Honda Insight Forum banner

1 - 20 of 40 Posts

·
Registered
Joined
·
591 Posts
Discussion Starter #1
In my quest for improved mileage I am pursuing many things. One of these is the A/F ratio sensor. It is very difficult to weed out the good information from the garbage. Since I have a spare sensor, I decided to swap the old one tonight. But when I pulled it, I found it had white deposits.

Kazam_screenshot_00000.png

The scratches are likely from when I pulled the sensor out of the hole.

Those deposits, I have read, can be from fuel additives (which I have used), an antifreeze leak (of which I'm not aware), and silicon (of which there has been some use around the car, in the form of silicon spray to deal with window track, plastic, etc.).

Note that in the holes the white deposit is also visible, which is Not A Good Sign. In one hole, however, it looks like the deposits have flaked off a little. I did not put anything into the hole; more on that next post.

I used an exacto knife and found that the deposits could be scraped off:

Kazam_screenshot_00001.png

Note, below the scratched area, is another area that looks like a piece of deposit that has cracked off in the past. More on that in the next post...
 

·
Registered
Joined
·
591 Posts
Discussion Starter #2 (Edited)
Air Fuel Ratio Fun (part 2)

I decided to experiment on this. You will find people talk about getting these red hot and then dunking them in water to break off soot. You will also found technical literature describing how these sensors should be placed so that moisture/condensation from a cold start does not splash into the substrate and cause it to crack from heat stress.

I decided instead to try heating it with a blow torch. This was the result:

Kazam_screenshot_00002.png

It looked at first like the torch may have helped remove the deposits. (Note, inside the hole, there is a flake missing from the deposit inside. I don't know if that was there before I heated the sensor. It is possible that it was.)

So I decided to take the knife to the side where the material really changed:

Kazam_screenshot_00003.png

What I did here was rolled the edge of the blade along the outer cylinder with a good amount of pressure. The deposits flaked off! What that suggests to me is either that the heat melted the material and it formed a glaze which the focused pressure of the blade edge cracked off, or the sensor was designed with a protective glaze which I cracked off.

Silicon == glass. So maybe this really is silicon contamination.

Finally, I tried rolling the knife on another side of the sensor. This time, a large piece of the deposit flaked off:

Kazam_screenshot_00004.png

The area where it flaked off is shiny. (You can see where the old flake was from the previous post. On the left is where I scraped off deposits before I heated it with the propane torch.)

The question now is, was this caused by a one-time event (some bad additive in gasoline or a bad selection of additives on my part, or silicon spray used in the garage, or similar) or do I have a small coolant leak from the exhaust manifold? Or did I at one point due to a temperature-induced leak?

So I'm not going to put the new sensor in yet (unless my torch test broke this one) but instead replace this one and check the newly clean areas in a few thousand miles to see if the deposits have returned.

If my air/fuel sensor readings are broken tomorrow, I'll know I broke this and will put in the other sensor. Otherwise, we'll see if the heat made any positive or negative difference. I've been able to get 60 MPG on my last two commutes without too much effort, so I kind of have a baseline to compare to.

Any thoughts or experience on this is welcome.
 

·
Registered
Joined
·
591 Posts
Discussion Starter #3 (Edited)
<EDIT>The new plug's brand is redacted until I can confirm that I did not simply receive an improperly packaged plug.</EDIT>

The 2004 Insight did not get along with the new <BRAND REDACTED> sensor once the introductory 10-15 miles was over!

Today after 130 miles on the "cleaned" Denso AF sensor (above) (driven to confirm that the engine would not contaminate another sensor), I swapped in a new <BRAND REDACTED>.

The car runs fine for the first 10 or so miles after pulling the battery ground with this new sensor. But after that, if you are moving on flat ground and get it into lean burn, then slowly add throttle, it will actually gradually reduce power and drop in speed. You can kick it out of this mode with firm acceleration. The car eventually throws code P2A00 "O2 Sensor Circuit Range/Performance", aka "A/F Sensor B1 S1 Performance".

More interesting: if you leave your foot in the same position while traveling on flat ground, it will eventually (and suddenly) drop out of lean burn and immediately start accelerating. If still on flat ground and you still have not moved your foot, it will eventually drop back into lean burn and start losing power again. I tried this around 25-35 MPH tonight in 3rd gear and the length of each cycle was about 5 seconds.

Here are pictures of the old Denso (left) and the new <BRAND REDACTED> (right). Note that the holes in the <BRAND REDACTED> are much smaller.

