starting issues

Post your E12 technical questions and comments here. Please, no off-topic posts.

Moderators: Mike W., Pierre

Post Reply
HELPNEEDED
Posts: 4
Joined: Wed May 06, 2020 1:43 pm

starting issues

Post by HELPNEEDED » Fri May 22, 2020 11:33 am

my 78 530 was my daily driver for many years.. when gas hit $4 , i put it aside for a couple of years. i have been trying to get it running for the last 4 years at least. recently i discovered, by accident , that if i left the coolant temp sensor unconnected, the car would start. it would idle very high (over 2500) and would die out if the coolant temp sensor was connected.
so that is the symptom. here is what i have tried.
I first found a short in the tach. since this also connects to terminal one of the ecm. this was the first problem found. the tach is not connected at this writing. I went through the diagnostics listed for the l jetronic. all resistances in the air flow meter, the coolant temp sensor, the idle sensor, are within normal limits. I also traced all wires in the connections under the hood. the dual relay seems to work fine as i have 32 pounds of pressure at the cold start relay. I also have a good spark. and i also have new injectors installed.
i have tried another ecu, which does not make any difference. and also a second air flow meter, which doesn[t make any difference.
i have tried to decrease the idle speed by adjusting the idle screw (not the one on the air flow) which helped a little. changing the timing also helped somewhat, but the idle was very rough.
there is new gas in the tank and i am using a pertronix ignition. wires and distributor cap seem good
where do i go from here:

HELPNEEDED
Posts: 4
Joined: Wed May 06, 2020 1:43 pm

Re: starting issues

Post by HELPNEEDED » Fri May 22, 2020 1:57 pm

additional info: no hose leaking is found. i did not visually inspect the aux air valve , but i did disconnect the hose at the manifold and blow through the aux air valve to make sure it was open.

User avatar
528i-1981
Posts: 300
Joined: Wed Feb 07, 2018 5:50 am
Location: Boston North Shore, MA

Re: starting issues

Post by 528i-1981 » Fri May 22, 2020 7:06 pm

You may have a big vacuum leak somewhere. Take off the intake hose between the air cleaner and the AFM and put an unlit propane torch in the AFM and see if it starts and runs better. If yes, go on the hunt with a can of starter fluid. I was having bedeviling problems recently and finally found a vacuum leak at an intake runner at the lower plenum (which I had removed while replacing injectors). I swear I tightened those bolts to spec, but that's where the leak was.

The brake booster hose is another common place for vacuum leaks, and the check valve might need replacing. If your car has a charcoal canister check that as well. Of course, the injectors are another suspect spot if they were recently replaced.

However, the temp sensor disconnected should be from the ECU's perspective close to a cold reading, which enrichens the mixture. If everything was in working order in that circuit, disconnecting shouldn't affect a cold start so drastically. Are you absolutely sure there's not a short in the harness and that the sensor is in spec?
(oo=00=oo) Eric
1981 528i Manual

HELPNEEDED
Posts: 4
Joined: Wed May 06, 2020 1:43 pm

Re: starting issues

Post by HELPNEEDED » Fri May 22, 2020 11:05 pm

thanks for the information. I will look for a leak tomorrow. when i said that i traced all of the wires in the engine bay, i meant that i checked all for shorts and grounds. none were found I also measured resistances of all of the listed divices. afm, coolant sensor, throttle sensor. and everything on the ljetronic troubleshooting faq . i have checked the hoses, but i havent checked with the propane. i hadn;t thought of that. thanks. the part that really has me baffled is why does connecting the wires to the coolant sensor shut the car down. like i said, i have two ecm;s , changing them does not matter. i suppose that the short from tach could have done something to the ecm, but to have i happen to both of the ecms is questionable.
would a vacuum leak cause an initial slow sluggish idle and then go to a very fast smooth idle?

User avatar
528i-1981
Posts: 300
Joined: Wed Feb 07, 2018 5:50 am
Location: Boston North Shore, MA

Re: starting issues

Post by 528i-1981 » Sat May 23, 2020 7:15 am

Well, it depends on what else has happened with the throttle body screw and the mixture. As I was playing with mine trying to get it right, I was pulling levers all over the place and the engine behavior wasn't making any sense. It could be very lean from false air and disconnecting the temp sensor enrichens the mixture so that it's running better. But I don't think that shouldn't matter on a cold engine, so I'm a bit baffled too, honestly.

I would pick a measurement from the temp sensor pin on the ECU connector to make absolutely sure the correct signal is making it there. I'm not sure how you measured for shorts in the harness - did you verify continuity and ensure there's no cross-connections?

Did you check for cracks in the bellows between the AFM and the throttle body?
(oo=00=oo) Eric
1981 528i Manual

HELPNEEDED
Posts: 4
Joined: Wed May 06, 2020 1:43 pm

Re: starting issues

Post by HELPNEEDED » Tue May 26, 2020 9:15 pm

i finally got around to checking for vacuum leaks and you nailed it, there are leaks at the injectors. so that will the next issue addressed. i should be good at working on this as i have taken it apart so many times for various reasons, but i still hate doing it.but i already paid an "expert" 1000 bucks to fix things and didn't get a penny's worth of value.
as for the wiring harness, i literally seperated all of the wires, checked for continuity and verified that none were crossed, none of the insulation was damaged. then i repaced all of the 2 pin bosch connectors , soldering the new ones on and keeping polarity correct just to make sure i wasn't causing a problem. I measured all of the connections at the harness plug for the ecm to verify the resisstance values were within tolerance of the values in the ljetronic troubleshooting guide. I also checked the values of the air flow meter , the throttle position switch and the coolant temp switch.
just to make sure , i got another coolant temp switch, but that one measured resistance below the listed value in the troubleshooting guide. I also placed both the new and old ct switch in the fridge to ensjure they reacted to teemperature and both did chANge so i kept the old one and returned the other.
will keep you p0sted when i "solve" the vacuum leak im glad you mentioned to check around the injectors, because i probably wouldn't have thought to look there since new ones were just replaced

User avatar
528i-1981
Posts: 300
Joined: Wed Feb 07, 2018 5:50 am
Location: Boston North Shore, MA

Re: starting issues

Post by 528i-1981 » Wed May 27, 2020 6:43 am

L-Jet really hates vacuum leaks. The stout and reliable M30 goes from daily-driver mode to total basket case, and without ODB it's all trial-and-error.

I'm really impressed by your diligence on the harness. I thought that could have been the problem, too. But as Lock once wrote, there are (usually) no coincidences. You replaced the injectors, and that's where the issue turned out to be. Same for me with the intake runner leak. I was absolutely sure that couldn't be the problem, even though I had just replaced them with E21 runners and replaced the gaskets. Too bad you have to tear it all down again, that really is a pain (I got off easy, the runner leak was number 4), the bolts were easily accessed).

There was a recent Hack Mechanic where a faulty distributor was diagnosed, and an earlier one regarding a hot temp reading on a road trip in a 2002. Both involved a faulty assumption in finding the source of a problem, and then isolating that assumption. It's a great lesson when working on a car without OBD.

Cheers, and let us know how you make out,
Eric
(oo=00=oo) Eric
1981 528i Manual

Post Reply