CIS idle question... open loop behavior

[fireant911]

Folks,
I have a question regarding CIS Lambda behvior during idle on my 81SC. It is my understanding that during normal driving the Lambda brain is controlling the mixture shooting for an average Air/Fuel ratio of 14.7. Upon removing one's foot from the accelerator results in the decceleration valve slowing down the rpm drop for emission reasons. My question pertains to the behavior of the 'fuel control' during the time the engine is at idle (engine at normal operating temperature)... does it immediately go from closed loop to open loop or is there any sort of time delay? I can hear and see (via an A/F gauge) that my car goes from Lambda control to open loop if I am stopped at a traffic light - but I am stopped there for approximately 20 seconds before it goes to open loop. Is this normal (the microswitch appears to be functioning properly - it is just the time delay that I am curious about)?

[psalt]

does it immediately go from closed loop to open loop or is there any sort of time delay?

but I am stopped there for approximately 20 seconds before it goes to open loop


Daryl,

An 81SC is in closed loop at hot idle, cruise and under 35% throttle. Your questions states that your car goes to open loop after 20 second at idle, is that correct ?

As soon as you open the throttle more than 35%, the TPS grounds the same circuit as the 15C degree switch and the system goes open loop. There is no delay built in to the circuit for the WOT switch from closed loop to open loop.

If you want to see what the system is doing, the best method is to view the duty cycle directly with a dwell meter. What kind of A/F gauge are you using ? A narrow band sensor is only accurate around a few tenths of stoich because of the slope of the voltage output. It was designed to function as a switch, not a linear sensor.

[fireant911]

Paul,
The Bentley manual has the following passage, "The throttle switch contains two integral microswitches. One switch provides a closed throttle input to the oxygen sensor control unit and one switch providfes full throttle position (closed at 35° throttle angle). When the idle contact switch is closed, the control unit delays the oxygen feedback cycle to prevent erratic idle. When the full throttle contact is closed, the oxygen sensor control unit switches to an open-loop, fixed (richer) control mode and the A/C compressor is turned off.

I also have both Watson's "How To Tune & Modify Bosch Fuel Injection" and Probst's "How to Understand, Service and Modify Bosch Fuel Injection & Engine Management" books in my library. I think I remember reading somewhere that adjusting the fuel mixture via the 3mm screw has the greatest effect at idle since Lambda takes over when the engine is operating at cruising speed as long as the throttle is less than the 35° point.

I scanned these books this morning to refresh my memory and you are right on target with the statement that the engine is in closed loop at hot idle. The behavior in the bold section above is what I am curious about - what is meant by "the control unit delays the oxygen feedback cycle to prevent erratic idle"?

[psalt]

Darly,

I think they are referring to the frequency of the feedback cycle at idle, not a delay in switching from open to closed loop. If you watch the needle of a analog dwell meter hooked up the the lambda test port at idle and compare it to cruise, the rhythm of the needle is different. The system trims the mixture at a higher rate at cruise than at idle, which gives a rhythmic beat to the idle. Most programmable EFI ECU's I work with let you change the target AFR for the idle and cruise, and the frequency that the system dithers the mixture, for the same reason. No fuel management system is accurate enough to hold a specific AFR with precision, they have to dither the mixture around a target to make an average AFR. If you want closed loop control at idle, this is how it has to work, and a CIS or EFI engine without an active idle control device will have a fluctuating idle by design.

The Bentley manual has a few errors on CIS lambda and the main sensor testing contradicts Bosch and Probst's books. The 3 mm screw has a much more profound effect at idle than at WOT under load. Few OEM's allow significant user adjustment to the high load WOT mixture for a good reason. The main fuel curve is determined by the FD and the shape of the air funnel. The lambda feedback cycle trims the mixture to keep the cat in a safe range, but has no effect under heavy load. However, the FV must be pulsing in the default range for the fuel curve to be correct. It is curious that no one suggests that removing the O2 sensor in a new Porsche would improve performance, but it does not, for the same reason it does not on a CIS lambda system.

[ninesixfour]

How does one know if the car is properly going into open loop mode? I have a 79 USA motor running in my 82 Euro car and the AFR is fine at idle but goes lean (to 17) at WOT.

[Bob Kontak]

Originally Posted by fireant911

but I am stopped there for approximately 20 seconds before it goes to open loop. Is this normal (the microswitch appears to be functioning properly - it is just the time delay that I am curious about)?

Daryl - Are you meaning to say that after 20 seconds at a stop light your system finally goes into closed loop?

I have an 81 US 911 and will check the time delay in the garage with my dwell meter for grins if you are interested.

Paul - good info.