IM240 fail, No DTCs, runs rich

Randy from Denver Analyst Posted   Latest   Edited  
Question
Emissions
2007 Mazda CX-7 Grand Touring 2.3L 6-spd

So in another thread I mentioned these failing vehicles with no DTCs. I had another one of them today. 

There basically three ways they can come to a repair shop.

  1. Emission failure for CO or NOx.
  2. Poor fuel economy complaint.
  3. Readiness monitor won't complete.

This one came in for IM240 CO failure, see the file "IM240 FAILS". I then ran 4 IM240 tests and they were all repeatable (meaning the data from my runs are not posted but similar). I gathered OBD data before the tests, after 2 tests and then at the end of all 4 tests (Those are the 3 images of Excel spreadsheet with each grab reading left to right for 1,2,3). 

A "U" code set during the first drive, it was caused by me messing with a scan tool so ignore it but I left the data in for you to see. All I did was cycle key off and restart 4 times to get the MIL off after the second test.

The scan data and 5 gas pictures are in order, all this is a lot easier to see without images but that's all I can upload here at this time. (EDITED)

There are TSBs relevant to this vehicle so feel free to check but they aren't clear so apply at your own risk.

Basically, enjoy looking at the data. I'm not sure I can answer all the questions but the data answers a lot if you can understand it. Maybe you can answer some for me.

I diagnosed it, the owner will get the part I failed installed and then bring it back to me for final analysis and I will post that here if he does come back. Of course, I might be wrong. I've been right twice before so fingers crossed.

The first thing you will notice is that they run rich all the time and at any speed and never set a DTC. They are happy as can be running rich in closed loop. Wait, how can that be? Closed loop is feedback of the exhaust sensors? Right?

As a bonus, these always go into an perturbating fuel control mode at about 30 mph during an IM240 run, I posted that in the last of the images including Lambda and CO.

Check which monitors ran and which readiness do you think completed. Which one(s) won't complete and why?

(EDIT) 9-26-18 Adding more images below:

Emission Test Data - 00

Emission Test Data - 01

IM240 1st Chance Calc-load-Cat-temp

IM240 1st Chance Cmd-Eq-Purge-Lambda

IM240 1st Chance LTFT-STFT-WR-Current

IM240 1st Chance Rear02-FRP

IM240 2nd Chance EQCommanded-EQ Ratio Rear02v

IM240 2nd Chance EQCommanded-EQ Ratio VSS

IM240 2nd Chance EQCommanded-EQ Ratio WR Current

IM240 2nd Chance LTFT-STFT-VSS

+2

Rick from Saint Charles

 

Technical Support Specialist
 

Looks like a great lesson. Unable to read the data. The sheets will not enlarge. Probably an old guy problem.

0

Eric from Edgerton

 

Owner/Technician
 

This works in Firefox, not sure about the other browsers. Right click on one of the images, hover over "This Frame" and in the slide out menu click "Show only this Frame" they will open up larger and you can click on them to zoom in.

0

Scott from Claremont

 

Manager
 

Hi Rick,

I had Randy send me the scan tool data so I could able to embed the Silver scan-tool report. Let me know if you have any trouble.

0

Steven from Spokane

 

Mechanic
 

That's a tough one. I see that the vehicle is running rich as you pointed out, and it looks like the ECM is commanding it to be there if I'm reading the data correctly.

You also mention we should have a look at which monitors are not completing and why. Looks like EVAP isn't completing, although I don't have any specific reason as to why.

What part did you recommend replacing? The EVAP purge valve?

0

Scott from Claremont

 

Manager
 

Hi Steven,

Randy provided me with additional data which I added to his original post just a few minutes ago. You may want to take a look to see if there are any helpful clues.

0

Randy from Denver

   

Analyst
   

Thanks Scott. 

Steven, I recommended B1S1, The wideband sensor. I haven’t heard back yet but I’m a little bit confident. Hey, it fixed the last one I had. 🤔 I’m open to any input but this thread along with when I posted the other 2 vehicles has been pretty quiet. Probably because my initial images sucked and I should not have posted them. Thanks to Scott for helping this week.

Look at the cat monitor, mode $06 and $09 data. The shop had recommended a new cat. What do you think about that? Also look at mode $01 pid $41. Lots of interesting data there if someone wants to explore.

