Are my O2 Sensors screwed?

[vbhero]

Hi All,

After months of reading, trial and error and help from my tuner ( who has be teaching me, OBI WAN - Padawan style :) ) etc, I finally think I have my SDCL cam tune done. Commanded AFR vs WBO2 AFR were close enough for my liking.
But after re-fitting the NBo2 last night, re-enabling STFT and LTFT etc I am noticing some fairly decent trim figures, along with what appears to be a lazy / failing o2 (Bank2).
Could someone please have a look at the attached log and see if I am looking at this right?

Car is a Holden VXII SS A4 LS1.
Mafless
215 / 223 .604 .610 112LSA 112ICL Cam,
Pacemaker 4->1, 100cpi 4in cats, 3.5in single exhaust.
Custom OS 2

I have attached the following logs for info:
20140922 3rd run was after I had PE etc dialled in to what I believe to be acceptable limits.
20140923 Final is a short run to confirm normal part throttle stuff, and again every thing appears to be in normal limits. Except for some random rich spikes...
20140924 CLSD is this mornings short log, with the WBo2 removed, and the NBo2 refitted back into bank two. LTFT and DFCO re-enabled etc. This is the log that I am shaking my head at in confusion.
20140924 CLSD(1) is another short log that I have done which yields similar info to this mornings.

This tune did my head in for a while, as I wasn't trusting the WBO2 data, as it was telling me that I was lean across the board, despite having done a Calc.VE with the MAF. After much coaxing and further logging etc, I am thinking the Calc.VE I done was crook. The car is driving a lot better, a lot stronger and smells better.

So based on todays two logs, would it appear that HO2S21 is on the fritz? So to me this would throw the STFT and LTFT out?

Thanks in advance
17529
17526
17527
17528

[joecar]

Swap the NBO2's between banks and see if the laziness follows the NBO2's.

Do you have any leaks in the exhaust headers/plumbing...?

Did you do Calc.VET from post # 1 here: Calc-VET-correcting-MAF-and-calculating-VE-(in-single-log) (https://forum.efilive.com/showthread.php?15236-Calc-VET-correcting-MAF-and-calculating-VE-(in-single-log))

[vbhero]

Swap the NBO2's between banks and see if the laziness follows the NBO2's.

Do you have any leaks in the exhaust headers/plumbing...?

Did you do Calc.VET from post # 1 here: Calc-VET-correcting-MAF-and-calculating-VE-(in-single-log) (https://forum.efilive.com/showthread.php?15236-Calc-VET-correcting-MAF-and-calculating-VE-(in-single-log))

I'll do the swap and see what happens.

I followed the tutorial available for download via the main efilive web site http://download.efilive.com/Tutorials/PDF/Calc.VE%20Tuning%20Tutorial.pdf and all seemed to be okay. The only exception at this point was I did not have access to a WBO2. I did this pre cam, then used the data to run SD and removed the MAF.
After fitting the cam, I refitted the MAF, and re-did the Calc.VE process, then again removed the MAF.

Trims seemed pretty good, but car seemed thirsty, but I put this down to short trips etc. When I got a hold of and put the WBo2 in, it painted a very different story telling me I was lean across the board. So I assumed I had stuffed something up in the Calc.VE process, trusted the WBO2 data and subsequently completed what was essentially the Auto VE process. The only deviation there, was I used a serial WBO2 not an analogue.

[vbhero]

Swap the NBO2's between banks and see if the laziness follows the NBO2's.

Do you have any leaks in the exhaust headers/plumbing...?

Did you do Calc.VET from post # 1 here: Calc-VET-correcting-MAF-and-calculating-VE-(in-single-log) (https://forum.efilive.com/showthread.php?15236-Calc-VET-correcting-MAF-and-calculating-VE-(in-single-log))

mmm odd, I replied earlier and it's not here....

I will swap the o2 over tomorrow and see what happens.

