Churchill

Over the last few days my car ('96 C2, 150K miles, very well maintained) has developed a serious poor running condition during warmup only. It starts fine, normal number of cranks. But during the first 3-5 minutes of operation, when the engine is cold and warming up, it runs terribly. It idles fine, but you can barely accelerate, have to feather the throttle, and it stumbles and bogs down, makes backfire pops (not loud), and feels like it's running on three cylinders when you give it gas. Within a few minutes, though, it's running normally again and when fully warmed up it's all better. It has not triggered a CEL yet (this just started happening).

So this must be something related to temperature. Cylinder head temp sensor seems the likely culprit, but in all my searching online I didn't find a description of identical symptoms. I should be getting a CEL but as I wait, wondering if anyone has additional thoughts. The car has recent plugs, fuel filter and pump, caps and rotors, etc, and otherwise runs great.

GC96

Do you have a scanner that will let you monitor sensor outputs? If you do, then you might want to see what the output value of the temperature sensor is during those initial few minutes of operation where you are having the issue. It may be providing an incorrect temperature value that is still within the plausible range. This may not automatically trigger the CEL.

GC

Gbos1

Defective spark plug / Defective wire / Cap / Rotors defective or small crack … If no CEL have your mechanic recheck what was just installed.

pp000830

Hi Churchill,
Before doing a diagnosis clean the idle valve.
Are you getting a misfire-related Check Engine Light?
Here is a page that discusses poor running:
https://993servicerepair.blogspot.co...r-robbing.html
Andy

Churchill

Mystery deepens. Connected my Durametric and fired up the car and the cylinder head temp sensor seems to be working fine. Warmed up with the engine. This was only idling in my garage, but if the sensor was bad it'd be obvious. Was pretty sure that was my problem. O2 sensors seemed to be behaving fine as well. I'm stumped. Will continue investigating.

Gbos1

German Gremlins

GC96

Does the problem only appear when the engine is under load? If so you might want to drive the car and log the data while you do it, specifically probably the readings from the O2 sensors and the corresponding fuel trims. From a cold start I think it can take a few minutes for the ECU to switch from open loop to closed loop mode where it actually starts to use the O2 sensor data to actively manage the air-fuel ratios. It's possible that you might have a small vacuum leak when the engine is under load for which the ECU is able to compensate once it is in closed loop mode but causes driveability issues in open loop mode.

GC

Churchill

Thank you GC -- great advice. Will do. Been crazy busy lately and haven't had time to dig into the car much. I had similar thoughts -- feels like an open loop problem. Once the engine temp reaches X degrees it basically goes away. Fully warmed up it runs great.

pp000830

Hi GC,
My understanding is that it only takes a few seconds from a cold start for the oxygen sensors to begin working in closed loop mode as the sensors use a heater wire to facilitate the oxygen sensors heating up into their operating range.
I suspect this does not preclude an older dirty sensor from taking some extra time to warm up and get into its working range and begin to give the ECU meaningful information.
I seem to recall if the Oxygen Sensor takes too long to begin giving a correct signal it throws a code and the Check Engine Light will illuminate.
I use Durametric and have found interpreting running data to be a little hit-or-miss for me. Because of this, I suggest that if folks are seeing a code related to the oxygen sensors or something that uses the oxygen sensors to evaluate and trigger a code one should just replace the sensors on the side of the engine the code references if the sensors are old and original to the car.
Andy

GC96

Hi Andy,

You're right that the Oxygen sensors do have heaters to bring them up to temperature quickly so that they can provide accurate signals. But it's my understanding that in addition to that, the engine itself has to be at a sufficient temperature as signaled by the temperature sender before the ECU switches to the closed loop mode. Essentially on a cold start the ECU runs a slightly rich mixture to get the engine started and then up to temperature before beginning to control the mixture for efficiency. I believe that is similar to the operation of the cold start valve on the older CIS injection systems used on many older German vehicles. On one of my old 90's VW's that also used a Bosch Motronic 5.x engine management, the procedure for monitoring the Oxygen sensor output required that the engine coolant temperature be above a certain threshold first. Similarly the 993 diagnostic manual for monitoring the Oxygen sensors calls for the engine temperature to be greater than 90 Deg. C and also specifies a time delay of 250 to 350 seconds following the engine startup. I presume that the time delay is to ensure the engine is up to temperature and also ensure that other events that could potentially cause confusing readings like any SAI activation are complete before you start using the sensor readings for any diagnostic purposes. I will concede that I'm not an expert and so may be mistaken in my understanding of how the system works.

GC

95_993

Churchill,

I am suspecting fuel regulator. Bogging down, etc makes me believe it's dumping fuel inconsistently.

IainM

But only when cold???

Im with GC. Small vacuum leak post MAF.
Also, the CAT needs to be up to temp for valid O2 sensor readings.

Churchill

Thanks guys for all the good suggestions. Just had a chance to smoke test the intake. Airtight. Absolutely nothing doing there. Doesn't mean of course that somehow with the engine running there isn't a vacuum leak but at this point I'm going to look elsewhere. Going to take it out and log data with the Durametric.

Fuel pressure regulator is interesting -- it does feel like during warmup it isn't getting enough fuel. Idles fine but drives terribly, and the more throttle you try to give it, the more it bogs. Never had a problem with an FPR before -- do they fail in this way?

95_993

A leaky FPR will cause fuel to get past the internal diaphragm and into the vacuum side, resulting in extra fuel on the intake side. As the engine warms up, it's less pronounced. That's my theory.

Churchill

Well, after having several days in a row of this odd behavior, my car appears to have fixed itself (for now). Hasn't done it the last few drives at all. Absolutely no idea what happened -- maybe water in the gas that settled and worked its way to the engine during the first few mins of operation?