Here's the deal. I got an Integra LS, all wires stock & unmolested. Old O2 had a bad heater, so I got a genuine Honda/Denso O2. It is very lean at medium loads below 3k rpm.

I've tried a known-good sensor, also a different ECU. Problems only go away when in open loop (no O2 corrections).




Any ideas? Any way to test the O2 signal wire?

 

Maybe the car is running lean at medium loads below 3k....

 

Not likely. When I added enough fuel to the cells in question, they ended up nearly double of those around them. It also goes away when it's unplugged or in open loop.

I'm thinking it's a short on the signal wire. It's just very unusual that only certain load/rpm points are affected. Only ~7" Hg below 3k rpm.


I just remembered that OEM GSR headers vibrate around that RPM. Maybe that's why.

 

What cells? I thought you said the vehicle was stock? there will be no reading in open loop, still could mean its lean...

 

I temporarily used my other car's EMS to test it out with a stock P75 map. It did the same thing, so I datalogged it with the sensor enabled and then disabled in open loop.

I tested all the wiring and can't find a short, so I decided to just readjust the final fuel value. Basicly told it the injectors were smaller. After that it was only lean around idle in closed loop. I haven't run it much with this "fix", but that may have been the problem. I just don't know why it was only too lean in a small section of the map.

I'll be swapping to the larger injectors (for a JRSC) sooner than later, since these injectors don't seem to be showroom condition anymore. You never really know when you buy an old Honda anymore.

 

The way to test the o2 sensor is to keep the rev at approx 2k, and it should fluctuate from 200mv to 800mv approx 2 times per 100milliseconds.

TO make sure it is working while holding the rpms at 2000 have somebody insert propane through the intake. You should see the o2 sensor got 800mv or higher (not higher than 1volt). Then stop the propane and watch the o2 sensor drop to approx 200mv then hit the throttle. It should react.

In reality you need an osciliscope because a multimeter is too slow.

 

A multimeter is fine for a standard o2.

 

so a reading of 070 to about 918 mv is way out of range correct???? Thats the readings mines putting out at 2k rpm.

 

Nope, thats fine as long as it is consistent and the O2 reading does not hang high or low and contiues to switch. The 200mv-800mv reading is only a guideline. Engine age, milage, wear and condition will affect O2 readings. Risetime (how quickly it will respond to changes in exaust gas) and switch rate(how many times it switches between high and low) are all that matter.

 

Ok. Thank you. It doesn't hang at all. Seems to move pretty consitantly.

 

what are you defining as very lean?

 

Around 0.01-0.04 volts on the o2 sensor as read by the ECU. So well over >18:1

I think a previous owner put in some really slow injectors. Low load/rpm was lean, WOT was fine. The smaller the pulse width (larger percent dead time), the leaner it got. I'm going with new style Bosch EV6 injectors for the S/C were exact dead time can be input.