if you check out at the pictures of the new AEM gauge PCB that I posted, there is no Bosch chip there. Just amps and switches. ( and a big inductor for the heater )http://forum.diyefi.org/viewtopic.php?f=6&t=2267&start=140
The trace that I got out of it, hitting it w/ CO2, is on the next page. It doesn't appear to have the characteristic "rounding off / tapering" as it settles on its target value, like a PID. It does go slightly over-rich ( ~ lambda 0.98 ) for about 20ms.
With the dv/dt it shows, approaching its target value, that "dip" cant be a characteristic of a control PID. But, it could be the pressure sensitivity of the sensor, if there is a pressure wave? The spec shows about +1% dIp/Ip per psi. I guess 20ms isn't too hard to believe.
Or... it could be a quick temperature variation from the cold shot? Spec gives dIp/Ip = 4% / 100C
Or, some combination. Or something completely unrelated? You seemed to predict this behavior. So, I thought I'd get your take on it
I have seen a similar "dip" on an LC-1 running in "instant mode", with the same test. But, there was not nearly as much resolution, time-wise ( maybe 3 samples ). And, it was harder to pick it out of the noise. But, it's there.
I guess I could repeat the test with the sensor already saturated at lambda 1.0 and look for the reaction to the shot. But, then, I'd have to set it all up again