PeachParts Mercedes-Benz Forum - View Single Post

stevebfl · #43 12-26-2003, 12:24 PM

I am not sure which question is one or two but let me rephrase the concept for the record.

It is wise to not unplug connectors with circuits in operation, but it is no problem with an O2 sensor SIGNAL. I yelled that because the other two wires in some sensor connectors are heater circuit wires that carry 1-2 amps and souldn't be disconnected when powered. The 300e has separate connectors for the heater circuit and the signal wire (green) can be disconnected while running although it isn't necessary.

The lambda fuel correction circuit manages "closed loop" control. Closed loop is a control concept used for many systems but in this case what it means is that an exact mixture is maintained by monitoring the results and continually correcting. When the mix deviates rich the voltage of the sensor (the sensor is a power generator) goes high (above .45v). The system sees this and changes the EHA to lean out the mix. The sensor then goes low in voltage (less than .45v) and the EHA current is dropped to lean out the mix. All of this takes time so the actual voltage goes to .8 mabe .9 before turning around at the rich end or .2 to .1v on the lean end. This all happens more than once a second in a warmed good running system. Going down the road it probably happens a couple times a second or more (the hotter the sensor the quicker its readings change).

So, as diagnosticians we wish to see whether the controller can do the EHA control necessary for closed loop control. We will probably be doing this to a system that is not in closed loop control. The range of control at the EHA is about -12ma to +12ma (from lambda control). In reality this is not quite enough to take a car from running badly rich to running badly lean. (THIS IS IMPORTANT to realize). It is enough usually to take a smooth running engine to roughness in both directions, but it may only be enough for one direction. This is because the engine will run fine in a mixture range defined in percent CO exhaust gas of from about .5%CO to about 6% CO. Lambda control keeps it to .1%CO to 1% CO. So if you send the maximum lean correction to a car runnning 6% it may only take it to 1.5% and the car still runs great. On the other hand a car at 5% sent rich will go off badly at over 6%.

So now we have the O2 sensor disconnected. If we attach a voltmeter to it, the range of voltage possible will be 0 - 1v. This gives us good evaluation of mixture in the 0-1% CO range. It gives no idea about the 1-6%CO range in which a car will run fine.

If we monitor the sensor (while disconnected) and apply our own fictitious values to the open lead that goes to the controller we will initiate lambda control reguardless of actual mixture.

It is easy to apply 0v by grounding the lead. It is a little more complicated to creat the 1v signal but the 12v through the Cap'n method does good except you may have to replace the Cap'n with your own body as he is busy. By putting one hand on battery + and holding the green sensor to controller lead in the other hand the necessary voltage is transfered to the controller to imitate a rich signal. If one has an exhaust gas analyzer one can see the full range of correction that can take place with these maximum value inputs. This gives a view to the capability of control.

A properly set up mechanical portion of the KE system will have a mixture that can be read on the open circuited sensor. Some early systems suggested adjusting mixture with the voltage from the disconnected sensor. The mixture was set to .5-1.0v. Then the sensor was reconnected and closed loop had no problem because its capabilities could handle that base setting.

During the range observation (explained above) one should be able to get a view of this proper voltage range somewhere during the lean correction. That is the 12v through the Cap'n connection. Controller sees over 1v, interpretes a rich mix and rather slowly changes the EHA to -12ma eventually ...it takes a while .. maybe15-20sec (I'm guessing I have never timed it).

If the car was actually lean when this process starts the engine will probably die. Most often the mix is rich and so far above 1% CO that full -12ma may only get the system down into the less than 1%CO range necessary.

These are the views you can get. Getting meaningfull answers will require aquiring realistic experience of these happenings so that conclusions can be formed.

The thought of replacing this exquisite Bosch system with anything domestic blows my mind. This simple gourgeous system replaced with a crude, undiagnosible, carburator....there is real progress...not!