Denso and Bosch.jpg

Some background: The 2000 and early 01 cars use a five-wire AF sensor and the late 01, I think, and 02-06 Insights use a four-wire AF sensor. (This also implies that the ECUs are not exchangeable between these model years for a number of reasons, but I digress.) Mine is a 2004 with a four wire sensor.

It would be easy to blame the <BRAND REDACTED> sensor for the problem and the fault. But it would be better to understand why the problem is occurring in the event that there is a more subtle problem lurking, like a vacuum leak. Or, perhaps I have not properly initiated a relearn process for this sensor. After all, the car does run great and seems to get great mileage for some distance after power is cycled by disconnecting and reconnecting the battery.

Others with similar issues reported in the forum:

@oefenamb describes a similar problem:

https://www.insightcentral.net/forums/modifications-technical-issues/117913-leanburn-acting-weird.html#post1364513

which @KLR3CYL diagnosed as caused by an old (119K mile) NTK AF sensor, fixed by a different NTK sensor.

@Lt.Knuckinfutz posted that a new Bosch 15906 was "very jerky and ... could not maintain speed" in his 01 5-speed. It was fixed with an NTK 24301.

https://www.insightcentral.net/forums/problems-troubleshooting/20548-will-these-cheap-o2-sensors-work.html#post1091610

@evox reports a (presumably new) Bosch (presumably AF) sensor gave a "jerky lean burn" on a 2000 that was resolved by an NTK sensor:

https://www.insightcentral.net/forums/honda-insight-forum-1st-gen-discussion/24249-what-did-you-do-your-g1-insight-today-72.html#post1385201

@S Keith fixed an 06 HCH throwing a P2A00 code with a new Denso 234-9034:

https://www.insightcentral.net/forums/other-honda-hybrids-discussion/36482-06-civic-chugging.html#post653290

Similar issues, unresolved:

@turbogizzmo reports the P2A00 code on a 2004 whose O2 sensor was wired in by hand. The thread is unresolved:

https://www.insightcentral.net/forums/problems-troubleshooting/120595-p2a00-p0171-o2-sensor.html#post1395809

@Natalya reports "lean burn is weak (even at the lower MPG range of LB) and the car inevitably slows down when in lean burn unless on like a flat or downhill."

https://www.insightcentral.net/forums/modifications-technical-issues/117913-leanburn-acting-weird.html#post1365665
 

·
Super Moderator
Joined
·
4,692 Posts
More interesting: if you leave your foot in the same position while traveling on flat ground, it will eventually (and suddenly) drop out of lean burn and immediately start accelerating. If still on flat ground and you still have not moved your foot, it will eventually drop back into lean burn and start losing power again. [/url]
This is a NOX purge and is normal.

Sam
 

·
Registered
Joined
·
138 Posts
The 2004 Insight did not get along with the new Bosch 15394 sensor once the introductory 10-15 miles was over!

Today after 130 miles on the "cleaned" Denso AF sensor (above) (driven to confirm that the engine would not contaminate another sensor), I swapped in a new Bosch 15394.

The car runs fine for the first 10 or so miles after pulling the battery ground with this new sensor. But after that, if you are moving on flat ground and get it into lean burn, then slowly add throttle, it will actually gradually reduce power and drop in speed. You can kick it out of this mode with firm acceleration. The car eventually throws code P2A00 "O2 Sensor Circuit Range/Performance", aka "A/F Sensor B1 S1 Performance".

More interesting: if you leave your foot in the same position while traveling on flat ground, it will eventually (and suddenly) drop out of lean burn and immediately start accelerating. If still on flat ground and you still have not moved your foot, it will eventually drop back into lean burn and start losing power again. I tried this around 25-35 MPH tonight in 3rd gear and the length of each cycle was about 5 seconds.

Here are pictures of the old Denso (left) and the new Bosch (right). Note that the holes in the Bosch are much smaller.

View attachment 78157

Some background: The 2000 and early 01 cars use a five-wire AF sensor and the late 01, I think, and 02-06 Insights use a four-wire AF sensor. (This also implies that the ECUs are not exchangeable between these model years for a number of reasons, but I digress.) Mine is a 2004 with a four wire sensor.

It would be easy to blame the Bosch sensor for the problem and the fault. But it would be better to understand why the problem is occurring in the event that there is a more subtle problem lurking, like a vacuum leak. Or, perhaps I have not properly initiated a relearn process for this sensor. After all, the car does run great and seems to get great mileage for some distance after power is cycled by disconnecting and reconnecting the battery.

Others with similar issues reported in the forum:

@oefenamb describes a similar problem:

https://www.insightcentral.net/forums/modifications-technical-issues/117913-leanburn-acting-weird.html#post1364513

which @KLR3CYL diagnosed as caused by an old (119K mile) NTK AF sensor, fixed by a different NTK sensor.

