450SL will not run when EFI coolant temp sensor is plugged in

[capeterson98]

I'm working on a friends 1972 450SL (107.044) because it failed emissions testing. She took it to a Mercedes mechanic who prepared it for emissions retesting, but my friend never took it back to be rechecked and the car sat for three years. I got it running last month and noticed the EFI coolant temp sensor (next to the thermo time switch) was unhooked, so I hooked it back up and the car would not start. I unhooked it then started the car and plugged it back in and the car dies immediately. I tested the two prongs with my volt meter to see if one or both were grounded and they were not. I found the specs online, pulled the sensor and using cold, warm and hot water, tested the resistance and it matched the specs I found on the web. I'm guessing the ECU is bad, but not sure. Does this sound like a good diagnosis? Does anyone have any other ideas?

Thanks.

[ctaylor738]

That's one possibility. What are the values at the ECU end of the harness with the sensor connected?

[rowdie]

Could critters have gotten into the wiring while it sat? They just love wiring. They ate my '72 in the barn. That is why it is the donor car for my '73. I now put critter feeders in anything that doesn't regularly move. (Kind of describes me)

I suppose you checked the blue wire and made sure it is not grounding. I think it is also tied into the Aux fan somehow so you might try disconnecting the sensor on the dryer.

[capeterson98]

Chuck: I'll check the values at the ECU this week.

rowdie: I had to replace the plug wires because they had been chewed, but I couldn't find anything else that had been chewed. I did check the blue wire and it is not grounding. I will try disconnecting the sensor on the dryer to see if that makes any difference.

I will reply when I have some results.

Thanks guys.

[meltedpanda]

sure sounds like its going to ground. On the 72 I thought the ECU really only relayed lean rich mixture information

[rowdie]

Ron,

Don't know exactly what it controls but every effing sensor leads back it. Been upside down in the passenger floor testing till my eyes crossed only to find I effed up the cam timing.

[capeterson98]

Last night I checked the resistance of the blue wire from sensor to ECU and it was 0 ohms. I tested the resistance of the brown wire from sensor to negative terminal on battery and it was 0 ohms. I didn't check the sensor values at the ECU yet. Since there is no resistance in both wires, I assume the values are the same as when I had the sensor out of the car. When I get a chance, I will check those values just in case. I don't see any connection between the ECU temp sensor and a sensor on the dryer, at least on the wiring diagram, but I haven't physically tried to see if disconnecting the sensor on the dryer has any affect either.

I'm starting to think power is being cut to the fuel pump. If the sensor is plugged in, the car starts up just for a second and then dies. When you turn the key, the fuel pump turns on for a second, but when cranking it may be getting turned off by the ECU. Tonight I'm going to check the voltage at the pump to see if it is getting cut off.

[rowdie]

Check the fuel pump thread by RC4U12 below, sounds like pump is operating as it should.

[RC4U12]

for the pump relay to work in the ecu properly it is tied to the tach signal and if it does not see rpms and started it will not kick it in again..so this may be your problem if it starts and dies its using the gas from prime..if yo remove the relay for the pump and plug a bannan plug in the upper right corner and another tied to it in the one next to it you should be bypassing the cutoff and have pump on all the time..{at least those are the holes in a SL}Jeff

[capeterson98]

I hooked up my fuel pressure tester in-line between the fuel damper and the right fuel rail, started the car then plugged in the EFI temp sensor. The car started to die, but the pressure stayed steady at 32 psi. when the car died I could here the pump make it's last bit of noise before it shutoff. The fuel pump is definitely getting consistent power.

I disconnected the sensor from the dryer and the car still died when I plugged in the temp sensor.

I also used a fuel injector testing probe, that senses the opening and closing of the valve in the injector, to see if the injectors stop firing when the car starts to die. I tested one injector on the right side and it was firing through the die down of the engine until the engine stopped. I'll check the rest tonight.

I'm still leaning toward the ECU being bad, but since temp sensors are cheaper I might try a different temp sensor before I buy another ECU.

[rowdie]

I am not great at wiring but from what I can trace this sensor is connected in some way to the switch over valve. The switch over valve has something to do with timing and it may have a bad relay. I find mention of a relay on the kick panel but can't find out which one it is.

When I replaced my sensor my auxillary fan started running, which it had never done before. Now it runs all the time when the ignition is on and I don't even have the AC connected.

[capeterson98]

rowdie: I think the sensor you are talking about looks to be the Temperature switch (62 degrees Celsius) that has a brown and blue wire connected to it. On the wiring diagram it shows the brown wire goes to ground and the blue wire runs up to the aux fan relay and from that relay there are some connections to the switch over valve relay. This switch definitely affect the fan operation, however it is not the EFI coolant temp sensor. The EFI coolant temp sensor has a brown wire to ground and blue with white wire to the ECU.

[Gurunutkins]

pull the ecu temp sender and put it in different temp water (tested with your trusty meat thermometer) if it doesent read within the range you need a new sensor. resistance (which is futile) is in the picture, otherwise its on the last page of my website for the djet conversion. If the sensor reads 0 then the car will assume full enrichment on startup and probably flood the engine
cheers
barri

[ctaylor738]

My slowly-developing theory is the car is running very lean for some reason. With the sensor unplugged, the brain is going full rich and is able to keep it going. Plug the sensor back in, and the brain goes to a more normal mixture, and the engine dies. Remember, there is no feedback mechanism in D-Jet to correct the mixture.

Ideas:

- low fuel pressure (should be 2 bars)
- vacuum leak
- MAP sensor
- throttle switch/vacuum reading (brain thinks throttle is closed and car is coasting, and cuts off fuel)

[rowdie]

Told you I wasn't very good at wiring. Had the right sensor, just misread the diagram. These old eyes ain't what they used to be.