o2 tuning with headers

[dunce79]

I've done all the basics and I have a solid running LS1 with a TSP 228R cam and mid length headers.
-BRAF
-MAF tune
-VE
-Idle

Then recently went back over MAF and VE. Have them looking great with the wideband. Following this I put the car back into closed loop and started working on the O2s. Back in CL i'm running a bit lean, maybe 15.4 to 15.3. Computer is trimming fuel out. (this is all part throttle stuff.. no WOT stuff yet). So I read up a bit and decided to try adjusting the integrator delay. Made a 12% change for more delay, didn't see a change. Then came across o2 switch points and so I did a log in open loop and have a histogram of the o2 values when my wideband says 14.7. I adjusted the switch points upwards in the airflow modes that were effected and achieved the desired AFR in closed loop. Now back in closed loop my O2s are switching well in airflow modes below 14, with 14 being a little iffy and in mode 16 they switch but only slightly, just kind of bouncing around the set switch point by about 50-100mv. No driveability issues or problems, I was just under the impression that the O2s should switch more drastically for optimal closed loop performance.

So right now I need some direction..
1. should I continue trying to tune the o2s so they oscillate more? if so, why? I have the o2 DTCs turned off, no cats and no emissions to worry about.
2. I've read that the closed loop proportional vs airflow and integrator may be the way to get the o2s to agree with the wideband when the o2s have been moved (headers). Should I put the switchpoints back towards stock and work on those instead?
3. Or leave the switchpoints and integrator and try adjusting the proportional for the airflow modes where o2 switching is less than desired?

My thinking with the switchpoints was this... if in open loop you have your AFR dialed in at 14.7 and the o2s read 600mv then with the current engine setup 600mv would seem to be where the switchpoints need to be regardless of closed loop proportional or integrator settings, right? or am I missing something?

Thanks.

[GHuggins]

What wideband and are you sure it's reading or displaying correctly in the scanner?

[dunce79]

Innovate LC-2, checked the reading from the unit against HP tuners before I started tuning and it seemed correct. Wideband sensor is about 10 inches back from the collector on the driver side and the narrowband o2s are just after the collectors on the mid length headers.

[GHuggins]

My gen 3 O2 settings for headers without cats... Hope they work for you...

[dunce79]

GH -thanks for the screenshot.

From your tables to mine I can see that..
-your integrator delay is virtually double that of mine through most of the table
-you have about 25% less gain on the CL prop vs o2 error
-your CL prop base vs airflow is roughly 30-40% of my current table
-my switch points are lower on the low end and higher on the high end

I'm thinking of setting the switch points closer to yours and then adjust the other 3 tables. Which one would you start with or would you tweak them all? When I had my switchpoints closer to stock I had good oscillation of the o2s in the higher airflow modes.. I was just leaner than should be because the STFTs were constantly negative between 5-7. Once I raise the switchpoints, AFR gets corrected, STFTs hover around 1 to -1 but o2 oscillation falls into a narrow band. So no quite sure if I should leave switchpoints alone and tweak the other related tables or start over with a focus on delay, gain and base adder.

[GHuggins]

Integrator will speed up switch points and so will increasing proportional - problem is you wind up with surging, so knocking the error way down is how you fix that... Basically for gen 3 NA's I multiply integrator and proportional by the same amount "distance increase of the O2's from where they were originally" then multiply error by .1 to .4... I watch my air fuel gauge and see how much it's swinging with different revs to dial in the error - try to get the swings to 1/2 of a point or less...

The above posted values are just some I came up with after lots of trial and error with some long tube headers without cats...

[pdsq98gt]

Are you running 100% gasoline? Or 10% ethanol? If it's 10% then you should tune for a stoich of 14.1. Your wideband will still read 14.7 if on gas scale. Thats why it's a good idea to tune with lambda. I used to tune with afr and stoich of 14.68 and every time I put the narrow band back in, I was seeing fuel trim of 9-10%.

[Eaglegoat]

Are you running 100% gasoline? Or 10% ethanol? If it's 10% then you should tune for a stoich of 14.1. Your wideband will still read 14.7 if on gas scale. Thats why it's a good idea to tune with lambda. I used to tune with afr and stoich of 14.68 and every time I put the narrow band back in, I was seeing fuel trim of 9-10%.

As long as he doesn't make charges now everything will be fine. There's only 4% difference between the two values. When you take into consideration the incremental errors of all the various sensors and tables that 4% is basically meaningless in how the pcm does is calculations.

