More ecu madness.

[boostjunkie]

I've been having random problems ever since the rain and the snow we've been getting lately. A/f ratios are completely off the mark, and I'll sometimes even get mutiple leds going off at the same time!! My CE light will go on, and then shut off every now and then as well.

Checked the wiring to the A/F gauge. Everything's fine. Checked the wiring to and from my O2 sensor. Fine. Checked the codes this morning.

31 - TPS - This TPS has been the thorn in my side for a while now. I'm gonna swap in another TPS I got from the summer and see if that will help. FWIW, I was never REALLY able to adjust the current TPS to spec.

41 - A/F Learning Control - Uh-oh. If this is really fried, is there another fix besides getting a new ecu? Also, could the fact that the TPS is screwed up cause the ecu to be inefficient in learning? I'm guessing since the ecu relies on TPS signals as a secondary input for fueling/timing that this might be the case.

I'm going to reset the codes and swap out the TPS just to be sure.

Any ideas?

[vrg3]

I don't know what exactly has to happen to make the ECU throw code 41. But if your A/F gauge was giving you bizarre readings, the ECU was also probably getting bizarre readings from the oxygen sensor, which would make it freak out and shut off the learning control.

Can you be more specific about how the A/F readings are off the mark.

[boostjunkie]

The o2 sensor readings are always in the lean area, with multiple leds lit up at once, but I can tell there is variance in the voltage. Once I go WOT the reading stabilizes and the o2 sensor reads stoich. Sometimes, the o2 sensor readouts would be normal again, but most of the time the o2 sensor exhibits the strange symptoms described above.

COuld a partially clogged cat cause faulty readings?

EDIT

One more thing. The car's been backfiring a lot more than usual.

[vrg3]

Stoich at WOT on a turbocharged car is way too little fuel! You should use boost as little as possible while resolving this problem.

I don't know if a clogged cat could make the oxygen sensor read incorrectly, but it's possible I guess since the sensor works on pressure differentials or something (oxygen sensors are so amazing).

Or maybe whatever clogged your cat has also clogged your oxygen sensor. Do you have any idea how old the sensor is?

[boostjunkie]

Yeah, I know!! Once I noticed A/F getting screwy at WOT throttle I let off.

Well, I put in the newer TPS sensor, set it correctly to spec for the idle switch and cleared the stored codes via connector method. This time checked to make sure the codes actually cleared and started her up. No CE light came on even when I went WOT for a bit. The same symptoms as before . . . except one time I got REALLY ballsy (yeah, I know). Held WOT through the conditions making sure there was no audible knocking. I was taking a gamble with the ecu ignoring the o2 sensor readings anyway.

Here's the weird thing.

Checking all sensor readings, boost hit .76 (normal), IDCs increased normally, and nearing 5500rpm all of a sudden a/f ratio went to rich again. Tried it again but I couldn't get the same results.

I'm pretty sure the 02 sensor was replaced when Adam got the downpipe, which would've been a few months before I got it. I'm guessing a little more than a year and a half ago.

One more thing.

At idle, or when the engine is at really low load (<1/10th throttle), the o2 sensor doesn't even register on my gauge. Under a little more load the gauge registers, but like described above.

Doesn't the o2 sensor reading change dependent on heat? I can't remember how but I know the accuracy of the sensor changes. That's why I was thinking there is a restriction in the cat. Wouldn't a clogged cat cause higher temps just before it as the exhaust gases become backed-up?

[vrg3]

The oxygen sensor reading does indeed depend on temperature. But a higher temperature usually makes for a higher reading, I think.

It's weird that you wouldn't get readings around stoich at idle. If your engine is warm it should idle in closed loop.

At extremely low load the ECU assumes you're coasting in gear and injects almost no fuel, which makes most "A/F gauges" go blank. This load level is much lower than at idle, though, and certainly much lower than you'd get with any nonzero amount of throttle input.

Do you trust all the oxygen sensor wiring, going to the ECU and to the gauge?

Also, do you have the ability to disconnect the sensor from the ECU but leave it connected to your gauge?

[boostjunkie]

I guess you're right.

"This is from an e-mail I got from Bosch a while back.

MinV MaxV A/F ratio
0.0 0.1 17.0:1
0.1 0.2 16.0:1
0.2 0.3 15.5:1
0.3 0.4 15.0:1
0.4 0.5 14.7:1
0.5 0.6 14.6:1
0.6 0.7 14.5:1
0.7 0.8 14.2:1
0.8 0.9 13.2:1
0.9 1.0 12.5:1

These values are at 1200 dergeesF. Similar readings will be found from 800-1400 degreesF. As temps varies the O2 voltage will vary also, for example:

@ 750 degrees a V of 0.9 will indicate a A/F ratio of 12.4:1
@ 1500 degrees a V of 0.9 will indicate a A/F ratio of 14.0:1

Taken from here: http://www.bbs.legacycentral.org/viewto ... light=band

I'm with 99.9% certainty that my o2 sensor wiring is intact at the downpipe, however, I need to take a look at the ecu connections. However, wouldn't the a/f gauge still show the proper reading even if the ecu connection is screwy?

No way to disconnect the o2 sensor without disconnecting the gauge, unless I just clipped the wire from the junction point to the ecu.

