am wondering if there is a problem with the computer. The jumper puts voltage straight from the battery to the fuel pump. It is my understanding that with the relay the DME controls the voltage to the relay, but I could be wrong.

[944 Ecology]

The pin marked 85b must receive a ground from the ECU to power the fuel pump.

[dstudeba]

Thanks George. With the relay pulled I put my leads in 86 (voltage from the ignition) and 85b (ground from the DME). When I turned on the ignition there was DC current so my conclusion was that the DME was providing proper grounding.

Is this the correct conclusion or am I misunderstanding something? Thanks!

[944 Ecology]

I'm not clear on what is the diagnostic situation. Please explain for me:

1) with the DME relay jumpered as clarks garage describes, does the car start and run and stay running?

2) With the DME relay not jumpered, does the car start and run and stay running?

[dstudeba]

With the DME jumpered the car starts and stays running with the foot on the accelerator.

Without the DME jumpered the car doesn't start.

I believe that the car not idling is due to a vacuum leak which I am going to try to find next.

[944 Ecology]

If that is the case, you have verified the circuits shown in this picture as pins 4, 8, and 2. Any of the other pins not having the proper input will result in a non start situation. (Of course, it could be that you have a bad DME relay, but you insist that they are good.)

[dstudeba]

Ok, thanks, I appreciate all the work. I am going to switch to trying to fix the vacuum leak and then come back to this. I might have to build a test rig to test the current flow with the relay hooked up to the car to see where it is flowing/not flowing.

[Monkey Wrench]

maybe its just the picture but it appears that that wire jumper has clear plastic insulation covering it,

if you try with a bare spade terminal you can feel the tension of each terminal separately, when you plug in the relay it is hard to feel since you are inserting more than one at once.

if the fuel pump was ever stuck or drawing excess power for some reason it may heat up that terminal and make it loose it's tension.

not hard to connect a couple of LEDs or small test lamps to show you real time when fuel pump as power or when the other circuit is latched, or the solid state relay comes with some LED's on it.

that way you can watch in real time and see when the circuit has power, that might just simplify the troubleshooting a little.

The solid state one may be worth the money or if you don't want to spend a couple of tiny lamps are cheap, I think I see myself tempted just to leave them there, maybe locate them in some spot I can easily look, so I can see at a glance, then if the FP has no power you know straight away.. the rest of the time maybe just tuck it away out of sight because you don't really need to monitor this constantly, just if you have a no start issue.

in my volvo 740 I just connected a tiny bulb in parallel with the pump wires, if its lit the pump is running ...its in my engine bay. the load of that lamp is so miniscule in comparison the load of the pump itself that it's power draw is a non issue. now under a no start condition , weather of not the pump has power is very obvious to me from the get-go. if I saw it out under a no start I;d go straight to troubleshooting the relay circuit. , if it is on I can forget about that being an issue. the 944 has a bit more complexity because it involves more than only the pump.

[Monkey Wrench]

great diagram !

I don't know if one would do this but it appears that if one were to draw a horizontal line through that diagram it could be separated into 2 very normal easily obtained 4 pin relays.
so long as the coils and contacts were similar in size and current draw I think that would be electrically equivalent and then the Porsche would no longer need to have a proprietary "double relay".

separating the circuit in this way, at least on paper. may help understand it?

probably not worth the effort, for one car, but I think one could in theory, gut a bad relay, reuse its base and contacts, and install two 4 pin sockets that sit on top of the base of the old proprietary relay, meaning no wiring changes to the car itself. It would look like a plug in adapter. the only real advantage is that a simple 4 pin relay is probably always in stock at your local parts dealer. In reality once it is fixed it may be fine for years .

[wwdwgs]

The problem with these DNE (fuel) relays is the microfractures of solder between the post and printed circuit board.
That is caused by the heat, which is caused by the high current (if I understand it right).
So-called updated DME relays (by Uro) encase relays, but don't address the core issue.

[Jfrahm]

People have deconstructed the DME relay into 2 standard auto relays (and if you do this make sure you have a snubber diode on the DME control section). However it's not worth the trouble since we can get a quality part that should outlast the chassis these days and not have a daughterboard and silly wiring to contend with.

Re: the OP's problem I suggest jumpering the DME (ECU) control part of the relay and then looking to see if you get a strong ground signal for the fuel pump with a test lamp across 85b and 87 when cranking. Also check for switched +12v on pin 86.

I like to use a turn signal bulb or the like for a test like this, a DMM might see DC volts but there might not actually be a circuit that can carry current.

If the DME 9ECU) is not sending the signal to close the fuel pump relay then it could be a DME issue or related wiring. The DME relay jumper works around any issues with the signals on pins 85b and 86.

-Joel.