Hi, all

I've been trying to track down consistently negative trim values on my '97 convertible. I often see both long-term trim values around -15 at idle, though they normally hover around between -8 and -12. I don't get any codes, but the idle is a little rough when the trims are more negative.

One thing I've noticed is that one of the O2 sensors is much more active than the other. Here are screenshots from Torque showing how 1x1 has a lot less response than 2x1. Sometimes they will be about equally spikey, but about half the time I see 1x1 being flat like this. 2x1 is always switching.

Is this a clear sign of a failing O2 sensor? Is it likely related to the negative trims? I'm doubtful because both trims are negative when only one O2 sensor is flatter.





Based on advice in these forums and TSBs, here are things I've done so far trying to bring trims back in line:

• Replace air filter
• Clean MAF sensor multiple times. Clean connector.
• Swap MAF sensor with a used part. (The early model's MAF is rarer). No changes.
• Check fuel rail pressure. It's spot on the TSB's 38 psi spec.
• Verify engine coolant temperature sensor.
• Hard reset.

I'm running out of ideas. Seems like negative fuel trims don't have that many causes.

 

It could be O2 sensors but that both sides are pretty close to each other and both negative suggests a common component.

Are you confident that the fuel pressure regulator and manifold pressure reference are working properly? You mentioned 38psi, but how is that measured? With a gauge on the fuel line schrader valve? Under what engine condition?

 

I measured 38 psi at idle using a test gauge hooked to the fuel rail Schrader valve.

Now that I'm at my computer I can include this link to the procedure I used, which was a step beyond just measuring fuel pressure at idle. I went through the steps of revving to 2000 rpm, seeing the rise above 40 psi before the drop back to 38 psi, exactly as the procedure specified. That gives me a good bit of confidence that the pressure reading was right... even though I admit it was a Harbor Freight test gauge

For anyone interested, see page 10, "P20 Fuel Pressure Regulator"
http://www.jagrepair.com/images/TSB/...Flowcharts.pdf

 

1x1, 2x1 etc are confusing - non-standard and do they mean bank 2 or sensor 2???

You seems to have LTFT1 twice but no LTFT2.

 

Intake temperature looks high, maybe because it is static but check the air in pipework,
1x1 O2 sensor is bad and on your display you have LTFT 1 twice...

Would also expect a higher voltage at idle unless you have a very flat battery!

 

Also worth noting that the upstream O2 sensor (the one that controls the fuel trims) outputs current not voltage and most standard OBD readers can't get the actual current reading from it, just the fuel trims.

The sensor that outputs a voltage is the downstream sensor that monitors the catalyst.

 

I'd say make the TorquePro screen a little more direct. Show upstream sensors long and short term trims. Also try and find the reading to the downstream sensors (from memory, there is a trim for that). Trims are a function of Load and RPM, so make sure that is on the same page as well. Do the test at idle and at some higher RPM (2k?). Check if the trims somehow come back close to zero at higher RPM where leaks would be less important. You can also try and add Load at rest by pressing on the brakes in "D"rive and put some gas. Try and take a screen shot of that too.

Negative trims can be from a bad air flow meter. Cleaning it might help, but would not bring a failed unit back to life of course. The case against your AFM is that both bank show low trims...

Other possibility is that there is extra fuel. Check the purge valve circuit (wheel well behind the liner). See if you can temporarily block the line and retest. Normally, there is a green-capped valve somewhere in the evap circuit that can be used for smoke injection to detect leaks. One would think you may not have external leaks but maybe a purge valve stuck open and always releasing fumes in the intake.

Last would be to try and determine if 1 or more O2 sensors are somehow "lazy" and replace them. I would check that the short term trims bounce around some as an indication. Long term trims only update every 2 min AFAIK.

 

Oh! Right you are about having two of the same LTFT. That was a picture from a month ago, and I forgot that I'd made that mistake on that one.

I went out and took new captures today. This was after driving about five miles to warm things up. I was sitting idle for about three minutes before taking this screen grab, just long enough for fuel trims to stabilize, and for the O2 sensors to start showing different levels of activity.

