Community Forums:

Brett, the 87 uses LH2.2, so you can use an LED connected to the "CO test point" which I believe is the pink wire in the two wire plug near the driver's side strut tower. This will monitor the ECU as it switches back and forth adjusting the mixture in response to the oxygen sensor's measurements. The adjustment is physically on the AMM, but not actually part of its circuitry. If you remove the tamper cap, there's a potentiometer underneath that serves to fine-tune the match between the analog output of the AMM and the measuring input of the ECU. I think Bosch intended it to be adjusted only once, when first an AMM is paired with an ECU, but did not anticipate the degradation of the AMM as a reason to reset the adjustment.

I've done the sensor without removing the manifold by removing the AMM and its piping, disconnecting the brake booster hose, oh, and the flame trap to throttle body hose, then using a 10" extension 3/8 drive, deep 6pt 19mm socket. Surprisingly no appreciable coolant leaked, and the ECT was obviously the trouble.