The 2nd O2 sensor does affect AFR after-all.

[LBV]

Some background info first:

With the precat removed from my Corksport downpipe I soon got a P0139 code where the 2nd O2 sensor (the 1 after the precat but before the main cat) thinks that the pre-cat is malfunctioning and thus throws a check engine light.

So I tried a couple of things to get rid of the code and the only thing that worked was the diode trick explained elsewhere on these forums but basically is a re-write of this: AudiWorld Tech Articles

Since my DP went in I installed a scan gauge with AFR normally 1 of the parameters in view. My AFR has always been around 14.1 to 14.3 in normal closed loop driving such as cruising. I attributed that to the 10% ethanol in my gas as it's got a lower stoich than pure gas (although when I switched to normal no-ethanol gas I still saw the same AFR so was unsure if it really was the ethanol).

Anyway, with my car throwing out some smoke on wot, I decided to install a Magnaflow 3" hi-flow metal substrate cat into the CS DP (part # 59959 btw). I also removed the diode from the 2nd O2 sensor and spliced the wires back together so it was back to "normal" in it's operation.

Anyway, guess what, the AFR's are now upto 14.5 to 14.7 . In the past, the 2nd O2 sensor's purpose was to just check the efficiency of the cat but it appears now that it does more than that.

BTW, the diode was a .5 volt one and richened up the AFR by the same amount (ie. .5 points).

I only noticed this on part throttle/idle ... can't say for sure that it had any effect on wot AFR but I'd say likely not as the O2 sensors are out of the picture in open loop anyway.

Leads one to the conclusion that you can manipulate the closed loop AFR by using different diode voltages.

[Laloosh]

been saying this for a year now lol

[kingpin748]

I thought that the second 02 sensor controlling the AFR was common knowledge?

[LBV]

Originally Posted by Laloosh been saying this for a year now lol

Well I just proved it.

Kingpin: the 1st O2 sensor is what actually controls the AFR much more.

[itzl0l]

Originally Posted by kingpin748 I thought that the second 02 sensor controlling the AFR was common knowledge?

you would think so....but alot of people still argue that it doesnt on this car.

for those of us that dont want a 1st cat, i wonder if u could use a diff volt diode and still keep the CEL off, without throwing the afr off as bad

[bova]

i dont' think the first one controls the afr but is only used to monitor it.

[LBV]

Originally Posted by bova80 i dont' think the first one controls the afr but is only used to monitor it.

Technicalities, technicalities ... but yes - the ecu controls the afr based on feedback from both O2 sensors.

[superskaterxes]

wait you got a cat welded into your CS DP? picts?

[LBV]

Here ya go -- I had about an inch or so cut off each side of the cat before having it welded in. It took a wee bit of the burble I like so much out of the exhaust sound but that's to be expected.

[dadasracecar]

This is EXACTLY what Lou said in a post on M3F about the action of the second narrow band O2 sensor. It's actually VERY important.

[LBV]

Now that you mention it I do recall something about that but I never got the impression that it was definitively stated by Lou ... gonna try and find that post and re-read ...

[MS6_Auburn_Fan]

Lou from cp-e speaks - Mazda6 / Atenza

The link above might be the post being referred too. It was posted here on MSF and then I copied it and posted in on 6 club.

[LBV]

Beat me to it Auburn, but the key section in relation to this post is this (with my bold added):

"The dirty little secret is that the only sensor in the car that has good absolute accuracy at all time is the Nernst-cell narrow-band O2 sensor. Yes, I did say narrow band, and not wide band. In the 80’s the narrow band o2 sensor was developed, and its only purpose in life is to tell if there is excess oxygen in the burnt exhaust gas. This sensor has the unique trait of either working 100% or not working at all. It is a binary device, but an absolutely accurate device, that tells the factory ECU that the combustion gasses have just crossed over - or crossed under stoichiometric (14.7:1) burn. Keep that in mind. This is the only sensor in the engine management system, including the wideband O2 sensor, that can never be wrong, its only modes are working (chemically accurate) or broken and not providing output. So what?

