momotaro

I have a 4100 series Wideband O2 with Bosch sensor. Car is 1996 Miata running a stock motor. Ground is to ECU ground. Power tapped from the stereo. White wire (analog) to megasquirt ecu. Sensor is screwed into stock down pipe about 12 inches downstream of the head. No exhaust leaks that I can see or hear.

Controller, powers up just fine. Idle reads correctly: about 13.5 which smells about right (this is with the stock base tune provided by megasquirt, I know it is rich). Analog read out to the megasquirt is correct at idle.

The problem: as soon as add throttle, the controller goes rapidly to full lean. This can’t be correct as the car runs just fine. As soon as I add any throttle, it goes full lean. If I go steady state light throttle, I can sometimes get a stable AFR, but even then, it is leaner than I suspect is true (17s).

I tested the sensor with the gasoline soaked rag test, and it seems fine: sees full lean in the air, and full rich and stays full rich in the gas rag.

So is it a controller problem, a sensor problem, an install problem, or something else?

AEM can't figure it out. It is a used unit out of warranty.

18psi

iTrader: (76)

you have lean tip in?
tune your tip in enrichments

momotaro

Thanks for the reply.

I might have to tune my lean tip in, but I think the problem is something else, as it will go to full lean and stay full lean so long as I am in throttle (partial, full, does not matter). Car runs fine, so it can't be that lean, right? Don't know if I should try a new sensor, or just get a whole new unit.....

18psi

iTrader: (76)

oh. I thought it just dipped, because that's pretty common.

no idea in that case.

momotaro

Yeah, weird.

Should I risk buying a new sensor, or just get a whole new unit?

Fireindc

iTrader: (13)

I had an AEM sensor that failed, it would gradually get leaner and leaner up until the point it would PEG lean, and no matter what the car did it was 3-4 points leaner than it should have been. It would start out fine the next day, but at the end of the day it would return to showing leaner than it actually was.

Replaced the wbo2 sensor, same gauge/controller and it works fine now.

momotaro

thanks for the reply. Might be worth trying a new sensor then.

Fireindc

iTrader: (13)

fwiw, i found one on ebay for ~60 shipped brand new, so you might want to check there.

18psi

iTrader: (76)

also check your connection/pins on the actual big plug from the sensor to the gauge.

I had a buddy who had funky issues with his turned out to be one of the pins was slightly pulled out and wasn't making full contact

momotaro

again, thanks for the posts. I will go back and check my connections another time.

momotaro

Ok, I checked my connections and they seem fine. Pins look fine. I changes the sensor and I still have the same issue.

So I guess it is either the controller or the aem harness? Someone have any other suggestions before I get a new wideband. I have already paid once for being wrong (new sensor).


Thanks

momotaro

I am sharing the ecu ground and. The radio power with my othe gauges. I don't think this makes a difference? Do I need a dedicated ground only for the wb?

dlvandyke

Grounds can cause funny problems. Can't hurt, or cost much, to run a dedicated ground to chassis or engine block. Might even try running 12v straight to the batt, don't know if you unit has an on/off switch though. I suspect the heated o2 sensors can draw a bit of current too. hth

concealer404

iTrader: (3)

1) Make sure your sensor is grounded separately from anything else, and is grounded to ECU. Don't tie anything else in with it.

2) Don't pull controller power from radio. Illumination power from radio is fine. Sensor/controller power, jank.

3) If you do need to buy a new sensor, do some digging around. Too lazy to do it again right now but i found them in the $48-$50 range shipped with some creative cross-referencing.

4) Buy an Innovate MTX-L.

momotaro

I am hopeful it is my ground. I have coming off the chassis ground below the TB, but the wire passes by the back of the head close to the plug wires before it enters the cabin. Maybe the hight voltage from the ignition is creating noise? I am going to redo grounds and power. Using some shielded wire if I can find some.

curly

iTrader: (25)

Do you have a stock ecu you can try, just to eliminate it?

momotaro

Not sure what you mean curly?

concealer404

iTrader: (3)

He wants to make sure that your tune isn't in fact, garbage. Can you swap to stock ECU and see what happens?

(I doubt that's the problem.)

momotaro

I doubt it is the tune. It drives fine, and would not be drivable with 18+ afr.

I think it is the install, the cables or the controller.

I have the stock nb in for the drive to work and it drives perfect (smells a tiny bit rich).

I am hoping that it is just the ground.

For electrical engineers: if my controller ground wire passes over the plug wires (high voltage). Will this create a small current in the ground wire?

momotaro

What the fudge am I doing wrong? I tried a different ground, I tried a different power source. Same problem. I feel like it must be a wiring problem.

Could the base map really be that lean? It drives fine.

It wbo2 is behaving like a vacuum gauge. At idle it is 13.5, with throttle it goes full lean, in decelleration it goes full rich.

Is it the controller? I feel like I am missing something simple.