Page 1 of 2

Crazy AFR fluctuations

Posted: Sat Oct 08, 2016 9:28 pm
by spoon.ap1
Just started having this idle fluctuation issue recently. It seems that when the car is idling, the AFR's are all over the place but not only that, it varies from 10 - 29 on the data log but does not register on the gauge. I disabled and unplugged the factory O2 sensors and also ordered a replacement harness from the gauge to the sensor. I also tried to run the car in open loop without the wideband connected. Anyone have this problem ever? I'm using Analog0 to run my AEM UEGO gauge which is all plug and play. Also, K-pro and AEM AF readings match.

The voltage (blue) looks really stable but for some reason the AF is fluctuating like crazy.
Image

Below I posted two separate datalogs. "Squamish" is about 2hours long and the other one labelled "today" is only a couple minutes long.

https://www.dropbox....uamish.kdl?dl=0

https://www.dropbox..../today.kdl?dl=0


Here is the calibration:

https://www.dropbox....eltune.kal?dl=0


THANKS IN ADVANCE!!!

Re: Crazy AFR fluctuations

Posted: Mon Oct 10, 2016 8:11 am
by Hondata
The links don't work.

Re: Crazy AFR fluctuations

Posted: Mon Oct 10, 2016 7:10 pm
by spoon.ap1

Re: Crazy AFR fluctuations

Posted: Tue Oct 11, 2016 9:47 am
by Hondata
I'd suggest putting a voltmeter on the wideband output, and see if the voltage is swinging from 0V to 5V. If so, it is the wideband.

Re: Crazy AFR fluctuations

Posted: Tue Oct 11, 2016 10:18 am
by spoon.ap1
Hondata wrote:I'd suggest putting a voltmeter on the wideband output, and see if the voltage is swinging from 0V to 5V. If so, it is the wideband.
Haven't tried that yet but I've ruled out the wideband by using the factory o2 sensors in open loop resulting in the same driveability.

Re: Crazy AFR fluctuations

Posted: Tue Oct 11, 2016 3:26 pm
by Hondata
I don't think the wideband is accurately measuring the lambda. The reasons are that I don't see a corresponding change on the injector pulse when the wideband voltage swings, also that the wideband swing is very rapid and extreme (full lean to full rich), and finally it is under conditions which would give a misfire but I can't see any misfire on the rpm channel. Look at 1:07:17 to see what I mean.
BTW is Squamish.kdl from Squamish BC?

Re: Crazy AFR fluctuations

Posted: Tue Oct 11, 2016 4:57 pm
by spoon.ap1
I see, I'll double check the voltage at the gauge. What do suggest if that doesn't seem to be the fix?

Yeah it's Squamish, BC haha. Are you from Canada?

Re: Crazy AFR fluctuations

Posted: Wed Oct 12, 2016 10:01 am
by Hondata
Nope, but I know Squamish quite well.

Re: Crazy AFR fluctuations

Posted: Fri Oct 14, 2016 1:43 pm
by spoon.ap1
Nice!

Okay, so the gauge seems to be fine. I ordered some EV6 injector pins and plugs(bosch injectors) to directly crimp them on to the OEM wiring because I'm also getting a heavy fuel odour from outside of the car. Assuming maybe the conversion plugs are faulty.

Edit: repined the connectors but still have same issue

Re: Crazy AFR fluctuations

Posted: Mon Oct 31, 2016 11:00 am
by spoon.ap1
Checked the gauge again. It's reading 2.75 volts which is too high, it should be reading 2.35 volts. Does that sound alarming enough?

Re: Crazy AFR fluctuations

Posted: Mon Oct 31, 2016 12:24 pm
by Hondata
How have you got the wideband wired up?

Re: Crazy AFR fluctuations

Posted: Wed Nov 02, 2016 4:56 am
by spoon.ap1
I used an "add a circuit" under the dash that allows you to add another fuse off of for switched power. The wire for the signal going to :Analog 0".

Re: Crazy AFR fluctuations

Posted: Wed Nov 02, 2016 7:32 am
by Hondata
How have you grounded the wideband?

Re: Crazy AFR fluctuations

Posted: Wed Nov 02, 2016 7:11 pm
by spoon.ap1
It's grounded with one of the nuts on the steering column under the dash.

Re: Crazy AFR fluctuations

Posted: Wed Nov 02, 2016 9:24 pm
by Hondata
It needs to be grounded to the same place as the ECU, which on the S2000 is on the back of the cylinder under the cowl. Otherwise you get a large ground loop and possible voltage offsets. Almost all widebands are not well designed as they shunt all the heater current down the same ground as the analog output, but adding in a ground loop does not help.