@Lt.Knuckinfutz posted that a new Bosch 15906 was "very jerky and ... could not maintain speed" in his 01 5-speed. It was fixed with an NTK 24301.

https://www.insightcentral.net/forums/problems-troubleshooting/20548-will-these-cheap-o2-sensors-work.html#post1091610

@evox reports a (presumably new) Bosch (presumably AF) sensor gave a "jerky lean burn" on a 2000 that was resolved by an NTK sensor:

https://www.insightcentral.net/forums/honda-insight-forum-1st-gen-discussion/24249-what-did-you-do-your-g1-insight-today-72.html#post1385201

@S Keith fixed an 06 HCH throwing a P2A00 code with a new Denso 234-9034:

https://www.insightcentral.net/forums/other-honda-hybrids-discussion/36482-06-civic-chugging.html#post653290

Similar issues, unresolved:

@turbogizzmo reports the P2A00 code on a 2004 whose O2 sensor was wired in by hand. The thread is unresolved:

https://www.insightcentral.net/forums/problems-troubleshooting/120595-p2a00-p0171-o2-sensor.html#post1395809

@Natalya reports "lean burn is weak (even at the lower MPG range of LB) and the car inevitably slows down when in lean burn unless on like a flat or downhill."

https://www.insightcentral.net/forums/modifications-technical-issues/117913-leanburn-acting-weird.html#post1365665
My hacked wired O2 sensor was wired wrong. (well wrong as in it didnt work but appeared "right" to what the instructions want you to do) It was only after trial and error that I finally got it all working. I thought i updated the thread i started with the solution but perhaps not. I am searching to see if it put it in the Facebook group or some other thread.

FOUND IT: Here was the other thread where I explain a little more: https://www.insightcentral.net/forums/problems-troubleshooting/121021-o2-sensor-two-black-wires.html

Also because I assumed right off the bat before troubleshooting that it was the O2 sensor I had already ordered one and put it in. http://a.co/jbyyGNV Walker Products 250-54005 4-Wire Air/Fuel Ratio Sensor
 

·
Registered
Joined
·
591 Posts
Discussion Starter #6 (Edited)
This is a NOX purge and is normal.
Sam
I think you are referring to the drop from a lean mixture to stoichiometric that lasts only for a couple of seconds, and then returns back to a lean mixture?

What I was describing was different. The secondary observation was that the car was accelerating into and decelerating out of lean burn every five seconds with my foot held in the same position and the road grade pretty flat, in third gear, around 30 MPH plus or minus.

I thought that the periodicity of the NOX purge was much larger than 10 seconds, or can the NOX purge be triggered in other ways?

The primary observation is that if you are in lean burn, and moving at a constant velocity, and the road is fairly flat, and THEN you gently press the accelerator, the MPG falls; then, instead of feeling a weak acceleration you feel deceleration and the car slows down. The MPG rises even as you press harder (gradually, and not too much) on the pedal and the car continues to slow down as if you've lost power. If you are watching the injector pulse width you will observe it drop gradually. Eventually the car reaches one of the boundaries of the lean burn... map?... and returns to stoichiometric burn. If you have been traveling flat, this is a noticeable, sudden transition and the car begins to accelerate.

The condition described by @oefenamb and which @KLR3CYL diagnosed as being caused by the LAF sensor sounds like what I'm experiencing.
 

·
Premium Member
Joined
·
422 Posts
I have also observed this "pulsing lean burn" behavior and can generally reproduce it, but under a fairly limited set of conditions which include:

-OEM primary O2 sensor
-One particular flat stretch of road about six miles from where I start the drive home after work. This section is below a two mile downhill section where the car has been in FCO.
-Fifth gear
-A speed of about 50 to 55 MPH
-The air temperature around or below freezing. I do not monitor Iat.
-Ect of about 70-75 C (158 - 167F)

I monitor Tpv (TPS voltage) upon entering lean burn and then keep it relatively constant (0.74 to 0.8 volts) upon entering lean burn and the car will then go into and out of lean burn in about 4 second pulses. I have yet to capture this behavior via data logging and likely will not this season as it is starting to warm up lately.

Other than this pulsing behavior, which is common, but not every single time, under the conditions described above, I can achieve lean burn under all other conditions where one would expect it with this car.
 

·
Registered
Joined
·
591 Posts
Discussion Starter #8 (Edited)
I suspected that not performing the idle learn procedure after pulling the battery was the cause of the strange behavior of my new (and then old) LAF sensor, but another drive after properly performing the procedure did not remediate the problem. I still think that there may be value to performing the relearn after removing power from the ECU.

