Hunting down lean condition P0139 Honda Fit 07

Joined
Aug 6, 2010
Messages
5,835
Location
Atlanta, GA
Application:

2007 Honda Fit Sport w/ A/T 292k Miles

History: 2 years ago replaced a failed aftermarket CAT

PO had stolen maybe? It was a cut out replacement rather than a Direct Fit so the factory CAT was gone when I bought it at 156k.

So, it imploded due to a misfire event at interstate speed and not stopping soon enough to address the blinking CEL. I suppose enough unburned fuel took it and the Upstream A/F sensor out. Both generated CEL codes for each respectively.

Fix 2 years ago:

1) A shop welded in a Universal MangaFlow Fed Emissions CAT; no codes since

2) I replaced Upstream A/F sensor w/ OE Denso; no codes since

Since then:

The ECU also generated an intermittent code P0139 for the Downstream 02.

I thought it was also taken out by the misfire event and CAT implosion, so it was replaced with an OE Denso. Only $50 bucks and my own labor.

Of course, the problem persisted. :ROFLMAO:

I plugged in a cheap OBD-II scanner; which helped me diagnose the A/F sensor failing.

Then, it was pegged excessively lean almost all the time; like 46% additional fuel being added so it was obviously fouled. New one fixed that problem.

For this current issue, Scanner shows short-term and long-term fuel trims.

The short-term at idle fluctuates some.

During light to moderate acceleration, I get 4% to over 10% additional fuel being dumped in by the computer.

During heavy or WOT, the short-term goes to 0.

I’ve tried resetting the computer by leaving the battery disconnected and turning on lights, etc.

Question: Could an exhaust leak with the welded in CAT or an intake leak be causing the computer to think the mixture is lean?:unsure:

I’ve tried listening to the fuel injectors and running cleaners through the gas tank, and they also all sound the same through a long screwdriver. The performance is fine under WOT.

I suppose it could be a problem with injectors, but I suspect the problem goes away when the engine is overcoming any air leaks at WOT, but they persist at light to moderate acceleration when vacuum is higher.

Just trying to get a sense before taking it into the shop. It’s my only vehicle.

Thanks!
 
Last edited:
Try a new OEM gas cap?
Pay attention if the code happen after filling up the fuel?

You said, you have not replaced the downstream sensor, may want to try that first.
 
Try a new OEM gas cap?
Pay attention if the code happen after filling up the fuel?

You said, you have not replaced the downstream sensor, may want to try that first.

So, let me clarify that the Downstream has been replaced, but the code persists.

I have a new Honda gas cap as well, installed at some point after the code began.

What I “think” is happening is this short-term fuel trim on light to moderate throttle is related to the P0139.

I may be completely wrong and they are unrelated.

The code triggers because the computer expects a high oxygen / lean condition during Deceleration Fuel Cut Off; which is represented by a reading of .2 Volts or lower within 7 seconds by the Downstream 02.

I “think” because the computer is adding +10% or more fuel during light to moderate throttle, that it’s creating a situation of unburned fuel in the exhaust, causing the code even when DFCO kicks in.
 
I am at a similar prickly w/ using Denso sensors. I been told that using none OEM sensor in some cases it does not work.

For instance, I have a 2004 Honda CRV in which calls for an AFR/ up stream sensor 2394005. This is suppose to fit different Honda cars according to Rack Auto.

If yo look at Honda's OEM website these cars do not share the same part number or price. There has to be some difference that the cars computer is looking for a specific signal or pattern.
 

Attachments

  • DENSO-IMG_9361.jpg
    DENSO-IMG_9361.jpg
    245.8 KB · Views: 20
I've heard of air leaks with weld-in universal cats before :unsure:


I am at a similar prickly w/ using Denso sensors. I been told that using none OEM sensor in some cases it does not work.

For instance, I have a 2004 Honda CRV in which calls for an AFR/ up stream sensor 2394005. This is suppose to fit different Honda cars according to Rack Auto.

If yo look at Honda's OEM website these cars do not share the same part number or price. There has to be some difference that the cars computer is looking for a specific signal or pattern.

For your car, are they different depending on federal or California emissions? The OE supplier is probably Denso or NTK anyway. Maybe one part# is the federal sensor, and the other the Cali sensor.
 
I've heard of air leaks with weld-in universal cats before :unsure:




For your car, are they different depending on federal or California emissions? The OE supplier is probably Denso or NTK anyway. Maybe one part# is the federal sensor, and the other the Cali sensor.
I agree w/ your statement on Toyota cars.

Honda's another ball game, if you look at attachment from Rack Auto it claims to the fit cars Civics and CRV.

I took an OEM sensor from the a Civic into the the CRV and got codes w/ it.. The Civic had no codes at all.

The other difference is that the upfront sensors are Air Fuel Ratio sensor and they do not work like old fashion oxygen sensors.

The range 0.4mv difference.
 
I agree w/ your statement on Toyota cars.

Honda's another ball game, if you look at attachment from Rack Auto it claims to the fit cars Civics and CRV.

I took an OEM sensor from the a Civic into the the CRV and got codes w/ it.. The Civic had no codes at all.

The other difference is that the upfront sensors are Air Fuel Ratio sensor and they do not work like old fashion oxygen sensors.

The range 0.4mv difference.
For the Honda Fit, I used my BIL’s car as a test donor (same year/model) and swapped AF sensors (his was working).

Those codes and behaviors associated with the Upstream AF sensor went away; I replaced the bad one with a Denso and it behaves the same as the working OE from my BIL’s… thread here:


I think the current problem(s) could be with a leaky Injector or unmetered air causing the short-term fuel trim to be elevated during light to moderate throttle.

