240 replies to this topic

[3VOLUTIONIST]

<sub>this is my oxygen sensor at idle


and this is yours</sub>

I was just watching JafroMobile's newest video, his Wideband install video, and I was looking at his narrowband values on his ECMLink. He was simulating it with an innovate wideband, but yeh it was jumping between like 0.2 and 0.8, whereas mine seems to be a lot "flatter". Tomorrow I'll measure the simulated narrowband output from my PLX Wideband module. I can snip the narrowband o2 sensor's input and wire in the simulated narrowband if I have to.

Jafro's YouTube channel is great by the way, everyone should check him out

Ok so, thinking thinking thinking.
My narrowband doesn't seem to be sensitive enough. Or, is intermittent. Or I have a wiring issue (as it did it with my previous 2 wire). The fact it did it with the two wire indicates a wiring issue. Maybe a frayed connection somewhere..

Edited by 3VOLUTIONIST, 17 May 2015 - 01:49 PM.

[vr401]

Something is wrong.
Wideband O2 sensor works different to stock narrow band.
When you said that it's all good once you turn off closed loop it made me thinking that all your scaling for afm and injectors are good.


You know when you said this:

I was thinking, if open loop doesn't use STFT, then it must be a software issue, not a hardware issue, because it idles perfectly in open loop (AFR 14.7 . So that rules out a vacuum leak & injector scaling doesn't it?

That proves your injectors and afm scaling are spot on but closed loop using STFT and bad signal from O2. Not software issue only bad O2.

[3VOLUTIONIST]

Something is wrong.
Wideband O2 sensor works different to stock narrow band.
When you said that it's all good once you turn off closed loop it made me thinking that all your scaling for afm and injectors are good.


You know when you said this:

I was thinking, if open loop doesn't use STFT, then it must be a software issue, not a hardware issue, because it idles perfectly in open loop (AFR 14.7 . So that rules out a vacuum leak & injector scaling doesn't it?

That proves your injectors and afm scaling are spot on but closed loop using STFT and bad signal from O2. Not software issue only bad O2.

Awesome, means I dont have to shell out cash I dont have for another MAF at least haha

 

My PLX wideband module has an output for simulated narrowband, allows you to replace the standard narrowband if you like. im not using it at the moment, but tomorrow ill wire it into the ECU (i have some spare ECU pins) and ill see how I go, fingers crossed it solves the issue.

Edited by 3VOLUTIONIST, 17 May 2015 - 02:49 PM.

[3VOLUTIONIST]

Righteo! Everything is remedied
Used one of the old wires from when the OBDII port was on the drivers side to send the narrowband output from my wideband module over to the ECU, put a pin on the wire, unplugged the Lamba sensor pin, plugged in the wideband's narrowband signal. No problems! Idles/drives/cruises/boosts fine. Running 15psi at the moment with minimal knock, a few slow CEL flashes, but not consistently. Should be good to get tuning
Surprisingly it pulls all the way to redline at 15psi. I was expecting the 14B to drop off at about 6500rpm.
Maybe porting out the turbine housing and exhaust manifold did make a difference

Edited by 3VOLUTIONIST, 18 May 2015 - 12:43 PM.

[leadfoot]

Great news all these issues and down to a bad 02 sensor. Probably fowled up from other issues earlier on.

[3VOLUTIONIST]

Great news all these issues and down to a bad 02 sensor. Probably fowled up from other issues earlier on.

I dont see how, it was freshly installed with the new turbo and front pipe. Maybe I just got a dud, bad luck I guess

Unless it was damaged in the post

[leadfoot]

well looking at those nice square waveforms from your log something was definitely fishy about it

[3VOLUTIONIST]

well looking at those nice square waveforms from your log something was definitely fishy about it

Yeh, not happy. Will be speaking to the seller about it.

[vr401]

Great news mate. Enjoy tuning.

Btw: I believe CEL flashing is tephra option for knocks so watch your tune

[evo-gsr]

Agreed...mine is setup to flash CEL for knock also ^

[jack be nimble]

Woohoo.

[vr401]

Woohoo.

 

 

[3VOLUTIONIST]

Yup. I get a few slow flashes (3 counts of knock), but not recurring which is good.

Next, figure out how to log my wideband via the 0-5v signal it outputs, I've read I can send it to the rear o2 ECU pin and log that with some tweaks in evo scan.

The other thing is the deceleration fuel cut. I need to stop the ECU turning off the injectors as soon as I come off the accelerator, it does it between shifts sometimes, makes for a jerky drive.

[jack be nimble]

I think there is a timer for it that might fix it. Or nlts!
I actually dont take my foot all the way off the throttle when changing, just on a bit, and on a bit more for the down change (that i often double clutch anyway). Around town if im just pottering i double the upchange (without the blat) as well.
I dint like changing synchro rings much. My2c.

[3VOLUTIONIST]

I think there is a timer for it that might fix it. Or nlts!
I actually dont take my foot all the way off the throttle when changing, just on a bit, and on a bit more for the down change (that i often double clutch anyway). Around town if im just pottering i double the upchange (without the blat) as well.
I dint like changing synchro rings much. My2c.