I found that it took over 30 minutes to do this on a cold engine. After reconnecting the battery, I let it idle until the coolant fan turned on twice, then let it idle another 10 minutes. It took over 20 minutes on a 70 degree day for the fan to turn on the first time.
 

·
Registered
Joined
·
591 Posts
Discussion Starter #9
Target lean burn ratios vs. LAF sensor response, for my old and new sensors:

two-densos-annotated.jpg

This leaves more questions than answers. How does the ECU determine what the sensor response should be at the stoichiometric mixture - is this a hard-coded value that all sensors should be? (Does the resistor in the cable of the 00-01 5-wire sensor calibrate the sensorr for this?) How does the ECU know what the value should be in lean burn? Is this also fixed, or is it learned?

It's pretty clear that my new sensor, which appears to be a Denso intended for an 00-02 Civic, and this ECU (2004) do not get along. There is so much I would like to know about how this system works, but that knowledge appears to be locked up in ECU code.
 

·
Registered
Joined
·
591 Posts
Discussion Starter #10
References:

This post describes 5-wire vs 4-wire sensors and is the first place that I found that shows the calibration resistor for the 5-wire sensor in a schematic.

https://www.scannerdanner.com/forum/diagnostic-tools-and-techniques/327-air-fuel-ratio-sensor-testing-thread.html

This article describes the operation of 4-wire sensor operation. This is of interest to late 01 to 06 owners. It refers to Toyota sensors; the previous article suggests little difference between Honda and Toyota four wire sensor tech.

http://www.partinfo.co.uk/files/4 Wire Wide Band Oxygen Sensors.pdf
 

·
Premium Member
Joined
·
9,683 Posts
Why aren't you using the NTK sensor?
 

·
Registered
Joined
·
591 Posts
Discussion Starter #12 (Edited)
Why aren't you using the NTK sensor?
EDIT: Pulled a Denso from a wrecked 2006. There was nothing to suggest it was newer than the rest of the car. /EDIT

Because I'm just now learning how different sensors matter (and why), and I'm also not certain that NTK was the supplier for the late 01-2006 models.

Is NTK the supplier for Honda for 4-wire sensor used in the late 01-06 models? Mine has a Denso in it. The previous owner had the car serviced at the dealership. I don't know if the old sensor is the original sensor. If not, and if replaced by a Honda dealer, why would they use anything other than a Honda part?

So I'm not sure if NTK or Denso (or both) were Honda supplies for the 2004 Honda. Or if Honda switched suppliers or both were suppliers. Or?

The sensor for the 00-late 01 and late 01-06 cars are engineered fairly differently, so it would not surprise me if the late 01-06 supplier to Honda is different.

What I'd like to know is what brand (and, ideally, part number) sensor is running in @jime 's 2006, or, for that matter, any other late 2001-2006 owner who is getting great mileage. That could eliminate the sensor from this equation.
 

·
Registered
Joined
·
591 Posts
Discussion Starter #13 (Edited)
I pulled a Denso LAF sensor from a wrecked 2006 yesterday. The part number on the side is "DU2 DENSO 192400-1061 05H29" and stamped 103 around the collar. There was no indication that it was newer than the rest of the car. 0414191047_1555253679863.jpg 0414191001a_1555251959890.jpg
 

·
Registered
Joined
·
591 Posts
Discussion Starter #14
More information.

RockAuto indicates for many oxygen sensors whether they are for the 5 speed or the CVT. But the Bosch 15394, which was found to cause the engine to lag instead of accelerate in lean burn (above posts) doesn't have this distinction. It also appears on Amazon as compatible for a 5 speed, which is where I got mine.

I got a CVT recently and pulled the LAF sensor to look for contamination (it was clean, and new). Surprise, it looked just like the Bosch 15394. (It had no manufacturer marking but resembled the Denso which the Bosch actually was.)

Then I read that the CVT doesn't do lean burn like the 5 speed, to permit an SULEV rating. Thus the LAF sensor would not need so wide a response.

Looking again at RockAuto's site, the information about the Bosch 15394 mentions a 15395, the grainy photo of which has the same hole pattern as the proper Denso sensor for this model year of MT Insight.

So, I now believe the Bosch 15394 is really intended for the 2002+ CVT.

Someone putting a sensor intended for the CVT into an MT may have the issues I saw.

Finally finally, I installed a new Denso LAF sensor and my mileage and power increased noticeably over the old contaminated Denso sensors, perhaps a 5 MPG improvement.

