1997 SL1 starts and dies only in the morning.
 

My 1997 Saturn SL1 with 165k miles on it usually runs like a champ. This last week it has been dieing on me right after I try to start it up. It will turn over and run for a few second and then sputter and die. If I do this a few times, I can eventually get it to stay on so I can drive to work. Funny thing is that this only happens in the morning. On my home for work it starts right up. This morning I got gas on the way to work, maybe 5 minutes from my house, and it started right up after I filled the tank.

Any ideas?

Thanks.

You might want to replace the ECTS. It's a known problem with the S series cars and can cause all kinds of problems. Replace the connector too. Be sure to source them from a GM dealership, though. For the two, about $50. Aftermarket sensors are NOT calibrated correctly and won't improve anything.

Longshot GUESS:

Battery voltage droop overnight.

Right to the ragged limit of being able to supply enough current to the starter to rotate the engine. After a few false starts, those rotations of the engine have allowed the alternator to charge the battery just enough to supply fully the starter motor with enough current to spin the engine as needed for it to start.

The alternator likely successfully charges the battery all the way to work.

The voltage droop occurs while you are at work.

So why does it start? Because the 8-9 hours you spend at work are less than the time the car sits overnight at home (12-13 hours). So the battery voltage droops less.

You can pseudo test this by using a multimeter to measure the V across the battery BEFORE you start the car at home in the morning and at work before you leave.

I am thinking that it is the ECTS as I have also noticed that my dashboard temperature dial is higher than it ought to be even in the mornings when the car has been cooling all night. I am confused by some of the guides that I am finding though. Is there one sensor for the comptuer and one for the dashboard? Or just one for both? If there are two, should I replace both sensors? Do they both use the same type of wire? I called the GM dealer and the guy there said that I should replace both and they both use the same wire, but I think I read somewhere that the dashboard sensor uses 1 wire and the computer sensor uses 2. I dont really have a lot of confidence in the GM dealer. If there is two, did they both break at the same time? I checked and my coolant level is fine, and the car runs fine once it gets going.

1995 was the last year with separate temp sensors for the PCM and the dash gauge.
For the record, the single wire in that configuration was for the dash gauge. and the 2 wire sensor is the actual ECTS.

So in 97 there is only the 2 wire ECTS which screws into the head and the PCM infers temperature based on the change in resistance of the sensor.

When the sensors fail, they usually go open circuit, tell the PCM the coolant is always -40 degrees, and therefore the PCM doesn't know how warm the engine actually is. If the car sits in traffic or idles for a long time, the temp of the system keeps rising, the aux fan never turns on, and the car overheats. If your fan is coming on somewhere near the 3/4 mark then your ECTS is not open circuit. When they fail shorted, the fan turns on and runs continually while the car is on.

But, per Rube's comment, if it's never been changed, it is likely the original flawed design and is providing inadequate information to the PCM.

Rube already covered the importance of using a GM sourced ECTS. It should be all brass.

I'm going to ask a stupid question here, but everyone should be used to that by now. Do you have the original spark plug wires on the car? If so, the resistance of them could be wonky. I have seen weak plug wires keep a car from starting all together. Usually it's more noticeable on older cars and cool mornings, but you just never know. And yes, once the cars started, they would be fine for the rest of the day. Next morning, no start again.

I did replace the plugs and wires 2 years ago I think. So yesterday, I got a new sensor from the dealer. Replaced it and it now starts up just fine. Thank you all for your help.

Now I am curious, what is the car changing with coolant temperature that makes it idle and then die when cold?

Air to fuel ratio.

At start up, the car runs in open loop mode where it ignores most sensor feedback until the sensors are warm enough to provide reliable and "accurate" readings.

One thing it does not ignore is coolant temp, which it uses to determine the appropriate air/fuel mix for start up. Sets it richer as the temp decreases.

When an ECTS goes bad, it usually goes open circuit or the equivalent and its resistance will be interpreted by the PCM as -40 degrees. So the PCM sets a crazy rich mix. Sometimes so rich that the car will not start.

When it goes into closed loop mode, the O2 sensor tells the PCM via the O2 in the exhaust that the mix is too rich. However, the PCM still thinks the engine is at -40.

I'm not sure what happens from then onwards, other than it usually causes numerous drivability issues which vary from car to car.
----------------------
Oops -- your temp was reading higher than expected. That would mean the starting air/fuel was too lean (aka not rich enough) for the car to start at the temperature inferred by the PCM. The equivalent of not pulling out the choke lever on classic cars. The mix would then presumably keep leaning out prematurely based on the bogus air/fuel mix..

And another longshot guess of mine goes down in flames.....

I'm so overconditioned to ECTS's going bad in the other direction (reads too cold) that it didn't hit me.

But Rube was all over it.
That's why a forum is more than 1 person.

Glad you got it back in running order!

Glad to hear it's back on the road. Don't run off on us. Stick around and learn some things about the S series Saturns. They're really great cars. Too bad GM had to get involved with them and kill off the Saturns all together. Frigging idiots!