11sORbust

actual af ratio via wideband

11sORbust

the first pull via narrowband

11sORbust

last pull

11sORbust

My eyes tell me that the narrow band has 0 use for wot tuning.

Grim Reaper

A critic could argue that you should disclose the mV for both and plot them on the same graphs. But as you can see, the mV looks quite flat for all rpms in both graphs, while the WB shows the differencebut as you see.

I think plotting both on the same graph would only strengthen the argument that only a WB is useful for tuning. But it would offer definitive proof as to why the NB is useless.

11sORbust

I don't know how to add polt lines with my software. I can say that the mv was 960-925 both runs. You can see that the stock O2 looks the same in both runs. Mv's dip a hair in the same spot and everything. I think it's clear that the NB missed the 17:1 ratio on the first run.

Grim Reaper

Put them in a table in Excel and then graph them.

Low C1500

Was the wide band sensor at the tailpipe? I would have to say something is wrong with one of the sensors (or setup of them) in the first run, maybe the narrow band wasn't warmed up? The narrow band is fast reacting, I find it hard to believe that it should showed a 925 mv when the afr was 17:1.
Good post though.

And yes the wb sensor is the only way to tune wot. But I think a gtech and a narrow band sensor are pretty good too.

11sORbust

what is excel?



yes, and this is a car w/o a catalytic converter. It also haves the borla exhaust, wb on the straight pipe.

The wb had a new filter and was calibrated. The nb was a new heated o2 from a zr-1 vette. It was put on before the trip to the dyno. I can't say 100% that they was both working totally correct. Made sure of closed loop before each run. But they both seemed to be in great working order. I also think the cam could have something to do with the mv readings. The results even shocked me. I really wanted to believe the nb could be used for WOT. But after looking at 8 dyno runs the stock o2 never changed much. New o2 that worked great for part throttle and the wideband had a new filter ad cal'd.

One other little thing that scares me about using the nb. A pal from team3rdgen told me an odd story. He had DFI and nitrous. The nitrous company told him to start with 2X% more fuel for the spray. So he sarted spraying it with 1000mv output from the stock o2. He would reduce the fuel and the car ran faster. The car ran the hardest with the mv at .200-.300. It didn't blow up either. Nice little scarry story.

Low C1500

This the first time I've seen a nb vs. a wb, good info man! I gotta tell ya, I don't trust my nb as much now.

So were you making changes to your VE tables, or your PE afr. And where these baby steps or big changes?

11sORbust

Thanks, I worked really hard to document that day at the dyno. Like I said in other threads, I have the bins,log files and scanned dyno sheets for each run(8 total). I made changes between each pull and everything. I have all 8 dyno sheets/logs/bins in a zip file if you want to check them out.

Tim

11sORbust

I had to lear how to get the a/f lined out, cause it was far off and I was working with static injectors. But if you look at all the data, you can get a feel for how much to add/subtract. I experimented with spark and fuel...

TRAXION

Hey Glenn, I think this is what you were looking for. This is from my 11.66 run. The anamoly in the middle is me letting of the gas slightly when shifting from 2nd to 3rd. I didn't even realize I did that until I compiled this data. You can also see the excessive pump shot that needs to be reduced ... while decreasing the decay rate. I need less pump shot but I need it to hang around longer.

Tim

Low C1500

Nice log trax, the argument could be made that shooting for .880v on narrow band can yeild an ok afr. Thats also the mv that my motor pulled the hardest at .

TRAXION

That argument definitely could be made .... for my setup. I noted awhile ago that 880mV on my O2 sensor yielded a respectable WOT AFR. However, on someone else's setup it was different. Furthermore, VERY small changes around 880 yielded fairly substantial AFR changes. Now, with all that said, since I actually have a comparison of NB vs. WB for my setup ... whenever I see anything in the 900+mV region I know I am way too rich. Anything 850mV and lower then I know I am way too lean. However, I can't emphasize enough that this is on my setup. A friend's setup produced different results.

Nothing acts like a Wideband though. In the graph above you can clearly see that you can get several NB O2 readings for the SAME WB AFR reading. Thus, the NB O2 is clearly not accurate.

Tim

rooster433

Would you care to replot that graph, with either the wide band or the narrow band in the other direction.

The way you have it, the two outputs deviate from each when the car gets richer or leaner.

The reason why a wideband works better for tuning is because its easier to read the 11.0-13.0 AFR's. the wideband is extremely linear while the narrow isn't. The narrowband can read the same AFR but a small change in NB voltage corresponds to a larger jump in afr and its not linear so you can't say that .20mv changes the AFR ratio by X.

A wideband is definately not magic, i'm figureing out if you learn to read both on the same datarun you can find a wealth of information.. particularly useful in finding misfires

ALSO, turn up the sampling rate in your scan tool, alot of times the stock o2's won't see the quick change that your wideband is reporting because your don't sample your stock o2's fast enough.

rooster433

also rescale them correctly so they more closely follow each other.

for instance, graph the wideband from 13.5 TO 10.0
then graph the narrow from 850 to 950


I hate graphs cause you have to take special attention to graph them. They are alot like statistics, you can play with them to get the results you want.

Grumpy

And so far no one's looked at EGT's and backpressure which further skew the NB's.

880 might work in 9 out of 10 cars. Personally I don't want to be running car number ten long term.

Not to mention that it seems like lots of folks are falling prey to ONLY watching a NB O2, and thinking that it's an good answer, when in fact it's not. There are other forums where folks to this day are thinking 850+ is safe and constantly blowing HG's and wondering why.

rooster433

exhaust pressure effects the widebands in the exact same way.

i'd still like graphs replotted, i'll think they will be surpizingly close for you guys.