The lessons learned, I think, are these:

- old LAF sensors get contaminated and while they may not through a code, they will decrease performance in lean burn and reduce mileage.
- the general "use the OEM sensor" advice found in this forum may be due a sensor with a range suitable for a CVT being installed in a MT. This needs to be researched, however.

And, replacing the LAF sensor should be one of the last things one does when restoring a car to its original mileage:

- replacing a LAF sensor in an engine which has other issues or in which non-O2-sensor-safe materials/fluids are causing contamination can lead to the demise of the new sensor.
- other issues can contribute more significantly to reduced mileage (ie, bad EGR valve, improper valve lash) and may be the true cause.
 

·
Registered
Joined
·
121 Posts
2006 MT
I've suspected my vehicle has sensor issues despite no DTCs. One symptoms is it has more power when cold. Actually runs very good when cold. I believe this is because it is running in open loop mode and hence, ignoring the sensors. On rare occasions while driving at highway speeds, it will gradually loose power despite depressing the accelerator further and further. About the time I think its going to stop, the power returns. This has happened several times in the year and a half I've owned the car. It also has been known to occasionally "jerk" while cruising at constant speed. I've done all the normal things one does to these cars such as valve lash adjustment, cleaning EGR path and replacing EGR valve, ect. Maybe I need to pull my AF sensors. How many are on an MT?
 

·
Premium Member
Joined
·
9,683 Posts
Sounds like you are in the "Maximum" vac. for lean burn and then going into the "Purge" cycle. That is all normal. If you had a monitor for vacuum you would see it readable. Do you have something to monitor with, such as "Peters Gauge"? You are right on the loop mode.
 

·
Registered
Joined
·
121 Posts
Sounds like you are in the "Maximum" vac. for lean burn and then going into the "Purge" cycle. That is all normal. If you had a monitor for vacuum you would see it readable. Do you have something to monitor with, such as "Peters Gauge"? You are right on the loop mode.
I thought it was being fuel starved. I had no idea this is normal. What exactly is being purged and why? If you are you referring to Peter's OBDIIC&C device then no. I'd like to have one but its difficult to justify the expense.
 

·
Premium Member
Joined
·
9,683 Posts
Gone is the encylopedia for the G1 per the new forum setup. That would explain a lot. I suggest you search "Lean Burn" at this time and see what you come up with.,
 

·
Registered
Joined
·
591 Posts
Discussion Starter #19
2006 MT
I've suspected my vehicle has sensor issues despite no DTCs. One symptoms is it has more power when cold. Actually runs very good when cold. I believe this is because it is running in open loop mode and hence, ignoring the sensors. On rare occasions while driving at highway speeds, it will gradually loose power despite depressing the accelerator further and further. About the time I think its going to stop, the power returns. This has happened several times in the year and a half I've owned the car. It also has been known to occasionally "jerk" while cruising at constant speed. I've done all the normal things one does to these cars such as valve lash adjustment, cleaning EGR path and replacing EGR valve, ect. Maybe I need to pull my AF sensors. How many are on an MT?
Of the three oxygen sensors in the car, only the upstream sensor is a wideband air fuel sensor aka LAF sensor.

I put the LAF sensor off until last because of its cost and because mine showed signs of contamination. I don't know a good way to confirm or eliminate a subtle contamination problem if in fact what I observed was not simply normal after 150K miles. I also did not want to find that my mileage problem was elsewhere. After finding small improvements in some places and eliminating other potential issues I decided to finally try the new sensor.

I am now running a new Denso LAF sensor. I am very happy with the improvement in mileage it brought. I still, however, feel a slight stumbling when at the edge of lean burn, when trying to get the most power out of it. So there is more to do.

One thing you could do is pull the LAF sensor, write down the numbers on it, and post a picture. I can tell you which of mine it matches up with, if any. The holes in my CVT sensor are smaller and fewer than my MT sensor. Dont try to clean it; try to get an accurate representation of its color and deposits by taking it under different light so that ant courseness or color nuance stands out.
 

·
Premium Member
Joined
·
434 Posts
I don't know a good way to confirm or eliminate a subtle contamination problem if in fact what I observed was not simply normal after 150K miles. .... Dont try to clean it;
is it possible to 'clean' or remove contamination by using a hotter gas (not sure of it's name but burns much hotter) than propane to burn it away?
 
1 - 20 of 40 Posts
About this Discussion
39 Replies
8 Participants
Jcwhite57
Honda Insight Forum
We’re the ultimate Honda Insight forum to talk about Honda’s hybrid car and its fuel economy and specs!
Full Forum Listing
Top