Yeah, I noticed intake temperature looked a bit high, but when I open the hood and felt how hot the airbox was, it seemed reasonable. For this capture the car had been moving.

In the Torque terminology, 1x1/2x1 are the upstream O2 sensors while 1x2/2x2 are downstream.

Yeah, I'd noticed the voltage being a bit lower than expected. The battery's new, but the alternator is a little old, so I figure it might be on its way out.

fmertz, as I recall the trims do go to zero at higher rpm.

 

Ok if you have -ve trims at low/idle and rough running but normal at higher RPM then you have an air leak after the MAF introducing unmetered air,.
Also your LHS O2 (2 x 1) is dead!

Pretty sure they are narrowband sensors on a 97 so the voltage is measured directly.

 

I thought unmetered air always caused +trims as the car introduces more fuel to match the extra O2 in exhaust.

Yeah, I was wondering if this graph would be enough to say flat out whether that sensor was dead or not. I read so much in here about the O2 sensors being carefully monitored, as if a dead sensor would definitely result in a check engine light, but I don't even have pending codes.

 

This was an interesting suggestion especially since I did work on my purge system earlier this summer. Fortunately for these early cars the purge valve is accessible in the engine compartment (though the vent valve is still up in the fender). I pulled the pipe from valve to manifold and capped it, but fuel trims were still negative. As I recall, I pulled vacuum on the output of the valve before I installed it, so it looks like it's still not leaking.

 

Doh! Of course it does - that'll teach me to post way past bedtime
O2 sensor still dead though

 

Not seeing LTFT2, but if both banks have fairly similar -ve trims you would likely be looking at an item common to both banks so could easily be looking at the MAF.

The PCM has to mostly trust a few things, MAF included, then can tweak depending on the burnt outcome as seen by the O2s.

 

Ha, thanks. Well that's mainly what I wanted/expected to hear, that the graph on the 1x1 O2 sensor shows that it's definitively dead. At least it's something I can move forward with.

I agree that rich on both banks indicates something shared, but I've already swapped the MAF sensor with no change, so I don't want to blindly throw another at it. Here's hoping replacing that upstream O2 sensor somehow improves both banks.

 

Sounds worth trying.

 

I was looking at the ECU strategy for a different problem for someone else and I noticed that if the intake air sensor goes bad the ECU substitutes ... 122F , which reminded me of this thread (and my earlier comment about it) - because that would make the engine run rich at idle - so you would get -ve trims and things would normalise as the engine and intake gets hotter.

Also check the plug before you get a new O2 sensor - the Delphi did not fit my 97 Coupe it was a Walker - there is a different plug AND the heater resistance is significantly different.

 

Thanks for checking back. In my case the trims don't change when the engine is hot.

I've also mostly convinced myself that the air intake temperature is pretty close to what I think the temperature is in the airbox.

At this point I bought the new sensor, which wasn't particularly expensive from Rock Auto, and I'm going to install it in a few days, once the car has passed the routine DOT inspection. I don't have a check engine light now, and I want to avoid anything that could trigger one right before it's due!

 

No emissions testing for the DOT then...
Easy way to check the IATS - just turn on the ign without starting and see what it says!

 

Yep. I'm happy to not live in a city that requires emission testing.

Country roads for driving are a bonus.

I had to replace the swaybar links over the weekend for it to pass inspection, so the reinspection tomorrow will be fine so long as no check engine light shows up before then.

 

Quick update: I replaced the upstream O2 sensor that wasn't showing as much activity, and so far the new sensor looks significantly more active, but there is no change in the negative trims. LTFTs are still between -10 and -15.

The next steps seem to be pure shot in the dark territory. Around the internet I see different ideas about how downstream O2 sensors are supposed to behave, whether they're supposed to cycle at idle for example, and about what different cars might do in response to downstream sensor readings, so that's one shot in the dark I haven't worked on yet.

Or maybe I should run a round of Techron through the tank. People in these forums have said good things. Or maybe the MAF sensor connector itself is a problem, even though I really think the MAF and Intake Air Temp values look about right.