Yeh, downshifting I dont really have any issues because I rev match it, third gear synchro is getting old.

But upshifting can be a bit iffy at the moment. I doremember seeing a Decel fuel cut delay option in my Evo 6 ROM, ill check the 7 ROM when I get back from uni

[jack be nimble]

I remember you saying it, and me wondering what it might be used for...
can we have a screenshot of the nb sim trace, to compare to vr401 post please. When you get a chance.

[vr401]

Yeh, downshifting I dont really have any issues because I rev match it, third gear synchro is getting old.

But upshifting can be a bit iffy at the moment. I doremember seeing a Decel fuel cut delay option in my Evo 6 ROM, ill check the 7 ROM when I get back from uni

it's not there but this should do. Just add it to 90550001 XML

 

<table name="Decel Fuel Cut Delay #1" category="Fuel" address="4120" type="2D" level="2" scaling="uint8">
<table name="Engine Load" address="50a4" type="X Axis" elements="7" scaling="Load"/>
</table>

<table name="Decel Fuel Cut Delay #2" category="Fuel" address="412e" type="2D" level="2" scaling="uint8">
<table name="Engine Load" address="50a4" type="X Axis" elements="7" scaling="Load"/>
</table>

<table name="Decel Fuel Cut Delay #3" category="Fuel" address="413c" type="2D" level="2" scaling="uint8">
<table name="Engine Load" address="50a4" type="X Axis" elements="7" scaling="Load"/>
</table>

<table name="Decel Fuel Cut Delay #4" category="Fuel" address="414a" type="2D" level="2" scaling="uint8">
<table name="Engine Load" address="50a4" type="X Axis" elements="7" scaling="Load"/>
</table>

<table name="Decel Fuel Cut Ignition Retard" category="Fuel" address="4164" type="2D" level="2" scaling="Timing">
<table name="Engine Speed" address="4d76" type="X Axis" elements="10" scaling="RPM"/>
</table>

<table name="Decel Fuel Cut Resume Enrich" category="Fuel" address="3aaa" type="2D" level="2" scaling="EnrichmentAdj">
<table name="Engine Temp" address="5102" type="X Axis" elements="9" scaling="Temp"/>
</table>

<table name="Decel Fuel Cut RPM Offset" category="Fuel" address="37f4" type="2D" level="2" scaling="RPM8">
<table name="Engine Temp" address="50e8" type="X Axis" elements="8" scaling="Temp"/>
</table>

[3VOLUTIONIST]

Pasted that into my ROM, think I've got it working, first time I did it it just removed all the tables before ISCV control haha

Why are there 4 decel fuel cut tables though? I'm assuming just set all the values the same.

 

Still learning how to use evoscan, my table looks different to yours. But here it is, Thats only like 20 seconds of log though, for some reason it stopped logging the sensor after that, I'll have to try again later to get a proper reading.

 

Did some research on logging the wideband using the rear o2 input (pin 75)

Founds this, looks interesting, but its for an evo X. http://www.evolution...out-serial.html and this http://www.geekmappe...hread.php?t=337

 

Actually my fuel trims are surprisingly good, I was expecting them to be huge, looks like theyre about -6%

Attached Thumbnails

• 

Attached Files

•  EvoScanDataLog_2015.05.18_19.39.02.csv   3.46MB   100 downloads

Edited by 3VOLUTIONIST, 19 May 2015 - 06:04 AM.

[vr401]

Pasted that into my ROM, think I've got it working, first time I did it it just removed all the tables before ISCV control haha

Why are there 4 decel fuel cut tables though? I'm assuming just set all the values the same.

 

Still learning how to use evoscan, my table looks different to yours. But here it is, Thats only like 20 seconds of log though, for some reason it stopped logging the sensor after that, I'll have to try again later to get a proper reading.

 

Did some research on logging the wideband using the rear o2 input (pin 75)

Founds this, looks interesting, but its for an evo X. http://www.evolution...out-serial.html and this http://www.geekmappe...hread.php?t=337

as long as you keep your fuel trims between -5% to +5% is ok. As for logging the AFR do a bit of search on how to log AEM wideband using.  

Read here,

http://www.evoscan.c...wideband-review

[3VOLUTIONIST]

So, if I send my wideband signal to pin 75 (rear 02). I add a logging parameter to EvoScan.

Display: Wideband

LogReference: AFR

Request ID: The guide says 3C, however im not sure what this is or whether it changes between ECU's

Eval: 0.039216*x+10   (Wideband has an AFR range of 10-20)

Unit: AFR

MetricEval: 0.039216*x+10

MetricUnit: AFR

ResponseBytes: 1

GaugeMin: 10

GaugeMax:20

ChartMin: 10

ChartMax: 20

Scaling Factor: 10 (guide says this however I'm not sure if it changes)

 

I'll do the wiring then figure out the software