2008 Saturn Aura No Start

[Fishinguy40]

4 cylinder. My daughter complained that her temp gauge would sometimes oscillate hot to cold. This started about a week ago. Then last night the car would not start. I drove out to where she was and of course the car started on first try. Today a friend of hers replaced the engine coolant temp sensor. Temp gauge still oscillates. She gets home with the car and then 10-15 minutes later the car won't start. I go out and look at it. She cranks but won't fire. 60 psig fuel pressure. I look at the temp gause and it is at minimum cold temp. Car was only parked for 10 minutes and engine was still warm. I disconnect the temp sensor and try to read voltage at the pins but read 0. Pretty sure that I got good connection with the Fluke but its night and cold in the driveway. I'll check again tomorrow.

Would the ECT keep it from starting? I will also get scan reader from the parts store tomorrow to check codes. Engine light is on.

Any ideas?

I just hope the problem isn't with the ECM.

[Fishinguy40]

Here's an update. Later last night after writing this post the car started with no problem. Temp gauge also came up slightly and steady as the engine was still a little warm. So the problem is definetly intermittent. Tonight after work I plan to carefully check out the wire harness and connections between the temp sensor and ECM.

I'm still hoping someone on here pipes up and says "Yeah, Iv'e seen that before. Here's the fix"

[Jeremy airjer W]

How many miles?

[Fishinguy40]

Approx 61k

[LightningBG]

I really have no idea, but.... if it was mine. I'd get it scanned and maybe disconnect the battery for a while to reset the computer?

Truthfully, I have no idea, I just start taking shots in the dark, then start replacing parts.

FYI, an ECM on Rock Auto only runs $186

ACDELCO Part # 216140 Click for more information about this part {#12597125, 12612397}

Engine Control Module (W/ 2ND MPU)

$185.89

[Jeremy airjer W]

was the check engine light on before the no start problem?

If so it may be worth seeing what code(s) are present before proceeding further.

[Fishinguy40]

According to my daughter, the check engine light did come on a day or two after the temp gauge started swinging last week. So it was on before the no start condition but she stated that it also cleared for a day or two then came back in again when the gauge would swing again.

When her friend replaced the ECT, the battery was disconnected and the computer allowed to discharge/reset. Engine light went out.

I do plan to get the codes tonight.

[quickfish6]

Take a look and pull out the temp sensor and test that. If it is shorted out the car will not start. Looks like the ECM going bad is a common issue on these cars. You almost need a TECH 2 to try to turn things on and off and see what the ECM is controlling and not.

[Fishinguy40]

Just checked the codes, a total of 4. P0117 Low Engine Coolant Temp and P0118 High Engine Coolant Temp. Each one came in twice. No other codes present.

Guess I'll be looking at the wiring and connections between the ECM and engine coolant sensor. After that I'm out of ideas

[JOHNBIGDOG]

well considering you didn't see a circuit code I would assume that the wires to and from/pcm and ect sensor are ok.

I would let the car warm up and feel the upper radiator hose. see if you feel the same hot/cold swing that your ect sensor shows. To me, may be a t-stat or a head gasket. See if the cooling fans kick in at some point, or if they will turn on when the a/c is turned on.

I'm just throwing out some things to check...

(ECT= Engine Coolant Temp)

[Fishinguy40]

I guess I should add that when my daughter's friend changed the ECT sensor he drained/replaced the coolant and changed the T-stat. I've also heard the cooling fans kick in.

[Jeremy airjer W]

Eat low should be a problem in the winter. Default is -40 with an open sensor. It would be a little rich but it should start unless the engine was up to temp the it would flood. Temp to high would indicate a short in the sensor, this would make the temp read extremely high, not a good thing when your trying to start a cold engine.

I would like to see what the temp sensor value was when it doesn't start with a scan tool. It's possible that the connector may be the issue and not making good contact.

[LightningBG]

Fishinguy40,

Sounds like you may have a trip to Woodbury coming up....

Jeremy Airjer W

ASE Certified Lead Night Tech - Woodbury Autocare

[JOHNBIGDOG]

well considering you didn't see a circuit code I would assume that the wires to and from/pcm and ect sensor are ok.

(ECT= Engine Coolant Temp)

Darn it, should have googled your two p codes, there I go trusting people again. They are circut codes, high and low as Jer stated.

[Fishinguy40]

My bad. They are high and low voltage, not temp.

So here is what I did today:

1. Started the car and let it warm up to normal temp on the gauge. No gauge oscillation. Pulled it into the garage and shut it down.

2. Pulled and inspected the two ground wires on the front frame. Battery ground was good and clean. The other ground that went into a wire harness had a little corrosion underneath. Wish I had measured the resistance before removing. Cleaned it up nice and shiny and re-installed.

3 Removed the ECM and disconnected the wire connectors. No corrosion or bent pins. Everything looked nice and clean on both the ECM and wire harness side.

4. At the ECM wiring harness I inserted small pieces of solid wire into #4 and #58. (This is where I spent most of my time. I had to take coax cable, strip it and use the solid wire in the center. I had to carefully grind it down so that it would fit into the female plug. Had to do the same with a couple small finish nails for the other end of the wire harness at the ECT sensor.) This allowed me to read through the wire harness and ECT sensor. I read 3.2 Kohms which should equal 68-77 DegF, according to resistance tables. I checked again 30 minutes later and got 4.4 kohms which should be about 59F. So I'm confident that the ECT sensor and connector are good. I twisted/tugged and flexed the cable/connector at the ECT sensor and my ohm readings never faultered.

