Many of the sensors have a fall back, so you could unplug them and see if problem continues. Like Cam position sensor, fuel rail temp sensor, etc. MAF not a candidate, nor is crank position sensor.

 

I agree 1 @ time.

 

Drops in?

 

Well, you can always throw parts at it. But, if there are parts you could unplug, and clean connections, that could help. If there are parts you could unplug, the engine run at a reduced fallback state, but problem continues; that might reduce the number of sensors to throw at it. Some salvage yards are really cheap on sensors, and my yard even refunds, but not your environmental fee. Some sensors show up on the scanner, like ECT and IAT and MAF and TPS, so you can see them in real time. You already did the CKP. The remaining list is getting smaller. You should be able to unplug fuel temp sensor and truck run. Same for CPS.

 

ET - perhaps one way to eliminate the electrical wire issue - if the codes will comeback from a cold start at idle, or in the driveway revved up slightly; consider connecting your meter and watching for any fluctuation of current in the injector fuse circuit at the point problems happen. Or, if yours only draw say 3 amps, put in a 5 amp fuse and see if it blows, from a momentary short. Of course be prepared for Disco with no power and no way to re-crank quickly. Can also shake and prod wiring while metering the current.

 

Haven't had time to do more detailed diagnosis but here's some curious food for thought:

In the morning I start the truck and let it idle for a minute while I feed the cat. Cold the engine is high 30s or low 40s, after a minute it's maybe 60 - 70 F. By this point it's in closed loop. I start driving down the street and a minute later goes into open loop and the check engine light starts blinking (signaling misfires). By this point the engine is up to about 120F so I'm not sure why it's in open loop when it was closed earlier. This goes on for about 20 seconds and then it goes into closed loop again and the check engine light stops blinking.

The question is.. Why does it go into open loop? And, why does it misfire when in open loop? Or perhaps it's in open loop because it's misfiring.

 

Closed/open loop is not entirely driven by engine temp. Mine goes closed within 10 seconds after restart. It will go open loop on WOT. It may go at other times as vehicle tried to adapt to conditions.


From the GEMS manual:

During Stage 2 monitoring lambda feedback control is switched to use the downstream oxygen sensors.

Won't it be surprised when there is no signal from the deleted cats....

And there is more:

Road speed must be greater than 20 mph.
Steady engine conditions must be true
The catalyst must be deemed to have reached "light off" indicated by the following:
Engine Coolant Temperature is > 70

0 C AND Average Oxygen Sensor Heater Supply < 90%.

9. Monitoring Time Length / Frequency of Checks
The time taken for catalyst monitoring to complete the test depends on the switching frequency of the feedback control system, which in stage 2 depends on the catalyst's performance. If the switching frequency is taken as an average of 1.5 Hz then the stage 2 catalyst monitoring test will take 7x15x1/1.5 = 70seconds at stable valid conditions. Stage 1 will take a similar amount of time to complete.

So perhaps the internal test procedure that checks for cats actually working, which is a background test, is happening about then. Missing cats throws incorrect number into the mix. There is a Gulf States flag in the ECU for "no cats", and I suspect some markets only had front cats.

 

Forgot to mention one other observation. On one occasion, when the CEL was blinking and it was in open loop, I decided to punch the accelerator a little. As soon as I did it stopped blinking and went closed loop.

Perhaps the same aliens that build the pyramids are experimenting with my ECU in a grand psychological experiment.

 

Buzz this might just be on the right path..

"Steady engine conditions must be true"
That would explain why punching the throttle kicked it out of that mode.

I'm guessing it may be the rear O2 sensors, and not the de-CAT causing it for two reasons. I only ever replaced the front O2 sensors, and the misfires were present long before the CAT delete. I'm pretty sure the rear two are bad and actually unplugged them a few weeks ago. I figured they were useless emissions garbage.

So either I'll replace the rear two O2s, or figure out how to set that Gulf state flag.

 

What brand O2's did you use? Were they exact fit or did you need to splice the connector on?

The reason I ask is the only time I have seen what you have now described is when using sensors from a bad batch (Bosch), or using universal sensors spliced onto the original connector. When my buddy purchased his "96 it was running like crap. We went through everything before finding the sensors were spliced to the old connector. The GEMS uses Titania elements in the O2's and these seemed to be Zirconia, making the ECU very unhappy.