[IggDawg]

Hey man, why not check your TPS just for kicks? if it's screwy, that may be your problem. the computer does indeed calculate timing and fuel delivery based on TPS inputs. TPS is VERY important. prolly more so than the o2 sensor.

just give it a check. should be a simple test with a voltmeter. if it's busted, you can probably get a new one from a junkyard for a couple bucks. it's worth a shot.

[boostjunkie]

I swapped in a new TPS sensor last week thinking that was the problem. I checked all of the voltages as described in Legacy777's diagnostic check. Even the idle switch is set to EXACT specs. The problem is the ecu doesn't throw a code anymore like it used to for the TPS and the a/f learning control (hasn't come on in almost a week). But I'll still randomly get the crazy a/f readings.

I'd need to check the TPS when the a/f readings are out of whack, but that's really hard considering that the symptoms don't manifest themselves for that long anymore.

How does voltage affect the o2 sensor readings themselves? I remember reading Legacy777's post about the 4-wrie o2 sensor and morgie's statment in the fcd post about a better ground. I've done the grounding mod and checked the voltages from the contact point to the battery power post. With the engine idling, this is what I found:

terminal at intake manifold: 13.13v
terminal at strut tower: 13.17v
terminal at stock ground by battery: 12.25v!!!

Why such a discrepancy? Now this ground is what feeds the ecu, correct? Could the ecu just not be getting enough voltage to properly run?

One more thing. When I changed the alternator last year there was a gfb pulley on the old alternator. I wasn't able to pull that off and installed the stock alternator with the stock pulley attached. From what I know of the gfb pulley, the crank pulley is smaller AS IS the alternator pulley. I'm assuming the alternator is not running as efficient as before. Why? The lights do dim at lower rpms (below 2500rpm). Do you think this could be the culprit? Maybe my battery's going out?

[Legacy777]

The gfb pulley set does have a smaller alternator pulley which helps increase the speed of the alternator. with a stock sized alternator pulley and underdriven crank pulley, your alternator will definitely not be spinning as fast. I have that issue on my car now, I'm going to be alleviating with a smaller gfb alternator pulley.

[boostjunkie]

Now how would the decrease in alternator output possibly cause the problems I'm having with a/f? And why the variance in voltages between the grounding points on the car?

[Legacy777]

Don't know......I just know that the size difference will affect alternator performance.....

[vrg3]

Wow, those are big discrepancies. Even the 40 millivolts between the strut tower and intake manifold is undesirable. What points are you using for the grounding mod, and are you sure all the terminals of the grounding mod are making good contact?

That big of a drop between the block and the battery ground is definitely hurting your charging, too, regardless of the condition of your alternator and any underdriving.

I believe the ECU is grounded through the wiring harness that's mounted underneath the intake manifold. If I'm reading these diagrams right, it's a ring terminal that's attached to a bolt going into the rear-facing side of the intake manifold near the rear driver side runner. The bolt appears to be directly below the fuel rail flange. The ring terminal disappears into the same harness that feeds the fuel injectors.

The stock oxygen sensor grounds through its body (and so through the downpipe and whatever the downpipe's connected to).

Wait a minute -- your downpipe's stainless steel, isn't it? Stainless is not a good conductor of electricity. If Adam didn't switch to a 4-wire sensor, and so you have a stock 3-wire sensor grounding through a stainless steel pipe, you should try running some wire to the sensor body (use a small hose clamp to hold it on) and grounding that wire to the intake manifold.

[boostjunkie]

The downpipe is an hks unit. I don't think it's stainless. I've heard mild, coated steel. I scraped off the paint on the strut tower contact point as well as the stock grounding point.

I can't remember right now where the intake manifold grounding point is. I'll take a look at it this afternoon and get back to you.

The grounding point that shows a REALLY high voltage drop is the replacement ground I made from the battery's negative terminal to the stock point to the right of the battery on the chassis.

[vrg3]

The downpipe is an hks unit. I don't think it's stainless. I've heard mild, coated steel. I scraped off the paint on the strut tower contact point as well as the stock grounding point.

Oh, okay. Then it's weird that you should see so much potential difference.

The grounding point that shows a REALLY high voltage drop is the replacement ground I made from the battery's negative terminal to the stock point to the right of the battery on the chassis.

You may want to run some fat wire (like 4 gauge) from one of the alternator mounting bolts to the negative battery terminal to help charge get through there.

[Legacy777]

the ground point for the injectors and wiring is on the intake manifold where vrg3 described. I took it off when I had my manifold off.

[boostjunkie]

Seeing as my lights were dimming I decided to spring for a new battery. Whereas the old battery was giving varying volages through each of the grounding points at ~12-13v the new battery is giving rock solid 14.47v at idle at every grounding point (I wasn't able to find the ecu grounding point since I didn't have your description handy.

No change in the a/f behavior, athough my lights don't dim at idle.

I'll check the ecu grounding point later. I saw the harness but couldn't tell were the actual grounding point was.

I also need to check whether the o2 sensor is actually tight in the bung. If the o2 sensor is grounded through the body then it might be a little loose (I have hit a lot of speed bumps with my exhaust with those stock wheels).