You are reading it correctly. Go look at Scott’s tft thread and the discussion Robby, Paul and I had and look at the pertebation swings rich. It should look familiar.

0

Steven from Spokane

 

Mechanic
 

Hello Randy,

For the cat monitor, the mode $09 data shows that the monitor hasn't been run, and the mode $06 data shows the last recorded results were at the max limit. I think what you are getting at here is that because of the fuel control issue, the cat monitor is at the limit and the ECM is no longer running the monitor because the conditions aren't quite right for it.

In other words, the other shop suspected a bad cat, when the actual issue was that the cat couldn't perform well given the fuel control issues.

As a side note, when I see mode $06 data that is at the limit, I don't necessarily call that bad, because I've seen several results like that in mode $06, so I attribute that to the way the ECM runs its tests. If you look at the Evap Purge monitor in your data, it is also at the max limit. Maybe this is an issue, but maybe this is just a result of how the ECM ran it's test. Just my thoughts.

I think I see now why you recommended the air/fuel ratio sensor. The amperage for the sensor is negative (indicating lean), and the ECM is commanding rich to try to bring it back towards zero, but it looks like it isn't responding as it should, therefore the engine is running slightly rich all the time. Interesting that it hasn't set a fault. I wonder if it would after driving for a long while.

0

Robby from New Market

 

Mobile Technician
 

Hi Steven, 

I'm having similar thoughts about the cat. Maybe, with the rich condition, there's just enough oxygen available to get a pass.

Since the actual (5-gas) lambda matches the measured lambda in the data, that makes me think the front O2 is accurate. With a wideband, if the amperage is negative, that would indicate a rich condition. In the fuel control systems I've seen, the ECM responds to a front O2 rich/lean condition through STFT and/or LTFT, not by adjusting the target AFR.

In the first run, STFT = -17.2%, LTFT = +18.8%. The common though is that would be a total FT of +1.6. In Scott's FT thread a week or so back, I shared some data that indicated that wasn't accurate fuel control. I'm wondering how the results would look if the FTs were cleared and then retested.

I'm anxious to hear the outcome of this.

0

Steven from Spokane

 

Mechanic
 

You are correct. I got the amperage positive/negative backwards.

In that case, the air/fuel sensor does appear okay, at least as far as being on the rich side, and that matches up with the exhaust gas analysis.

I agree that a reset of the fuel trims would be interesting to see what happens.

Your comments about the STFT and LTFT not truly representing the TFT is what I have experienced with the German vehicles I see here. The Germans have their own way of dealing with fuel adaptations, and it doesn't match up with the old school definitions of STFT and LTFT.

0

Randy from Denver

 

Analyst
 

Let me throw you a curve ball. I am still struggling with this vehicle and it's mostly because I don't have much information on Mazda vehicles and I do not have any contacts for Mazda's. But, you and I do see a lot of Ford's and happen to know a guy that knows everything there is to know about Ford's.:) What if this was a Ford with a Bosch UEGO (it is a Bosch UEGO).............

"fordservicecontent​.​com/ford_content/c…​.​pdf

Take a look at the section on UEGO's. Especially this on page 81:

"For "Non-Stoichiometric Closed Loop (NSCL)" air/fuel control applications, a continuous open IA diagnostics (Air Rationality Test) is required since the lambda error is more significant in this mode. The air rationality test will always monitor the UEGO sensor voltage or pumping current reading during Decel Fuel Shut Off (DFSO) event. The monitor compares the UEGO sensor voltage or pumping current reading in air against the expected value for pure air. If the UEGO sensor voltage or pumping current during DFSO exceeds the maximum UEGO voltage/pumping current in air threshold, then the fault timer increments. If the fault timer exceeds the fault time threshold, then open IA DTC P2626 and/or P2629 will set."

NSCL, that should look familiar.

"UEGO pumping current: > 0.00309 Amps for >= 2 seconds in test conditions." to set DTCs. Those DTCs aren't available on the 2007 CX-7 but....

Here's the graph:

process​.​filestackapi​.​com/resize=h:1000…

Now take a look at my vehicle during decel:

process​.​filestackapi​.​com/resize=h:1000…

0

Robby from New Market

 

Mobile Technician
 

Hi Randy, I'm late to the party - again 😒

Reading 5-gas data is out of my league, but your data does intrigue me. 

