As far as I know the German "KAT" models have the Lambda system with a catalyst. If the catalyst was removed then what you are seeing at the tailpipe are "engine out emissions", and I can only estimate what they should be, but CO should probably be about 1-2 percent, a few tenths percent O2, and and probably no more than 300-400 ppm HC.

On a Lambda system the basic mechanical mixture should be set to center the system at about the middle of its control authority range. This is established by setting the "duty cycle" at near 50 percent and the reading at idle and 2500 revs, no load, should be within ten percent of each other.

As you deviate from this spec the Lambda system is forced to operate farther from its center point and at some point can reach the limit of it's control authority.

You can familiarize yourself with the Lambda system by viewing the various service documents at http://mb.braingears.com

I think there is a lot of misunderstanding of what the mechanical mixture adjustment in the air flow meter does. On a non-Lambda K-jetronic system it does vary the idle mixture. On a Lambda system it merely sets a point from which the Lambda system will correct the mixture to maintain stoichiometry, unless the adjustment is so far out of whack that the Lambda system is forced to the limit of its control authority. Within reason, diddling with the mixture screw will not alter the actual mixture on a properly functioning Lamda system. It will just alter the center point from which Lambda correction is initiated and if you diddle it too far you can force the Lambda system to the limit of its authority, which essentially defeats the system - something you DO NOT want to do!

The purpose of measuring and the duty cycle and adjusting the mixture screw is to establish a basic mechanically derived mixture that is close to the middle of Lambda control and that duty cycle is 50 percent. I prefer to set the idle duty cycle at about 45-50 percent, which improves cold starts IMO, especially in warm weather, as the starting and warmup mixture is established by the air flow meter prior to the system going into closed loop mode and a duty cycle of less that 50 percent means the basic mechaically derived mixuture is slightly rich. If the idle duty cycle is set to the 45-50 percent range the reading at 2500 should be in the 50-55 percent range, which meets the 10 percent max difference spec.

So it's not clear to me what your mechanic did, but the above is what should be done. Also, your car probably has a variable EZL resistor - a round knob next to the X11 diagnostic socket. Your owners' manual should have info on its function, and so does the the braingears site. You want to use the setting that provides the most aggressive spark advance curve for your fuel quality. If it's set too low, fuel economy can suffer.

When faced with a KE-system emission issue, the first thing I recommend checking is the duty cycle and scoping the output wave forms of the O2 sensor and ignition system. Then I go from there. If the Lambda system (including the 02 sensor) and ignition system are operating properly then reducing the spark advance by either adjusting the R16/1 resistor (a rotary knob on most models except USA, Japan, and AUS, which have a fixed resistors installed in a wiring harness pigtail) or disabling the vacuum advance can significantly reduce emission readings. This is particularly true of high mileage cars that have degraded catalysts due to catayst aging.

If you have to pass an emission test to register your car, it probably wont pass unless you install the proper catalyst.

I've posted more background information at the following link and the links this post refer to:

http://www.peachparts.com/shopforum/tech-help/117048-successful-ca-asm-emission-test-ke-fuel-system-post833484.html#post833484

Duke