Bogs for a second and then takes off

[Marc Haibeck]

Dominic,
Your question 1. I find that oxygen sensor performance is better on OE manifolds. The OE manifolds are designed to jet each of the four cylinders across the sensor. With headers usually only two cylinders are in the sensors path.
Q 2 and 3. Extending the oxygen sensor cable has no effect on the sensors voltage to the ECM. Voltage drop is determined by resistance x current. Ohm's law. The sensors's current is so low that if the cable was lengthened by three feet the voltage drop would only be a millivolt or two. An insignificant loss. Adding tape or sealer to an extension cable can cause problems. The sensor gets its reference air through the connectors. Only use crimp connectors.
Q 4. I have a nice scan of the fueling going falsely rich on deceleration. At 2300 rpm the throttle was closed. The FI pulse width is 3.2 ms, then the sensor goes to 898 mv, very rich. The sensor stays at about 880 mv for two seconds. Meanwhile, the FI pulse is ramping down to 2.0 ms. Then the sensor snaps quickly to lean. About one second later everything is normal with a FI pulse os 3.2 ms. The new sensor switches faster from rich to lean, without running down the FI pulse to the point to where the engine misfires.
I always set the secondary throttle opening delay to zero, even though it doesn't matter if the secondary throttles are not present. The delay is for the opening of the ports, not the fuel.

[Paul Workman]

Originally Posted by Marc Haibeck

Dominic,
Your question 1. I find that oxygen sensor performance is better on OE manifolds. The OE manifolds are designed to jet each of the four cylinders across the sensor. With headers usually only two cylinders are in the sensors path.
Q 2 and 3. Extending the oxygen sensor cable has no effect on the sensors voltage to the ECM. Voltage drop is determined by resistance x current. Ohm's law. The sensors's current is so low that if the cable was lengthened by three feet the voltage drop would only be a millivolt or two. An insignificant loss. Adding tape or sealer to an extension cable can cause problems. The sensor gets its reference air through the connectors. Only use crimp connectors.
Q 4. I have a nice scan of the fueling going falsely rich on deceleration. At 2300 rpm the throttle was closed. The FI pulse width is 3.2 ms, then the sensor goes to 898 mv, very rich. The sensor stays at about 880 mv for two seconds. Meanwhile, the FI pulse is ramping down to 2.0 ms. Then the sensor snaps quickly to lean. About one second later everything is normal with a FI pulse os 3.2 ms. The new sensor switches faster from rich to lean, without running down the FI pulse to the point to where the engine misfires.
I always set the secondary throttle opening delay to zero, even though it doesn't matter if the secondary throttles are not present. The delay is for the opening of the ports, not the fuel.

I was unaware of the sensor sampling the air via the leads (i.e., though the wires) too. But, I can testify to the fact the sensor depends on the air it gets through the port where the wires enter the sensor.

After replacing one of my O2s, Just to add some reliability for "good measure", I sealed the sensor where the wires enter the sensor. I packed the area with RTV. Bad idea... That new sensor and one more failed in short order before I discovered my folly. So, I shelled out another $60 for a third sensor, and left the little opening alone. In the 5 years after it never failed again "AS IS".

[StickWalker]

I believe I found the issue. I did not realize how bad the TPS wire connector had become. I've replaced it and seems to be driving better

[XfireZ51]

[QUOTE=StickWalker;220044]I believe I found the issue. I did not realize how bad the TPS wire connector had become. I've replaced it and seems to be driving better

Occam's Razor.