Topic: Sigh, O2 sensors reading wierd (Read 5263 times)

BoostedSchemes · « **on:** December 24, 2009, 06:39:33 PM »

I have never gotten wideband o2 sensors to read right in this car. I have used an LC1, LM1 and now an LM2 - there are no leaks before the wideband after chasing after them for a long time - even on a 100% stock tune the car will read around 19AFRs with a burble free idle. With large injectors and a tune, it will still read ridiculously lean at idle and cruise.
Then when I get on it and I am 90% sure its running around 12-13afr, it will read 9afrs... like on both ends of the spectrum the car always reads obscenely high or low... it CANT drive around at 19 and run WOT like it does at 9...wtf. Any ideas? The LM2 gets power/ground from the lighter, but the others have been ground directly to ECUs frame ground and powered right from the battery. All have same shitty results.

keelay · « **Reply #1 on:** December 24, 2009, 07:51:37 PM »

Maybe you can fix it with vodka and violence

BoostedSchemes · « **Reply #2 on:** December 24, 2009, 07:56:08 PM »

hasnt worked for the last 2 years but ikeep trying

DmC · « **Reply #3 on:** December 25, 2009, 12:51:59 AM »

Quote from: BoostedSchemes on December 24, 2009, 07:56:08 PM

hasnt worked for the last 2 years but ikeep trying

try powring it from the battery and grounding to the thermostat housing ground.
Also try taking the sensor out of the o2 bung and doing a free air calibration. if that doesn't work try reinstalling the firmware to the lm-2 and reset it to factory defaults.

Joseph Davis · « **Reply #4 on:** December 25, 2009, 03:43:24 AM »

My LM-2 is prone to weird glitches when powered from the cig lighter on some cars. I use a set of large clips that feed a cig lighter socket to bypass that. If you have had the analog widebands grounded at the same point the ECU is - should be on the engine or IM, not the frame - that should read correctly if logged through the ECU. Their displays will always read correctly.

Yuriney, the only cars I've seen switch AFRs like that had crappy over bored injectors. Deutchwerks, Venom, etc. Like flipping a lightswitch, they'd go rich or lean, but it was primarily at idle or really light load.

BoostedSchemes · « **Reply #5 on:** December 25, 2009, 04:01:40 AM »

I have brand new Ford injectors, I can't think of anythign worse then that

but since the problem has lingered through my 3rd set of injectors and the car has run reasonably well at times especially at PT I cant help but feel I am homosexualing something in the WB... JD do you recommend using clips like you said, or just snip the power cord and wire straight to the battery?

Joseph Davis · « **Reply #6 on:** December 25, 2009, 11:01:49 AM »

A good connection is a good connection, whether good clips or good hard wiring.

You've been through other BMWs, and they do not have this problem?

BoostedSchemes · « **Reply #7 on:** December 25, 2009, 12:40:06 PM »

Every other bmw I have had my hands on has had a non innovate gauged unit installed (ngk, plx) that all worked fine. The only unit I have used on other bimmers was the LM-1 and it consistently got readings within .5 afr or much less as compared to mounted gauges. This is something retarded with my own car. I hate myself and smell badly.

I was sure I had a leak causing this, but after much ado and work I have determined I am leak free. Besides my anoose. I grounded the LC1 and wired it supremely well and even xenofraud who may or may not know a thing or two about them verified my install and rma'ed my original unit to no avail.

My next plan of action is to wire a new connector to the LM2 running power and ground straight from the battery... check for perfect batt voltage at the unit, and complete ground at unit...somehow. Fucking little things kill me worse than engine swaps and motronic tuning.

bgdriver · « **Reply #8 on:** December 25, 2009, 01:00:44 PM »

Don't run the reference ground to the battery, ground offsets FTL

http://wbo2.com/sw/wblinout.htm
^ how-to avoid ground offsets

BoostedSchemes · « **Reply #9 on:** December 25, 2009, 01:39:46 PM »

looking at that link reminded me of why I didnt get a tech edge

red to red black to black brotha

Joseph Davis · « **Reply #10 on:** December 25, 2009, 02:09:46 PM »

Techedge = failure.

I'll probably try to fix my 2C0, though, despite the two broken 2A0's being easier, just because I have SMT rework station, multiple cool shaped tweezers, and a cool lighted magnifying lamp. Anyone want to but the thing when I'm done with it?