[dunce79]

Hmm. Good question. Been buying the premium at BP around here, 91 octane but I'll have to double check. Don't think it contains ethanol but one of lower grades does.

[dunce79]

The thing I'm having a trouble wrapping my head around is how adjusting the tables other than the switchpoint values gets your o2s to agree with the wideband if at 14.7 your o2s are reading say 700mv. I guess its in the timing and amplitude ect of the wave.. how fast/long they switch?

[cookmeup272]

As far as running 91 non ethanol compared to 93 with "10%" ethanol I've read different things. If I'm running 93 with ethanol should I change stoich? Or is it not THAT big of a deal? Especially when it says right on the pump the gas could contain UP TO 10% ethanol, which means you never really know exactly how much is in there.

[GHuggins]

As far as running 91 non ethanol compared to 93 with "10%" ethanol I've read different things. If I'm running 93 with ethanol should I change stoich? Or is it not THAT big of a deal? Especially when it says right on the pump the gas could contain UP TO 10% ethanol, which means you never really know exactly how much is in there.

If your MAF and VE are already tuned in with the 10% fuel in the tank, then you've already compensated for the alcohol You change the stoic and ALL tables including open loop temp corrections will need to be redone... This is the reason I don't personally change the stoic setting unless it's running some serious fuel

[cookmeup272]

Currently I'm in the process of tuning the VE table. Haven't gotten to the MAF table yet. Although before I started tuning I wondered about the whole 91 vs 93 with ethanol thing and changing stoich. Started tuning for 91, realized it was easier for me to get 93 and that I figured I would make more power. So started to tune with 93 without changing stoich.

[Eaglegoat]

If your MAF and VE are already tuned in with the 10% fuel in the tank, then you've already compensated for the alcohol You change the stoic and ALL tables including open loop temp corrections will need to be redone... This is the reason I don't personally change the stoic setting unless it's running some serious fuel

If it's a pump gas car I change it before I start tuning. Aim small, miss small.

Currently I'm in the process of tuning the VE table. Haven't gotten to the MAF table yet. Although before I started tuning I wondered about the whole 91 vs 93 with ethanol thing and changing stoich. Started tuning for 91, realized it was easier for me to get 93 and that I figured I would make more power. So started to tune with 93 without changing stoich.

If you change the stoich setting you'll need to redo everything. If you leave it it'll be fine.

[dunce79]

GH, if you want to/have time can you take a peek at these logs and maybe let me know what direction to go in from here?
v42: basically stock closed loop/o2 settings with small tweaks except to the o2 switch points which I had set high to stop negative 5-7 STFTs. This was after VE and MAF dialed in to 14.7 with the WB and then closed loop had it running about 15.2. The switchpoints fixed the AFR but the o2s stopped switching and just kind of run along the switch point.

V43: took your screenshot/advice and made some changes... made a little progress

V44: went further towards your settings than v43, more progress.

What I see are switch points that are probably too fast/sharp.. not staying lean or rich long enough and not moving quite far enough. But I'm new to this so its hardly even an educated guess. Oh, I should note this one is setup to run open loop at idle for now.

I didn't go quite as far as your settings on most because this truck has mid length headers vs long tubes. I would say the 02s are about 12 inches, maybe slightly more from their stock location. 1 and 3/4 tubes. Other than that, 228 R cam with 241 heads, LS6 intake/injectors. Stock MAF, custom air box, 2.5 inch dual exhaust, H pipe.
will put the tune files in the next postv44.hplv43.hplv42.hpl

[dunce79]

tune files
Apache V44.hptApache V43.hptApache V42.hpt

[GHuggins]

Airflow mode - Looks like your Okish in the lower airflow idle areas, but I think your going too high too quickly = the shark teeth O2 waveform your getting in the 2000 rpm + areas... The quick changes are "sorta" what your after as long as your actual air fuel ratio isn't changing more than 3/4 of a point - typically you want it to stay withen a 1/2 point change (you use the error table to control this on the gen 3's - limit table on gen 4's for any future reference)... This is what I would recommend for you - go back closer to my above posted airflow mode and O2 bank to bank MV settings - use the mv bank settings to help control air fuel demanded... Since your mid length headers - multiply your integrator table and proportional tables by 1.45... Change your error tables to what I've posted - PM sent to make things easier

I would also like to comment that other than the cold start or initial startup that needs some work - you've got one of the best dialed in fueling I've seen posted on here in quite some time... Your VE table could use a touch of smoothing and your MAF could probably use some smoothing out up top, but just driving fueling looks pretty dang nice