Well, this sensor is used to make the “trims” that the wideband is calibrated from, and correct for all of the tolerances, dirt, etc of the rest of the car’s sensors. Yes, the wideband too is used for closed loop operation, but it is not as absolute as the narrowband. In fact, the wideband’s goal in life is to help the ECU “find” stoich faster. Anyway, the point is that the narrowband sensor is what the ECU uses to characterize all other sensors in the car. This characterization can only occur at 14.7:1 AFR - because that is the only region that the narrowband is operational. So, to summarize, yes, the ECU is constantly trimming all of its other sensors based on the narrowband sensor - but this “trimming” can ONLY be done at low engine loads (closed loop)."

I will add that my fuel trims have barely changed since I removed the diode so it seemed to have little effect ... guess the narrowband 2nd O2 sensor is still either on or off as Lou said but maybe it was happening at a different time with the diode in place??

[itzl0l]

so it changed ur idle/cruise afr's but not really ur fuel trims much?

[ElBartoRex]

so what would be better, the diode mod or the fouler spacer mod if you are running catless?

I did the diode mod and it works, but for some reason the CEL still clicks back on every 800-1000 miles. I clear it with the DH and its gone and but comes back on eventually.

I drove my car 15,000 miles with the 2nd 02 CEL on and have been driving the last 2-3,000 miles with the diode in. I have noticed no drivability differences or worse mileage. My AFR seems like it might be a hair richer with the diode in and the CEL off but thats about it. 98% of my driving is done at 2800-3000 RPMs on the highway out of boost in closed loop mode.

[ChrisK]

Good info.

On a side note I only got a cel once right after the DP and RP install (after 100 miles). Slow reacting o2. Cleared it with the DH and it hasn't been back, 3000 miles now. I was going to do the diode trick but now that's out of the question. Plus the DH can clear it and it stay gone for a long time. I'd much rather just do that, than to screw something up in the long run by messing with a sensor that actually does something to AFR.

[Lymerock]

^^ im with ya on that, i just assume clear it every 15 or so key cycles than cut up the harness and F with the a/f

is anyone running an antifouler on the 2nd 02? does it prevent it

[ChrisK]

I'm running a defouler, I only had the cel that one time. It's the one that came with the CS dp

[kingpin748]

I run the defouler on a CS. I've never gotten a CEL

[ElBartoRex]

I thought CP-E tells all of their customers to do the Diode fix, I figured they know what they are doing with this car.

I am tempted to take out the diode and try the anti-fouler trick instead.

[superskaterxes]

i have the nonfouler and i went about 3k miles before seeing a cel which was about a month ago. ever since then i see the same cel come back once a week. i just tried something i heard on this forum where u cutout part of the cat and u stick it up into the non fouler so that the sensor reads though some of the cat material. i havent gotten a chance to drive it like this as i just did it last night and my car wont be mobile until tomm night. i will be sure to post my findings........

[itzl0l]

^^^^^ were u getting a catalyst ineffeciancy code or a slow to react code

[ElBartoRex]

for the record I only ever got the inefficiency code (PO421) on my car, never the slow to react code.

[ChrisK]

You will only usually get the slow to react code if you are using a spacer or a non-fouler.

[superskaterxes]

slow to react

[Lex]

14.2-14.3 is not correct for cruising and idling AFR. It should be 14.5-14.7. If you are getting those low A/Fs I would expect the (too rich at idle) code to pop and your LTFTs to be way off.

[ElBartoRex]

Originally Posted by Lex 14.2-14.3 is not correct for cruising and idling AFR. It should be 14.5-14.7. If you are getting those low A/Fs I would expect the (too rich at idle) code to pop and your LTFTs to be way off.

my car is more like 14.3-14.7 at cruise. I have never seen any other CELs though. Looks like I might have to fix the diode and try the fouler fix.

Don't let Arkenzo see this thread-- he gets very defensive when you try to debunk his CEL fix.

http://www.mazdaspeedforum.org/forum/foru...cel-fix-3.html

[Laloosh]

Originally Posted by Lex 14.2-14.3 is not correct for cruising and idling AFR. It should be 14.5-14.7. If you are getting those low A/Fs I would expect the (too rich at idle) code to pop and your LTFTs to be way off.

ive been running like this for 7 months, never through a cel code for the cat or being rich

[LBV]

Lex - I never got a CEL due to the lower AFR's and I suspect that's within spec as the car ages and the cat's start to go, etc..