First, I've been looking at this graph:

Overall, it's commanding slightly rich mixture. I could see a front O2 being biased causing this, but is it possible the rear O2 could be biased? Did you happen to measure Lambda with your 5-gas and compare to the PID data?

Again, I have no experience reading or interpreting 5-gas data, so bear with me😊 . HC looks pretty low, but CO is high. To me, that doesn't make sense. I would except both to be high with a rich condition. Could the catalytic converter be doing this? Speaking of the catalytic converter, it looks like the cat monitor completed, but the test result is equal to the max value. However, if it's running rich, I suspect the cat monitor results might be inaccurate.

On the other hand, it the cat monitor is showing low oxygen storage, the fuel control system may be reacting like it did on that 14 Ford F550 I posted about in Scott's earlier FT post.

In your reply to Steven, you mentioned PID $41 (Monitor Status this driving Cycle). I'm trying to figure why that PID is important for me.

0

Steven from Spokane

 

Mechanic
 

Hello Robby,

Exhaust gas analysis isn't too difficult once you know what to look for. This chart shows what to look for.

As for the catalyst monitor result being exactly at the limit, I believe this is because of a "fast pass" by the ECM. It ran the test only long enough to get a pass (within the limit) and stopped the test right there and recorded the number. I'm sure Randy will correct me if I'm wrong about that.

Randy,

I will have a look at the things you pointed out later on tonight. A little too busy right now. I don't think your posts are quiet because of the pictures. I think it's because it's a lot of data and hard thinking that has to be done, and maybe not everyone is up to that, or maybe not enough time to dive into it. I think your posts are interesting, it just takes some time to come up with a good response. Thanks.

+1

Randy from Denver

 

Analyst
 

Too much data? I uploaded a fraction of what I gather, that's the real problem. I have no idea how to convey difficult concepts via the internet but I am open to suggestions.

I do not agree with your cat assessment.

The following is one I looked at today for a cat monitor (Camaro). Here's the data I gathered:

process​.​filestackapi​.​com/resize=h:1000…

And then zooming in on the monitor running:

process​.​filestackapi​.​com/resize=h:1000…

0

Randy from Denver

   

Analyst
   

This is in reply to Rob.

Lambda for the same time in the graph you posted:

process​.​filestackapi​.​com/resize=h:1000…

For entire 1st IM240:

process​.​filestackapi​.​com/resize=h:1000…

Lambda was pretty close to EQ ratio on scan data.

As for PID 41, I used it on the Camaro I just posted in this thread. Too much typing to explain all of this stuff though but 41 changed, IUPR did not (for cat) on the Mazda. The O2 pid 41 did. The mode 6 results look like cat is failing, a shop wanted to replace it. If the shop understood what was going on, including 5 gas, there's no way they would have diagnosed a cat. The mode 6 cat data did not change yet the O2 mode 6 did. 

Here's another problem with trying to explain this stuff, the Camaro is a 2010 and the Mazda is a 2007 so the IUPR and PID 41 behave differently.

0

Randy from Denver

 

Analyst
 

B1S1 was replaced, it now passes the IM240. I ran 5 IM240 tests and they range from 3.3 to 6.7 gpm CO so all are passes but not what I considered fixed. I have a secondary issue I want them to address and the shop has that repair scheduled for next week and I will see it again after that repair. Fuel trim is pretty close to the same as before, maybe slightly different. Here's a little data:

Before repair (intrusive test)

process​.​filestackapi​.​com/resize=h:1000…

After repair (intrusive test):

process​.​filestackapi​.​com/resize=h:1000…

Before repair decel:

process​.​filestackapi​.​com/resize=h:1000…

After repair decel:

process​.​filestackapi​.​com/resize=h:1000…

0

Steven from Spokane

 

Mechanic
 

Wow, interesting stuff Randy. Looks like this boiled down to a lambda error during Non-Stoichiometric Closed Loop (NSCL), and since this Mazda doesn't have fault codes for that, then maybe it doesn't monitor for that type of fault. That sure would explain the failure.

I'm still not sure why the ECM would choose to stay commanded at a slightly rich ratio though. It seems like it wouldn't continue to command slightly rich indefinitely. Maybe it is a safety default mode, or maybe it is just a bug in one of the algorithms in the ECM . Anyway, this is an interesting topic. 👍

0