I did Calc VE following the tutorial available here http://www.efilive.com/documentation-tutorials. The only exception to that is I did not use a WBO2 during this. I did a tune before the cam and again after the cam. The trims seemed okay from memory. But it did seem thirsty, but I put that down to short trips
When I put the serial WBO2 in, I did what is essentially the Auto VE process using a BEN table. That told me I was lean right across the board. I am inclined to believe that data, as the car seems to be running much better, and smoother.

Oh and I don't think there are any leaks in the exhaust

[vbhero]

Ok this forum is doing my head in... I can see that there are two replies, with mine being the last, but here I am again not seeing my last reply to answer your questions Joecar.

Any way, I have swapped the O2 sensors and this is the resultant logs

17551
17552
17553

The short log "O2 Swap" shows similar issue and that it has not swapped sides as I would have expected.
The next two logs show them running as expected with no discrepancy. I am still not happy with the trims though, and still not sure whether to trust the NBO2s, I can say that now it is in SDCL that the car doesn't feel as strong, almost very flat in the mid range. I am thinking of going back to OLSD and seeing how it goes.

To answer your question about how I did the CalcVE, I followed the .pdf tutorial available from the tutorial section of the main EFI Live page. The exception being I didn't have the WBO2

[joecar]

...

I did Calc VE following the tutorial available here http://www.efilive.com/documentation-tutorials.
...That tutorial is a little outdated... it is superceded by post #1 of Calc-VET-correcting-MAF-and-calculating-VE-(in-single-log) (https://forum.efilive.com/showthread.php?15236-Calc-VET-correcting-MAF-and-calculating-VE-%28in-single-log%29)

[vbhero]

That tutorial is a little outdated... it is superceded by post #1 of Calc-VET-correcting-MAF-and-calculating-VE-(in-single-log) (https://forum.efilive.com/showthread.php?15236-Calc-VET-correcting-MAF-and-calculating-VE-%28in-single-log%29)

Ah righto then. Well, that aside I did go through and do a Auto VE, using the Serial WBO2, so that should have picked up and corrected any issues that I had with the CalcVE?
Well to me that should be the case, yet here I am seeing trims much larger than I anticipated. Yes there was a degree of hand smoothing involved in the VE table and I expected some trimming, but not to the extend seen in parts of the log.

I am giving serious thought to running SDOL...

[joecar]

Is your FPR manifold-referenced or un-referenced...?

[vbhero]

If your FPR manifold-referenced or un-referenced...?

You lost me there... ??

[IJ.]

You lost me there... ??

Is your Fuel Pressure Regulator plugged into the Manifold vacuum via a small vac hose?

[vbhero]

Is your Fuel Pressure Regulator plugged into the Manifold vacuum via a small vac hose?

I will have to have a look. I have not played with any of the FPR stuff, so it should be stock standard for a Holden LS1

[vbhero]

Turns out standard Holden config is unreferenced.

[joecar]

You lost me there... ??Sorry... I meant to say this:


Is your FPR manifold-referenced or un-referenced...?

[joecar]

Turns out standard Holden config is unreferenced.Ok, check if your B4001 IFR table is sloped.

[vbhero]

17554

Does that help? These are stock standard settings, for stock injectors. I have not touched them at all

[joecar]

17554

Does that help? These are stock standard settings, for stock injectors. I have not touched them at allHmmm, Do you have a variable speed fuel pump...?

What is measured rail pressure...?

What injectors do you have...?

[darcy]

That looks like a standard Holden IFR table....

WRT the difference in how strong it feels; is fuelling enrichment controlled the same in both OL/CL tunes?
In logs 004 & 005 there are several instances of MAP>90kPa while fuelling is still at stoich. This definitely won't feel very sharp.

[statesman]

I think your O2s are stuffed.

[vbhero]

Hmmm, Do you have a variable speed fuel pump...?

What is measured rail pressure...?

What injectors do you have...?

