No Start...Pull CTS plug and it Starts!

[dcvee]

This is a wild one...for me anyway. Parked this car for 3 weeks. Was running perfect. Now, it won't start unless I pull the CTS connector and then it runs terrible. It appears the injectors are not firing properly per DME signal. I think pulling the CTS just richens up the injectors to the point where there is so much fuel in the cylinders they are firing out of order. Very high idle with the CTS pulled and hard to keep the car running. Fuel pump is fine, perfect pressure. 12v+ at injectors. Swapped DME, same issue. Spark is perfect. Timing on the money. Coil/plugs/wires are all ok. Tested both reference sensors, both ok. Swapped, same issue. Swapped out CT, ICV, Relays, O2 sensor, AFM, etc etc from my other 2 cars. I am now to the point of tracing wires from the DME since it HAS to be something on the negative side of the signal wires coming from the DME to the injectors. I mean I just can't see it being anything else? I know the injectors batch fire by the DME transistors grounding them. I'm just thinking something is interfering with that?

Is my thinking right or am I missing something obvious?

Thanks,
Don

[Chris W]

Hi Don,

I had that same problem with my car. The car would not start when cold unless I pulled the CTS plug. Then it would start and after about 30 seconds I could plug the CTS plug back in and it would stay running. With the CTS plug pulled, the engine seemed to be running rich and it idled a little higher than normal, but would return to around 880 rpm when the CTS was plugged back in. When the engine is warm, it would start fine even with the CTS plugged in. I also thought there was a problem with the injectors. I thought they might be leaking and had them cleaned, but the problem persisted.

I eventually traced my problem to 2 factors. One was the barometric pressure sensor, which was reading my altitude about 800' lower than my true altitude. When I adjusted the sensor to the correct range, the car would start, but run rough until the engine warmed up. It would also buck and stumble a bit when accelerating from very slow speed. The other factor that caused my problem was a faulty air temperature sensor in the AFM. It was a remanufactured unit about 15 years old and I thought it was fine. Then I reinstalled the 30 year old original unit, and now the car runs perfectly. It now has better throttle response than I ever remember it.

My best guess is that with the faulty air pressure and air temperature sensor the engine was running so rich that it wouldn't start. Then by pulling the CTS, it would lean the engine out just enough to be able to start, but run horribly rich.

[dcvee]

I'll check both tonight and report back. It's nice having spare cars!! Thanks Chris!

Don

[HANDBLT]

unplug the bar sensor and throw it in the trash. Its known for killing engines with bad readings.

T

[GG///M3]

Can wait to hear the full fix.

[dcvee]

Worked late...didn't have a chance to swap parts. Tony....a win-win. Weight savings to boot.

Thanks,
Don

[Chris W]

If you have the same problem I did, throwing away the barometric pressure sensor will only make the problem worse. In my case, the sensor was already reading the air pressure as denser than it actually was leading to over-fueling and being too rich to start. If you disconnect the sensor, the ECU will default to the reading for the car being at sea level, which in my case would have been denser still and even worse over-fueling.

Swapping parts is a good way to check things, but it might not be completely foolproof. I suspected my barometric pressure sensor earlier, and bought several replacements. I now have 3 of those sensors and they all read fairly consistently with each other, yet none of them are reading my altitude correctly. Eventually, I had to open one up and manually adjust it until it read correctly.

[RAD2LTR]

If you have the same problem I did, throwing away the barometric pressure sensor will only make the problem worse. In my case, the sensor was already reading the air pressure as denser than it actually was leading to over-fueling and being too rich to start. If you disconnect the sensor, the ECU will default to the reading for the car being at sea level, which in my case would have been denser still and even worse over-fueling.

Swapping parts is a good way to check things, but it might not be completely foolproof. I suspected my barometric pressure sensor earlier, and bought several replacements. I now have 3 of those sensors and they all read fairly consistently with each other, yet none of them are reading my altitude correctly. Eventually, I had to open one up and manually adjust it until it read correctly.

Just out of curiosity, how do you figure out what the BPS is reading? How does one adjust it if its not reading correctly? I'd love to check mine and adjust it if it needs it.

Will

[Ant's m3]

Just out of curiosity, how do you figure out what the BPS is reading? How does one adjust it if its not reading correctly? I'd love to check mine and adjust it if it needs it.

Will

Same here!