92CXyD · « **Reply #11 on:** December 25, 2009, 02:15:59 PM »

Quote from: Joseph Davis on December 25, 2009, 02:09:46 PM

Anyone want to but the thing when I'm done with it?

Did you mean buy and are referring the station or the Techedge WB for sale?

DmC · « **Reply #12 on:** December 25, 2009, 04:48:45 PM »

I'm Assuming youve changed the sensor before right as in a differant sensor has been switched out and you still have the same problems?

danz · « **Reply #13 on:** December 25, 2009, 08:01:46 PM »

normally when a sensor is on the way out it will spike to rich and slowly start to read lean and just stay there.

BoostedSchemes · « **Reply #14 on:** December 25, 2009, 10:12:49 PM »

I get a new sensor with each innovate product...

HiProfile · « **Reply #15 on:** December 25, 2009, 11:50:45 PM »

I have an LC-1 *and* a techedge...and oddly enough both work perfectly. Amazing, eh? Maybe because mine was the DIY version...

I call it "use the correct god damn power points".

People will spend hundreds on car parts, but skimp on a few bucks of wiring & connectors. My LC-1 and techedge have received power both from the battery terminals directly, as well as right at the dashboard fuse box from the single-blade accessory slot. I've also used the ground bolt on the t-stat housing to be as "perfect" as possible with zero change. Both log exactly the same via ecu's input as they do via their graphing software (serial data). I've even logged the serial stream of the LC-1 while logging it's WB output wire via an SSI-4, no offset.

All I do is take the single wire for WB output on both and hook it to the ecu's, use good wires, and solder if possible. I don't think my techedge even gets a ground reference via a shunt or ecu, and I have a switch on my LC-1 to use ecu ground or battery ground.

Joseph Davis · « **Reply #16 on:** December 25, 2009, 11:53:03 PM »

Quote from: HiProfile on December 25, 2009, 11:50:45 PM

I have an LC-1 *and* a techedge...and oddly enough both work perfectly. Amazing, eh? Maybe because mine was the DIY version...

Both my 2A0s were. One worked for a dozen cars, then died. The other I went through the rigamarole to troubleshoot and it came down to the SMD ADC chip and ship it back overseas for warranty.

I also have a dead LC-1.

BoostedSchemes · « **Reply #17 on:** December 27, 2009, 09:13:51 PM »

fucking garbage, finally got the o2 sensor to cal to 20.9 AFR and it reads absurd fucking numbers all the time now, time for yet another new sensor

lilpooh21186 · « **Reply #18 on:** December 28, 2009, 12:00:28 AM »

what type of fuel are you using mabey you can find another wb to source and check your afr. btw my lm2 is about .5 leaner than one of the dyno's ive used and .5 richer than the other i tend to just avrage the between them.

BoostedSchemes · « **Reply #19 on:** December 28, 2009, 01:03:28 PM »

im just running on premium... if i everget the car sorted ill retune for e85 so i can have new issues

bgdriver · « **Reply #20 on:** December 28, 2009, 01:43:56 PM »

Quote from: HiProfile on December 25, 2009, 11:50:45 PM

I have an LC-1 *and* a techedge...and oddly enough both work perfectly. Amazing, eh? Maybe because mine was the DIY version...

I call it "use the correct god damn power points". People will spend hundreds on car parts, but skimp on a few bucks of wiring & connectors. My LC-1 and techedge have received power both from the battery terminals directly, as well as right at the dashboard fuse box from the single-blade accessory slot. I've also used the ground bolt on the t-stat housing to be as "perfect" as possible with zero change. Both log exactly the same via ecu's input as they do via their graphing software (serial data). I've even logged the serial stream of the LC-1 while logging it's WB output wire via an SSI-4, no offset.

All I do is take the single wire for WB output on both and hook it to the ecu's, use good wires, and solder if possible. I don't think my techedge even gets a ground reference via a shunt or ecu, and I have a switch on my LC-1 to use ecu ground or battery ground.

This. Most wb02 problems seem to be caused by incorrect or half-ass installation. I have Techedge's cheapest controller with a $50 VW sensor and it's worked perfect for 3 years in two different cars and 40k+, same sensor.

lilpooh21186 · « **Reply #21 on:** December 28, 2009, 05:45:14 PM »

Do you tune using lambda or afr mabey your conversion is set for the wong fuel try lambda mode see if it's still jumpy

News:

Author Topic: Sigh, O2 sensors reading wierd (Read 5263 times)