Pump is stock
Injectors are stock
I don't know what rail pressure is.


darcy
That looks like a standard Holden IFR table....

WRT the difference in how strong it feels; is fuelling enrichment controlled the same in both OL/CL tunes?
In logs 004 & 005 there are several instances of MAP>90kPa while fuelling is still at stoich. This definitely won't feel very sharp.

Yes, that is the standard IFR table. I have not done any mods to the fuel system yet. The only mods are whats listed in my opening post.

The only difference between OL and CL is the disabling of the trims etc PE etc weren't touched between the two.
But in saying that, table B3605 comes into play when in OL, which alters fuelling from about 85kPa and up at operating temp. .963Lambda @ 85kPa, to .855Lambda @ 100kPa. My PE is .858 So that could explain why it feels better in OL?


statesman
I think your O2s are stuffed.

I am thinking that too. 152,000km on the clock and I have no idea if they are original or not, chances are that they are.

[joecar]

If that's the stock Holden IFR table, then the stock Holden VE table will match (i.e. is calibrated to the IFR table).

[vbhero]

If that's the stock Holden IFR table, then the stock Holden VE table will match (i.e. is calibrated to the IFR table).

Really, even though I just changed the cam? Which is the reason for this whole process?

[statesman]

Keep an eye on your injector duty cycle... with your new cam, the stock injectors might struggle to deliver enough fuel at full noise.

[vbhero]

Keep an eye on your injector duty cycle... with your new cam, the stock injectors might struggle to deliver enough fuel at full noise.

That is something that I have been considering after seeing the results for the VE table after doing the AutoVE. I have figures >100%, which according to the bloke who has been working with educating me, could be a sign that I am pushing the limits of the stock fuel system.
I might start logging INJDC over the next day or two and see what results it shows.

[joecar]

Really, even though I just changed the cam? Which is the reason for this whole process?I meant to add: if you now did AutoVE then the VE table will match the IFR table...

( instead of the VE table modelling the true cylinder airmass, it will have the IFR table's deviation from the squareroot-pressure relationship calculated into it... it should stilll cause to VE table to work, but there may be some conditions where it will be incorrect... I suppose this is how Holden did it )

[vbhero]

I meant to add: if you now did AutoVE then the VE table will match the IFR table...

( instead of the VE table modelling the true cylinder airmass, it will have the IFR table's deviation from the squareroot-pressure relationship calculated into it... it should stilll cause to VE table to work, but there may be some conditions where it will be incorrect... I suppose this is how Holden did it )

So your saying that the VE table that I now have as a result of doing an Auto VE should be matched to the IFR?

The other thing is I have just upgraded to COS3, I did this to give me access to the RPM vs MAP open loop fuelling table

[joecar]

So your saying that the VE table that I now have as a result of doing an Auto VE should be matched to the IFR?Yes...

the problem is this: by itself the IFR table is wrong (does not match the actual flowrate of the injector), and by itself the new VE table is wrong (does not match the airmass that the cylinder is actually filling with), but together the IFR and VE will provide correct fueling in *most* cases... the problem arises when the PCM wants to use just the VE by itself (for example, to calculate airflow rate for comparing to MAF or to sanity tables, or for calculating engine torque and transmission line pressure) it will have incorrect values... depending on the extent of the engine modifications, you may or may not run into problems with this;

if you look at other IFR tables (for example 2001/2002 Camaro/Firebird/Corvette) you will see that they follow the squareroot-pressure relationship (i.e. they don't have that hump that you see in the Holden IFR)... this means that the VE table (which was calibrated to match the IFR) represents more correctly the cylinder airmass (meaning that the other calculations more closely model what the engine is doing).


The other thing is I have just upgraded to COS3, I did this to give me access to the RPM vs MAP open loop fuelling tableThis helps if you're going to run in OL all the time; otherwise you'll be in OL only while engine is warming up and then it will be in (and stay in) CL mode (or a derived CL mode when you go to WOT/PE) unless the O2's fail.

