What's the purpose of an O2 sensor ?

[Trakrat]

I completely understand how an O2 sensor is used on my modern twin turbo car with an advanced ECU, and understand how the map changes between air and fuel mixture, along with other sensors, and changing the map from an open to closed loop.

But my 911 isn't that modern.... and I'm not really sure what is going on when the O2 sensor is connected and ALSO when it is disconnected.

Are all O2 sensors (excluding wideband) the same?

[darrin]

as I understand it, the 02 sensor provides our Carreras' computers with information that enables the computer to adjust the air/fuel ratio in real time -- i.e. the 02 sensor allows the computer to operate in a "closed loop" environment. See, e.g. What the Home Mechanic Needs to Know about O2 Sensors

Not much different than your modern twin turbo car -- just a less powerful computer at work

It's also my understanding that when our carreras' wide open throttle switch is activated, the 02 sensor is bypassed and a WOT mixture is applied. So, similarly, when the 02 sensor's disconnected, our carreras' ecu can no longer adjust mixture on the fly and default to a default mixture map that conservatively richens out the mixture to avoid detonation (at the expense of fuel mileage and emissions).

[spoke]

The O2 sensor on my '86 930 has never been connected for the 10 years I've had the car. It has CIS and Permatune ignition. Much of the emissions equipment has been removed like the air pumps to the exhaust.

I've tried a couple of times to connect the O2 sensor. The engine starts, runs fine until I give significant throttle. Then it misses and doesn't run well. Unplug the O2 sensor and it goes like hell.

The CIS doesn't seem to have any adjustments that would be used by the O2 sensor.

[Rawknees'Turbo]

^^^

On a 930 CIS system, the O2 sensor is part of the low and cruise RPM, AFR control system; the frequency valve responds to signals from the Lambda computer, which are based on the AFR reading from the OS sensor, and raises control pressure in order to lean the mixture (this is done in an attempt to create less exhaust emissions).

At about half throttle (based on the throttle position switch), the frequency valve goes into a fixed duty cycle that no longer affects control pressure. The same thing takes place when you unplug the O2 sensor.

[Walt Fricke]

Spoke - I suspect a previous owner had the CIS adjusted to run without the sensor input. As mentioned, you probably got somewhat worse fuel mileage, and put out more emissions.

Disconnecting a sensor works as Rawknees says. The sensor creates a voltage which is based on the amount of oxygen in the exhaust, I think, as a proxy for uncombusted gasoline. I think when the sensor starts failing, it puts out lower and lower voltages for the same air/fuel ratio. So I am at a bit of a loss as to how being, say, shorted, which means no voltage out, is going to cause the mini-brain which controls the frequency valve to make the car not run well above idle.

Jim Williams (google the name and CIS) on his website says that over time the voltage put out by the O2 sensor falls from what it should be. Leaded gasoline shortens the life of these sensors greatly, too. But he says this eventually leads to the frequency valve/mini brain system running at a 50% duty cycle, which is the default position anyway other than for a cold start. Which is why hooking it up leading to as noticeable a change as you report is a bit mystifying - reveals a gap (of which there are quiet a few) in my understanding of the system.

I also don't know what control differences there are on a turbo CIS compared with the NA systems I am familiar with. Maybe none - blow more air past the control plate, get more fuel, so on or off boost, and how much boost, maybe doesn't matter for the AFR?

But aside from the intellectual challenge, you seem to be in a good spot if you don't care about mileage and emissions - the car runs fine.

[Rawknees'Turbo]

Walt, the Turbos' WURs have a provision for boost enrichment - an increase in boost pressure to the WUR lowers control pressure and enriches the mixture. If the WUR is made adjustable, you can then tune the amount of enrichment (how much lower the control pressure goes) to get the full throttle/boost AFRs you want (within the limitations of the fuel head, of course).

[pmax]

Also called a lambda sensor, which forms the closed loop feedback control system for adjusting AFR based on the O2 levels. Closed loop is a good thing. Idle control on the 3.2 Motronic is another example.

[911pcars]

Just to fill in some blanks:
https://en.wikipedia.org/wiki/Oxygen_sensor

[3rd_gear_Ted]

An onion has many layers.
Does an S.Wong hi-octane chip interact with O2 sensors??
Do exhaust flow characteristics change anything like with headers??

[Walt Fricke]

To me, the interesting question is why is his engine doing what it is doing. I've not heard that connecting an O2 sensor would cause an engine to run worse, given that the sensor is a voltage source, and a disconnected sensor means zero volts, which, say, a sensor which is short circuited to ground should look like zero volts as well.

