This tale is that of a 2005 Stratus, 2.4 (vin J) with automatic transmission. The MIL is on, and a code P0491 (air injection system performance) is stored. The car only has 39.000 miles on the odometer and is very clean. After checking for applicable TSBs, and finding none, I pulled the code information to better understand how the code is recorded.
The air injection system feeds extra air into the exhaust when the engine is cold to more quickly heat the catalytic converter and bring it on line. This code is recorded when the PCM detects too little or too much airflow through the system, and is monitored with the engine running and pump commanded on. Feedback to the PCM is provided by a dedicated MAF sensor located on the intake side of the air pump assembly, and it is this feedback that is primary consideration in the PCM's decision to turn on the MIL. It is a 2-trip code.
Freeze frame information, however, did not tell me which way the system had failed. This was a hard fault, and not an intermittent, because the "Good Trip" counter was still at "0". How best to proceed with my diagnosis?
I like to structure my diagnostic procedure to mimic what the PCM is doing as much as possible. In the case of the P0491, the PCM commands the air pump relay on, which in turn powers the air pump motor. Air flows through the filter, then the MAF sensor, before exiting the pump. From there, the air flows through a tube to a one way valve located above the exhaust. This code occurs when the PCM knows it turned on the pump and then doesn't see the airflow it expects to see, high or low. The pump only runs when the engine is cold, and for 20 seconds or less.
Lucky for me, the car was cold when I got it. This made my choice of approaches a little easier. I decided to hook up my scope to see the system in actual operation. I could verify all I wanted in a few screen shots. Using three channels, I connected the scope to:
1. The PCM relay control wire. Connecting here would tell me that the PCM was commanding the system on. This pin provides the power to the air pump relay. This is the gold trace you'll see in the capture.
2. The MAF signal wire. Connecting here would tell me what the MAF was telling the PCM. The red trace is this signal.
3. The pump power wire, before the relay. Connecting a current probe here would tell me the pump was running, and is the green trace in the capture.
Here are the captures I saved:
This first capture is taken with the system connected as it should be. The gold trace tells me the PCM is indeed sending the power to the relay, and the pump current pattern tells me the relay is engaged and the pump is running. MAF increased only slightly, however, to just under 1.0 volts. Could it be the MAF was reading incorrectly, or that the correct reading wasn't reaching the ECM? Could it be a restriction to airflow in the plumbing? Let's disconnect the air tube at the one way valve and try again...
Luckily, there was enough "cold" engine to run the test one more time.
It didn't run as long this time, but the increase in the MAF signal led me to believe it was an air restriction. I removed the one way valve for a visual inspection...
How did this valve get so corroded? And why is the current flow on the motor higher with the restriction removed? While there is no specification, 34 amps had to be too much. The car was "new" to our customer, but I suspect that water had intruded into the pump somewhere in this car's past. I shared my concerns with my boss, and she (yes, I said "she" 
) authorized replacement of the air pump and one way valve.