5. I disconnected the ECT sensor. Installed two small finish nails in the end of the connector. I installed a small wire jumper into #4 and #58 on the ECM connector. From the ECT sensor connector I was able to read 0.4 ohms through the cable. At the ECM end, I twisted, tugged and flexed the cable. My 0.4 ohms never changed. So I am confident that the cable and connectors at both ends are good.

6. I re-installed the ECM and connectors. From the ECT sensor end I read 0.5 ohms to frame ground at pin A on the cable connector. Again I twisted and flexed the cable connection at the ECM module. No change in resistance.

7. I turned the key tro ON. From the ECT sensor end I read a solid 4.96 volts to frame ground at pin B on the cable connector. Again I twisted and flexed the cable connection at the ECM module. No change in voltage.

I'm hoping the cleaning the corrosion at the ground connection or reseating all the connectors fix the problem. If not, I'm sure the ECM will have to be replaced.

A couple questions for AIRJER:

1. Is there a reasonably priced kit for inserting/testing these male and female connectors? I'd hate to damage one some day inserting something too large.

2. How much would it cost to replace the ECM? Can I do it or is there programming involved?

[Jeremy airjer W]

I can count the number of computers I have replaced on one hand. It is likely not the issue!

Most of the time I use a paper clip. You don't need to insert anything into the pin. You just need Make contact with it. Most of the time I back probe with either a back probe pin or a JR Jumper. The pin is inserted into the wire end of the on metro. The jr jumper has a "bed of nails" that will pierce the wire insulation without any major damage to the integrity of the wire.

[Jeremy airjer W]

On a side note, I had two Chevys in the same night with dead clusters. One with the speedo stuck at 70 and both the oil pressure and charge gauge dead, and the other with an erratic temp gauge along with a speedo stuck at 50.

It is important to look at the data with a scan tool to make sure that the computer is receiving or at least communicating the appropriate values. If you took all those measurements and looked at scan tool data and it said the coolant temp was at -27 than I would say your on to something with a bad ECM.

[LightningBG]

On a side note, I had two Chevys in the same night with dead clusters. One with the speedo stuck at 70 and both the oil pressure and charge gauge dead, and the other with an erratic temp gauge along with a speedo stuck at 50.

Bad clusters are or were pretty common on GM trucks for a while. My envoy had to get a new one, i think they even extended the warranty for hat particular issue. Very erratic guages and one by one would die.

[Fishinguy40]

My bad. They are high and low voltage, not temp.

So here is what I did today:

1. Started the car and let it warm up to normal temp on the gauge. No gauge oscillation. Pulled it into the garage and shut it down.

2. Pulled and inspected the two ground wires on the front frame. Battery ground was good and clean. The other ground that went into a wire harness had a little corrosion underneath. Wish I had measured the resistance before removing. Cleaned it up nice and shiny and re-installed.

3 Removed the ECM and disconnected the wire connectors. No corrosion or bent pins. Everything looked nice and clean on both the ECM and wire harness side.

4. At the ECM wiring harness I inserted small pieces of solid wire into #4 and #58. (This is where I spent most of my time. I had to take coax cable, strip it and use the solid wire in the center. I had to carefully grind it down so that it would fit into the female plug. Had to do the same with a couple small finish nails for the other end of the wire harness at the ECT sensor.) This allowed me to read through the wire harness and ECT sensor. I read 3.2 Kohms which should equal 68-77 DegF, according to resistance tables. I checked again 30 minutes later and got 4.4 kohms which should be about 59F. So I'm confident that the ECT sensor and connector are good. I twisted/tugged and flexed the cable/connector at the ECT sensor and my ohm readings never faultered.

5. I disconnected the ECT sensor. Installed two small finish nails in the end of the connector. I installed a small wire jumper into #4 and #58 on the ECM connector. From the ECT sensor connector I was able to read 0.4 ohms through the cable. At the ECM end, I twisted, tugged and flexed the cable. My 0.4 ohms never changed. So I am confident that the cable and connectors at both ends are good.

6. I re-installed the ECM and connectors. From the ECT sensor end I read 0.5 ohms to frame ground at pin A on the cable connector. Again I twisted and flexed the cable connection at the ECM module. No change in resistance.

7. I turned the key tro ON. From the ECT sensor end I read a solid 4.96 volts to frame ground at pin B on the cable connector. Again I twisted and flexed the cable connection at the ECM module. No change in voltage.

I'm hoping the cleaning the corrosion at the ground connection or reseating all the connectors fix the problem. If not, I'm sure the ECM will have to be replaced.

A couple questions for AIRJER:

1. Is there a reasonably priced kit for inserting/testing these male and female connectors? I'd hate to damage one some day inserting something too large.

2. How much would it cost to replace the ECM? Can I do it or is there programming involved?

Well it's been a week since I did everything above. No issues. Hopefully it was a connection issue and is fixed. I am keeping my fingers crossed and will report if anything changes.