Do the turbos get enrichment by connecting the intake vacuum (or boost?) from some convenient place to the lower chamber of the WUR? I know on my NA WUR, if I sucked or blew into a hose I hooked to the attachment on the bottom chamber of my WUR, I could watch the change in control pressure. Can't recall which way it went, but pretty strong movement. Makes sense, as the diaphragm in there is like a barometer bellows, so to speak. Interestingly, it affects the inner of the two springs which push up against the seal plate where the pressure changes are governed. On mine, the lower chamber is just open to the ambient air, though some models have a line to this for control purposes - maybe altitude adjustment?

So why does his car run poorly with the sensor attached?

[spoke]

Here's a little bit of the fuel pump arrangement on the '86 930. When I was having issues with the Airflow Sensor Charging Relay, one owner sent me these fuel pump relay bypass instructions. It was recommended for me to get rid of the Speed relay. The suggested changes are in blue.

The only modification I made was to remove the Speed relay and jumper pins 1 and 2 as shown as mod #3 in blue.

[spoke]

Here's the schematic for the engine management system for the '86 930.

The O2 sensor is on the left and connects to the O2 Sensor control unit. The Speed relay that I've removed is on the top left.

[Rawknees'Turbo]

Originally Posted by Walt Fricke

To me, the interesting question is why is his engine doing what it is doing. I've not heard that connecting an O2 sensor would cause an engine to run worse, given that the sensor is a voltage source, and a disconnected sensor means zero volts, which, say, a sensor which is short circuited to ground should look like zero volts as well.

Do the turbos get enrichment by connecting the intake vacuum (or boost?) from some convenient place to the lower chamber of the WUR? I know on my NA WUR, if I sucked or blew into a hose I hooked to the attachment on the bottom chamber of my WUR, I could watch the change in control pressure. Can't recall which way it went, but pretty strong movement. Makes sense, as the diaphragm in there is like a barometer bellows, so to speak. Interestingly, it affects the inner of the two springs which push up against the seal plate where the pressure changes are governed. On mine, the lower chamber is just open to the ambient air, though some models have a line to this for control purposes - maybe altitude adjustment?

So why does his car run poorly with the sensor attached?

Yep, a hose runs from a fitting on the throttle body, beneath the throttle plate, to a fitting on the lower chamber of the WUR. When the WUR is made adjustable, then you can increase or decrease the spring tension on the diaphragm so that the control pressure is lessened or increased in order to get the boost AFRs you're looking for. You can also add a solenoid into the boost pressure hose that will delay the boost pressure to the WUR so there isn't such an overly rich fuel dump at midrange/boost onset, like is common with these WURs.

The only thing I can think of that would make Spoke's engine run worse when he connects the O2 sensor is that the fuel head and WUR have been tuned without it connected, and then it goes too lean with the Lambda system in play.

[WP0ZZZ]

Originally Posted by Trakrat

I completely understand how an O2 sensor is used on my modern twin turbo car with an advanced ECU, and understand how the map changes between air and fuel mixture, along with other sensors, and changing the map from an open to closed loop.

But my 911 isn't that modern.... and I'm not really sure what is going on when the O2 sensor is connected and ALSO when it is disconnected.

Is there something that makes you think that the O2 sensor in your 911 is used in a fundamentally different way compared to a modern car?

[Walt Fricke]

WPO - do the O2 sensors in modern cars with ECUs and EFI use a dither system like the CIS does? Drive it rich when it reads lean, drive it lean when it reads rich, over and over and over - somewhere I read the frequency involved (though the actual system is PWM, isn't it)at maybe 60Hz?

Or do modern systems operate off a wide band sensor, and fine tune the length of the injection pulses with just a little bit of hysteresis? I think the aftermarket Electromotive EFI I use on a race motor will do that if you want, though of course I don't as it isn't hard to tune for WOT only.

Raw - I bet if Spoke took his car to a shop with a wide band sensor they could stick up the tail pipe, like the DynoPro place I go to, they could confirm or deny your guess. Having a dyno chart of your car is kind of fun all by itself, too.

[WP0ZZZ]

Originally Posted by Walt Fricke

WPO - do the O2 sensors in modern cars with ECUs and EFI use a dither system like the CIS does? Drive it rich when it reads lean, drive it lean when it reads rich, over and over and over - somewhere I read the frequency involved (though the actual system is PWM, isn't it)at maybe 60Hz?

Or do modern systems operate off a wide band sensor, and fine tune the length of the injection pulses with just a little bit of hysteresis? I think the aftermarket Electromotive EFI I use on a race motor will do that if you want, though of course I don't as it isn't hard to tune for WOT only.

The purpose of the O2 sensor (OP's question in the title) is to have a sensor for a feedback system that adjusts the fuel mixture based on exhaust lambda.

Is the OP looking for an in-depth explanation about how the feedback system is implemented in their own car?