First, it sounds like your car is fine (we have a saying about not fixing what isn't broken).

Second, your English is no worse than many others I've seen, and it is their FIRST language. You communicated what was on your mind.

You already learned that you cannot use a "Scotchlock" or otherwise pierce the shielded wire leading to the CPU. You are correct in the reasons for the electrical noise barrier. The place to check O2 sensor output voltage operation is at the single wire connector available near the firewall (the heater + and ground are in the 2 position connector).

When you checked the voltage with the wire disconnected from the harness, the injection system was forced to run in open loop mode (no feedback) which is why you got the steady, rich indication. Normal operation requires the sensor to be connected to the harness, and a volt meter (analogue preferred) should show a constant variation in sensor output which crosses the .70 volt mark often.

Start your diagnosis by using a voltmeter and familiarising yourself with the sensor output activity. If you don't see the proper range, move the ground probe to the cylinder head. I found a difference in readings in my own car. Remember, the O2 heater ground is not the same circuit as sensor ground. After confirming normal operation, connect the sensing wire of the AF30 to the junction and recheck the voltage activity (be sure to use the best ground you found with the meter for attaching the AF30). If the voltage has dropped, the add-on is loading the circuit and should be returned. If the voltage is still properly swinging over the .70 volt mark, you can safely leave it attached and check its display. There may be a fuel type selector, or more likely, the small differences in ground potential may have caused the meter to read out-of-range.