[vbhero]

Yes...

the problem is this: by itself the IFR table is wrong (does not match the actual flowrate of the injector), and by itself the new VE table is wrong (does not match the airmass that the cylinder is actually filling with), but together the IFR and VE will provide correct fueling in *most* cases... the problem arises when the PCM wants to use just the VE by itself (for example, to calculate airflow rate for comparing to MAF or to sanity tables, or for calculating engine torque and transmission line pressure) it will have incorrect values... depending on the extent of the engine modifications, you may or may not run into problems with this;

if you look at other IFR tables (for example 2001/2002 Camaro/Firebird/Corvette) you will see that they follow the squareroot-pressure relationship (i.e. they don't have that hump that you see in the Holden IFR)... this means that the VE table (which was calibrated to match the IFR) represents more correctly the cylinder airmass (meaning that the other calculations more closely model what the engine is doing).

This helps if you're going to run in OL all the time; otherwise you'll be in OL only while engine is warming up and then it will be in (and stay in) CL mode (or a derived CL mode when you go to WOT/PE) unless the O2's fail.

Sorry for the late reply, but I have been all over the shop the past week....

Anyway, I am convinced that there is something screwy with the O2 sensors, so that is why I went to COS3 and ever since the car has been running beautifully. I will need to put the WBO2 back in at some point just to double check AFRs.
I just need to now fine tune the RPM vs MAP table. What I have done initially seems to work fine, but I am new at this, so I am sure that while my seat of pants says its okay, there is probably heaps of room for improvement. :D

[joecar]

Just make sure PE enables correctly (see 2002 Camaro table B3616), and that PE is sufficiently rich (and no more) verified with wideband (EQR 1.165-1.175 for NA, EQR 1.25 for boost).

[vbhero]

Just make sure PE enables correctly (see 2002 Camaro table B3616), and that PE is sufficiently rich (and no more) verified with wideband (EQR 1.165-1.175 for NA, EQR 1.25 for boost).

My B3616 is almost identical to the 2002 Camaro, but mine comes in sooner in the 3200 - 4000 cells.

I also noted that I have my PE modifier set at .858 Lambda / 1.165 EQR in all cells. But the stock tune and the tune of the Camaro you referenced have a stepped increase.
The stock PE modifier for the Commodore goes as rich as 1.3 EQR, the stock Camaro goes as rich as 1.28 EQR.

So I guess my next question is, is it better to "step" the PE modifier? Ie gradually richen as the RPM increases, and also, why do the stock tunes go so rich?

EDIT: I am confident that the PE is accurate as verified with WBO2

[Jester]

I wish my cam played that nice, good job. My thoughts on the stock tune PE being so rich is to allow a fair bit of safety margin, keep warranty to a minimum etc, maybe other reasoning that I'm missing.

[joecar]

...

I also noted that I have my PE modifier set at .858 Lambda / 1.165 EQR in all cells. But the stock tune and the tune of the Camaro you referenced have a stepped increase.
The stock PE modifier for the Commodore goes as rich as 1.3 EQR, the stock Camaro goes as rich as 1.28 EQR.

So I guess my next question is, is it better to "step" the PE modifier? Ie gradually richen as the RPM increases, and also, why do the stock tunes go so rich?

EDIT: I am confident that the PE is accurate as verified with WBO2

( the Camaro/Firebird PE table overfuels... many of us have found more power when we lean it a little )

Without a dyno/dragstrip, just set the PE flat...

( if VE and MAF tables are correct, then the PE EQR you set will show up on the wideband )

Verify that wideband shows PE to be EQR 1.165-1.175, regardless of what PE table is set to (if VE/MAF are correct, then wideband will match PE table )

If you have access to a dyno (or umpteen passes at a dragstrip), you can shape the PE table to give you best peak torque and lean it a little (relatively) as you reach peak RPM to give you best peak HP...

otherwise without a dyno it is guesswork.