I also have a shitty cold start and stumbles and bucks if i try to move it when its cold. I just thought Im in need of a tune up. Only happens in the winter months tho

[HANDBLT]

The S14 is notorious for fouling plugs on cold starts so its already on the edge of stumbling and running rough as normal. I don't have ICV or BPS but I have AN and I have the TB's set to idle at 900 on cold start until the warm up timer decays. You can hear and feel the engine change once this timer decays. I think its 90sec?

put a volt meter on the AFM signal wire and make sure it is getting the 0-5v to the ECU. also check the TPS at the ECU for idle and off idle. kind of hard to check RPM getting to the ECU without a scope to see the square wave? coming off the bell housing sensors.


T

[Chris W]

Post #4. Thank Jake.

If it's reading incorrectly, you'll have to pry the cover off the sensor and bend the contact arm so it contacts the variable resistor at a longer or shorter point.

[dcvee]

Well I changed AFM and Baro with no change. Removed Baro with no change. Pulled fuel rail with injectors. No fuel on start up. Zip. Not a drop. Pull CTS plug and I have fuel. Check Fuel pump pressure(brand new in-tank conversion) 44psi. Jumped fuel pump, same issue.

I did notice one strange thing about the injector circuit...I know the DME fires the injectors in batch. 1-2 are on the same negative output and 3-4 are on the other coming from the DME. Hot's are all 12vdc. With the DME disconnected, I have continuity across both of those grounds? I would think they would be separate so the DME can fire either circuit. I'm going to go over the schematic...granted I have not. But pulling the fuel rail and observing the injectors during start up confirms my findings in post 1.

Post back soon...thanks for any suggestions.

Don

[dcvee]

Ok...looking over the schematics and focusing on the negative side of the injector harness. DME grounds either 1&2 or 3&4 injectors. I see NO reason or other electronics that would allow continuity between these two points once the injectors are out of the circuit. So, if I remove all injectors from their electrical connectors, I should see no continuity between outputs 14&15. Am I right? I have also triple-checked both the engine speed and cylinder identification sensors. Both are 940-980 ohms. I have swapped them to no avail.

Don

[Chris W]

Don,

It really sounds like you have the exact same problem I did. When I was diagnosing my problem, I also had occasions when nothing was coming from the injectors. I took them off and found that no fuel was being pumped by the fuel pumps. In my case, what was causing the fuel pumps to not run was an intermittently failing fuel pump relay. At first, I wouldn't believe that the fuel pump relay could be faulty because I hooked a 9V battery to it and I could hear it click every time. Yet, as soon as I replaced it, I could get the car started, albeit very rough and only after unplugging the CTS.

These 2 unrelated problems, each causing a similar result had me chasing my tail for a few years. Having the fuel pump relay work one time and not another made it a bear to diagnose my problem and I changed every part in the fuel injection system, before I started making any progress.

To diagnose your problem, the first thing to do is to remove a fuel injector and make sure fuel runs consistently through the fuel rail when starting. Then check the fuel injectors are firing, and then that the spark plugs have a strong spark. If those 3 things are present, then re-check the barometric pressure sensor reading by comparing it with Jake's figures and adjusting it as necessary. Seriously, the CTS being plugged or unplugged has nothing to do with whether the fuel pumps run or the fuel injectors fire.

[dcvee]

Right with you Chris...rechecking fuel pump operation/flow through the system tonight. This system is simple since I went with a single walbro in-tank pump on this one. I don't know about the CTS signal to the DME and what it does...but I can tell you this. If I pull the injector rail and place shop towels under it and try to start the car, no fuel from the injectors. If I pull the CTS plug I get fuel from the injectors. Now, I don't think I get the correct amounts OR timed correctly. But I can drive the car in and out of the garages although it runs terrible and has a very high idle. Plug the CTS in, car dies and wont' start. That sounds pretty consistent to me.....LOL!! I've never had a problem like this on any BMW I've ever owned.

If the pump checks and fuel flow seems to be good I'm going to go after the injector grounds/outputs from the DME. I guess the only input that effects the ground coming from the DME to the injectors would be the flywheel cylinder identification sensor. I can't see the speed sensor having a role in that part of map? Dunno.... I checked both....they do vary..one is 940 ohms and the other like 980 IIRC.

Thanks for the help Chris...I sincerely appreciate it.

Don