legacycentral bbs

I noticed after I did my tranny swap (battery unhooked for a few days) that my car stopped hesitating. Most OEMs allow about 5-10% authority on the O2 sensor to trim good emissions while idling and cruising. At WOT the AFR is really good but under cruise and idle it's insanely rich. Like 9:1 AFR rich...

Besides poor fuel economy this gives really poor driveability. I wonder if anyone else has experienced these symptoms from a faulty/unhooked O2. It doesn't throw a CEL but that's non-OBD2 for you.

I suspect that someone messed up and gave the factory ECU way too much authority over the fuel trim based on the O2 readings...

-Michael

It's not really about too much authority... the roughly plus or minus 19% our stock ECU can trim fuel isn't that odd. It just seems that the Hitachi ECU at least seems to behave maladaptively in some cases. Like, it sometimes doesn't realize that it should be ignoring the signal from the sensor, instead allowing trims to hit the rails. Or it can get into an overrich misfire positive feedback loop.

I'm confused -- Is your sensor hooked up?

Sensor is currently unhooked.

On many implementations they will use a pulldown resistor or something so when the signal is out of range it immediately trips the CEL.

I'll see if I can hook up my wideband's NB output to the O2, reset the ECU and see if that's related.

-Michael

Our ECUs only throw a trouble code if you're cruising for a while with no oxygen sensor signal. Otherwise, it's content to just wait for the presumably cold sensor to reach operating temperature.

I think something's wrong if you're running 9:1 air/fuel ratios with the oxygen sensor disconnected... unless the engine's cold?

My car runs very close to stoichiometric at low load without the oxygen sensor.

Is your MAF sensor in good shape?

I'm not sure about the MAF being in good shape or not but I hooked my wideband's NB output up so the ECU thought it had an O2 hooked up. After a few minutes the AFR has reached the 12.5:1 area and I believe it may climb as time passes.

About my hesitation problem it only starts happening after a few days of driving after an ECU reset.

-Michael

That's weird that it shifted to around 12.5. It should immediately start oscillating around stoichiometric.

Looks to me like it's PID based (looking at the overshoot and settling time) and they've obviously got a map of EGO corrections with some sort of average learning.

On startup it used to start at 8.8:1 (ugh!) and climb to 12.5. Now it idling it starts at about 11.5:1 and idles up to 14.7 so it's slowly learning. Now if I start driving it immediately goes to the really rich correction and slowly starts leaning itself out.

When I installed the wideband I unplugged the battery for about 5 min but that may not have reset the ECU successfully.

-Michael

But the closed loop control system doesn't take hours to react, it takes like a split second. If it reads below stoich one instant, it will read above stoich the next instant. Am I missing something?

It does only take about 5-10 seconds to trim the fuel up to stoich but the starting point appears to be stored and averaged so as soon as you leave that particular RPM/airflow point it will load another trim starting point. Over time of seeing lots of negative correction it will eventually trim everything so it starts at the proper trim. Unfortunately I don't have the time to reverse engineer the factory ECU. Instead I will begin concentrating on decoding the factory trigger pattern and adapting my ECU to work with it. Then I'll make a plug and play unit. Install and tune that.

-Michael

When I had my '94 touring wagon, I ran it for 6 months without an O2 sensor. I never had a single issue with it...

Check those plug wires!

Twice, when I had weird misfire issues, it was just a bad plugwire.

I just solved my current misfire issue today. #1 plug wire = toast

I did some datalogging today. Seems that the O2 is only part of the problem. I'm beginning to suspect it's got a weak ground on the MAF.

When idling it will be sending 1.25v and running about 10:1. Then it starts leaning out as if the O2 correction is kicking in but the MAF voltage is dropping. It eventually settles to 0.85v and the idle is stoich. I've got more digging to do but I'm going to measure the ground potential on the MAF to see if it's sometimes floating at about 0.5v.

*sigh* I hate electrical problems.

-Michael