Knowing what I know, I would still not hesitate to do the diode trick again ... although I'd try the piece of catalyst in the extender referred to above first for sure.

As an example of the effect the removal of the diode had on my LTFT's:
- I've been writing down my LTFT's since I knew I was getting the Cobb AP back in late Sep iirc. Anyway, this is what the LTFT's were before the removal of the diode and after (with considerable driving to allow settling).
- idle/decel went from 3.1 to 3.9
- accel went from 0 to 3.9
- cruise went from -.7 to 1.5

So as you can see, it's a minimal change and certainly within the scope of normal fluctuations.

[ElBartoRex]

forget it then, im not messing with it again if its that minimal.

on the same note, has anyone here with the diode trick gotten inspected and passed?

I am nervous that when I get my emissions done (OBD Scan, no sniffer) it will detect a 'not ready' sensor or something.

[itzl0l]

^^^^^^^^ i was about to ask the same thing. ALSO if u clear the code with the dashhawk or AP will it pass or will the system be "not ready" right after the code is cleared?

[Lex]

Originally Posted by Laloosh ive been running like this for 7 months, never through a cel code for the cat or being rich

You idle at 14.2-14.3 according to the DH? My LTFTs and STFTs work very hard to make my idle 14.58-14.69 AFR according to the DH.

[LBV]

Thought I'd mention another possible solution to this and that's to add an O2 bung after the main cat and just extend the 2nd O2 sensor there.

If you've got an old O2 sensor from another car even sitting around just cut a section of wire out and splice it in with some butt connectors to get it extended ... it should work as I had to splice in some wire from an old sensor when the diode came out (couldn't get the butt connectors holding the diode in place to separate so I lost a section of black wire there).

These sensors aren't cheap ($260 Cdn quoted - the 1st O2 = $340) so be aware but if anyone trys this, let us know how it worked out.

[ElBartoRex]

my question is, if you used the anti-foulers instead of the diode, wouldn't you get the same slightly richer AFRs?

[ChrisK]

Originally Posted by ElBartoRex my question is, if you used the anti-foulers instead of the diode, wouldn't you get the same slightly richer AFRs?

I have a defouler and no cats and my AFR is exactly 14.7 idle and cruise. Fuel trims are almost prefect too, STFT usually stays 0.0, LTFT is 1.7 or so.

[LBV]

I never had a scan gauge for the short time I had the O2 non-fouler on so I can't say for sure ElBarto but I'd agree with ChrisK ... there's no tampering with the wiring voltages by just moving the O2 sensor outside the pipe a few inches as the non-fouler does so it should read 14.7.

[superskaterxes]

Originally Posted by ChrisK I have a defouler and no cats and my AFR is exactly 14.7 idle and cruise. Fuel trims are almost prefect too, STFT usually stays 0.0, LTFT is 1.7 or so.

+1 i report the exact same thing.....

previous of my experiment above

[evidence]

Incase anyone cares here is the original thread Lou posted in. There is some other good chatter if you can make it through all the Cobb vs CPE

LTFT's, Intakes, Piggyback and a closed loop ecu

[Laloosh]

i idle and cruise at 14.2 with a diode or fouler. The ltft do not change if i decide to run the stock o2 sensor config. They are always the same. 14.7/14.2, it doesnt matter

[Lymerock]

Originally Posted by ElBartoRex for the record I only ever got the inefficiency code (PO421) on my car, never the slow to react code.

thats the code I get. I guess my txs pipe has an extended bung on it already, so perhaps that qualifies as a defouler. anyway my afr's are so close to 14.7 often that I wonder if its possible, so my 2nd 02 is obviously happy, but it would be nice to figure out a way to not get the inefficiency code time to time.

now when people say take a peice of catalyst out and stuff it in the defouler I dont get it. when i look in the stock ms3 cats it looks like ceramic to me, not something I want to go trying to rip a piece off of. Am I missing something?