Now, is that related to this P0139 that is now present or are they separate? If the new cat wasn’t doing its job, it would throw a different code, as far as I know.

Posting pics later.
 
I agree w/ your statement on Toyota cars.

Honda's another ball game, if you look at attachment from Rack Auto it claims to the fit cars Civics and CRV.

I took an OEM sensor from the a Civic into the the CRV and got codes w/ it.. The Civic had no codes at all.

The other difference is that the upfront sensors are Air Fuel Ratio sensor and they do not work like old fashion oxygen sensors.

The range 0.4mv difference.

What kind of Civic is it? :unsure:

There are many different Civics of that gen 01-05. The Si with the K20, the HX lean burn, the hybrid, and of course base DX/LX/EX D17 with or without VTEC. One of those, and possibly only one, may in fact use that AFR sensor, just not your version. OTOH, your CR-V only came with the K24.
 
For the Honda Fit, I used my BIL’s car as a test donor (same year/model) and swapped AF sensors (his was working).


This study makes sense, same cars and model years. When switching from different models w/ different engine sizes is where a problem can come from.
 
What kind of Civic is it? :unsure:

There are many different Civics of that gen 01-05. The Si with the K20, the HX lean burn, the hybrid, and of course base DX/LX/EX D17 with or without VTEC. One of those, and possibly only one, may in fact use that AFR sensor, just not your version. OTOH, your CR-V only came with the K24.
Yep. Same sort of situation on my old 92 Civic VX. It was the only one of that gen with the lean burn and had a 5 wire wide band. I replaced the old OE with a Standard Motor Products 5wire for the car. Came with Made in Japan on box. Code went away.
 
@Luftmasse Here are the pictures while in park, no accessories or engine fan running but at full operating temp (including oil, car just back from commute)

Some are at idle. Some at 3k RPM. A couple after letting car come back to idle, but before the fuel cut off readings go back to showing fuel adding.

I normally don’t notice the short-term elevated at idle. These aren’t the worst readings either. Normally when driving I’ll get as high as 15% fuel adding.
1280ED03-59B7-4949-8A76-3DD96CA1A6C4.jpeg
3C03B6B8-A478-4E6C-AB65-ADE2198B99DC.jpeg
0B287E04-267B-4140-8126-851FF9FCA1DF.jpeg
854CD93E-778E-400D-AFAD-FFB7B0EF7DFC.jpeg
C2D49479-3BF3-4D23-BB04-032C0277A650.jpeg
80734806-364A-49F8-8E47-84D41C587FEA.jpeg
AAC68FB7-24F3-4D57-91A6-AEFC0DFB9CCA.jpeg
 
Does the downstream o2 sensor location match the location on the factory cat?
That’s a good question. I don’t recall actually.

It was an aftermarket CAT that failed; at least according to the mechanic. There is a section of exhaust near the CAT which was replaced; perhaps due to rust or from the OE being stolen.

I think both were mid bed. But I don’t recall if the the previous one had an angled port.

Maybe you are on the right track.
 
Here are the readings immediately after letting off the throttle from 3k RPM. Downstream 02 correctly reads high oxygen content as computer subtracts fuel-

At idle, it jumps all over the place as short-term fuel bounces around adding various amounts of fuel.
 

Attachments

  • 691C9393-0879-47B4-ACB3-2533B58E0307.jpeg
    691C9393-0879-47B4-ACB3-2533B58E0307.jpeg
    77.1 KB · Views: 20
@Luftmasse Here are the pictures while in park, no accessories or engine fan running but at full operating temp (including oil, car just back from commute)

Some are at idle. Some at 3k RPM. A couple after letting car come back to idle, but before the fuel cut off readings go back to showing fuel adding.

I normally don’t notice the short-term elevated at idle. These aren’t the worst readings either. Normally when driving I’ll get as high as 15% fuel adding.View attachment 100186View attachment 100187View attachment 100188View attachment 100189View attachment 100190View attachment 100191View attachment 100192

Don't see anything wrong with the lean fuel trims you were talking about.

Does not look like you have an exhaust leak around the rear oxygen sensor either.

On some vehicle, it is important to install the rear oxygen sensor in it's original location. You can't just install it anywhere. Can you take a pic of the rear o2 sensor?

Simulate a rich condition by stomping on the gas pedal and does the rear o2 sensor immediately shows more than 800 mv? After letting throttle does it goes down to under 200 mv?
 
Last edited:
@Luftmasse my reasoning was perhaps a leak before the Upstream A/F ratio sensor, causing a rich condition by thinking it needs to add fuel because the computer sees lean due to additional oxygen. That was the theory anyway. In that scenario, would the rear 02 throw a P0139 because of poor running of the engine? Based on my general reading and limited understanding, a bad fuel injector or something with throttle can stop the car from entering DFCO.

Would a poorly positioned rear 02 cause the readings of 10-15% additional fuel requested by the computer on light to moderate throttle? I thought those readings would be determined by the Upstream A/F ratio sensor. Or is that pointing to a fuel delivery issue and a separate issue from the P0139?

Getting more photos.
 
Here is the cat, rear 02 and pipe upstream. It looks to be the same location of the OE CAT, which had a bend and pipe with flange above the catalyst; which has been deleted as seen by the 2 welds Upstream of the CAT.
 

Attachments

  • 07CC3866-CE6E-47DF-90EF-EF5B797B446C.jpeg
    07CC3866-CE6E-47DF-90EF-EF5B797B446C.jpeg
    143.1 KB · Views: 32
  • 15602076-9C48-4B3B-8D17-CCCAAC29F3E1.jpeg
    15602076-9C48-4B3B-8D17-CCCAAC29F3E1.jpeg
    137 KB · Views: 31
Back
Top