It may have been what you said, but I did not get it until the Chris-Bob clarified!

I would think that you could use a breakout box to "echo" the values back into the ECU - but really just for troubleshooting.

I seem to recall that some of the sensors send the information back using an algorithmic statement, code or specific sequence key to try to eliminate the possibility of doing exactly this.

I am SURE you can get around it, but other than the troubleshooting assistance it would give (include/exclude the sensor you are "mimicking") I really am running out of ideas to justify the time and expense to put something like that together...

Realistically, I would think you could use a "scope" (Oscilloscope) to look at the input/output in order to generate the expected responses. (A laptop running LINUX could be configured as an O-Scope in a matter of minutes - challenge being the right "connection" to get the data from the sensor or ecu into the PC) I have done that literally THOUSANDS of times with computer interfaces in order to identify the value that needed to be changed, parity settings etc...

Interesting idea though!!

 

A BoB would work great for troubleshooting. If you wen through with a pot and DVOM hooked up to each sensor in question one at a time, you could tweak the numbers going back to the ecm to see if the sensor is the cause of erratic/high idles, misfires, hot start/running issues, so on and so forth. If someone built one that worked and posted the results, it would help a lot of persons that are unwilling to throw money at parts that we are unsure of. I for one refuse to spend money to replace parts that I cannot find a fault in unless I am upgrading the part for performance.

 

Well at least we have a worthwhile exchange of ideas here.

As far as the front O2 sensors, do they just supply a changing resistance or is there a waveform associated with that input to the ECM?

If it is merely a resistance shift, a BOB with the correct range variable resistor could be subbed in.

Might even be easier than getting to the damn connectors!

Anyone know right off-hand which sensors are just a resistance change? These would be easy candidates.

Sort of like the mechanical choke I had on my 1967 Chevrolet Sport Van 90 Window van with the 283 4 barrel mounted in a doghouse enclosure right between the front seats. You could raise the cover and adjust the carb while you were crusing down the road. Sit in the driver's seat and do a tune-up.

We jacked up the rear end and mounted the biggest mud grip on there we could find and took that beast places a lot of 4by4's would have a problem with today.

Also built a VW based hunting buggy with a 36 HP VW engine and a 3 inch tubular frame for the rougher terrain in central GA.

 

If you have a OBDII scanner that reads live data you can trouble shoot all kinds of sensors.
You can watch the resistence of the O2's, you can watch the MAF readings, engine temp, air intake temp, TPS %, timing advance, engine rpm, road speed...all the info a Shade Tree Mechanic could ever need.
My scanner shows the output values of all 4 O2's and I can watch them as they warm up and as I drive.
Both of your front O2's should show the same reading, or at least close to the same, same thing with your rears.
If you guys remember I diagnosed my low MPG with my scanner, it was a faulty coolant temp sensor, I watched the engine temps on my scanner and they jumped all over the place confusing the ECU.
All of the parameter's have a default setting, so if a sensor totally fails the ECU will go into "default" mode, set off the CEL and try and compensate with one less input.
The GEMS with distgribitorless ignition is only 2 steps away from being carburated, it ain't rocket science.

 

I have the GEMS version in my early 97. Where can I find some specific info about it?

Which 2 steps is it, A carb might not be a bad idea.

Which scanner do you have Spike, I might have to pick up one that will do live real time monitoring.

My basic idea was modeled after one of the tasks I performed on the Flight Deck of the USS Kennedy in the early 70's. I was a Aviation Electrician and was responsible for the Flight Controls, Instrumentation, AutoPilot and Automated Carrier Landing System when we qualified it at that time.

Every 21 days, I had to climb in the cockpit of an A5 with the engines running to supply hydraulic power and using a large "Simulator " box, run thru a sequence of test with two multi-position switches that simulated faults in the aircraft and make sure the dual yaw aug and dual pitch aug would cut back to single systems when the fault was present and reset when the fault was cleared. You set a switch then performed an operation, verified results and went to the next step.

If it is only a resistive change, you "Dial In" the value that provides the desired results and be free of the poor performance that was caused by a faulty/worn out sensor. Might not be perfect, but may be worth a try!

Anyone have a "Defective" ECM on hand that they replaced? I would like to open one up for study purposes.