When tuning, you should tune for what the car likes, not 880mv or even 13.0 on a wideband.

remeber, a wideband isn't a magic wand for the tuners

TRAXION

Rooster433,

I'm not replotting anything. It's on a different computer and I had to make a reasonable effort just to get the JPG over to here. Since you have similar data ... please plot YOUR data in the manner that YOU want. My plot is fine for the point I was trying to get across.

Tim

rooster433

i don't have that data in any excel format.


If you don't want to plot the data again fine, i'm not going to cry.. but i will tell you your not making a good point with the graph.

TRAXION

I'm making the point that I want to make. I'm just not making the point that you want to make. My points are as follows ...

1) The NB and WB show the same trends
2) You can get a really good idea of AFR based on O2 mV so long as you have put a wideband O2 on your own setup and compared that to your own narrowband O2.
3) Small changes in O2 mV can represent big changes in AFR
4) Sometimes the AFR changes with no changes in O2 mV.
5) Sometimes the O2 mV changes with no changes in AFR.

What points are you trying to make with my data?

Tim

rooster433

This is the only point i got to make
enjoy

305sbc

iTrader: (3)

Tim, I also like your graph. I do not see where you lifted off the gas? Are you positive that you did this?
To me it just kind of looks like you went slightly rich at the top of your gear, then lean at the shift recovery which seems pretty normal to me... especially give the big load the engine sees with a manual transmission.
Just wondering.

I also see the point rooster is making about scaling the graph so it's a more accurate visual comparison.

Grumpy

I'm not sure if that is universally true. And while it might be the same way, on some, the extent may be different.

Wouldn't it depend on how you define majic wand.

While it's not perfect, it's a tool that allows one to better use their resources. And sadly plug reading is getting to be a lost art.

rooster433

The tech edge guys have a warning on thier page about supercharged/turboed cars having different exhaust pressure and to be cautious that they don't as they should.


I agree I have no data saying by how much each are changed by exhaust pressure

Temperature also matters..

What matters most I've found is ignition health

Also, I don't know about you guys but i've personally had this happen to me on three different cars now. All supercharged. You can dial the wideband to whatever you want say 12.0 to one but the stock 02's will say 400-600.. well because the car isn't running right for kicks I'll throw in an obserb amount of fuel and the wide band will jump to 13-14:1. From then you can continue to tune and get it right with the wideband.

SO my point is somehow you can get the car so lean that the wideband will actually read in the other direction.

Agian, its happend to me maybe 3 times now, all on supercharged cars (don't know if thats the determining factor or not) and every time i've like trippled the PE values. I've had this occur with 3 different brand wideband controllers all useing the honda sensor.

For me, I'm happier getting it close with the narrowbands, verifying with the wideband, and tuning the rest of the way at the track.

Grumpy

Good old TE.
He doesn't mention that the backpressure difference was something that I first alerted folks about?.
And it was specific to turbo cars where in some applications, there EBP can be 2x the boost level so in a 20 PSI boost situation you maybe looking at 40 PSI. A search of DIY-EFI.org site would give the dates and specifics, of what I found. It's been established by many people that going post turbo is the best answer. The Bosch may not suffer from this at all, but I don't know of anyone stating that as fact.
In the sampling I've done at the tailpipe, vs at the collector, I've never been able to measure any difference. This was with 2 WB's operating at the same time. While not done as or with lab grade equipment I couldn't see where any reasonable amount of EBP effected the readings.
But, like I've said many times, tuning includes notes, and looking at trends.

The operating temp of the sensor's element is 850dC, or about 1,800dF. If your EGT exceeds that then yes, EGT can effect it.

Yes, timing and ignition do effect the O2 level in the exhaust. But, we were agreeing that the WB isn't to be used solely as the final verdict.

And I'd add plug reading to the process you discribe.

rooster433

Have you been able to notice any post turbo/supercharged differences?

The TE website actually says (if i'm not mistaken) that it could occur on the supercharged cars too.

I'm not quite sure how thats possible, or if its revelent enough to matter.

I believe the only reason why I see such anomolys with the widebands is because the cars I've seen this on have been 1 bar speed density supercharged cars. all of which need way more than usual PE settings.

Low C1500

I know its been covered before, but does anyone (TRAX?) want to share what kind of 3 wire NB sensor there running, and did you just get it off a gm vehicle listing (ie: 97 4.3L).

Gotta go get me one right away.

Grumpy

A whole list of them.

https://www.thirdgen.org/techbb2/sho...hreadid=209443

11sORbust

according to my results, the NB is good for nothing. I agree with tim that .XXXmv could be verified for your setup to be X a/f ratio. But when the sensor's output is like the charts I have, I wouldn't trust it.

On the flipside, could an exhaust leak effect both or one of the sensors,theoretically.?

11sORbust

trax has the same sensor used in my test....

rooster433

you and trax might have the same sensor... but you also have a horrible way of graphing them.

I'm trying to show you guys that you can get better results from them if you scale your tables right..

refer to my taco vs burger chart.. noticed how each are scaled.. i tried to make them have the same price curve threw the years BUT look at one goes down while the others going up... look familiar?

maybe i should make the chart hustler vs playboy with a few little visuals.. maybe that will open your eyes

11sORbust

or I could send you the data and have you graph it!? I am horrible at graphing. But I am a HS dropout, so what do you expect? I don't think it takes anything special to see the NB is way off on the first pull. Those are actual bits of log data. I took snapshot of the diag software and just slapped on some text.....

TRAXION

Maybe you need to chill out and contribute something instead of making requests and telling everyone what you think they are doing wrong. I know plenty about graphing and scaling graphs. Why don't you just say what you want to say regarding the NB vs. the WB instead of trying to get someone else to cater their data to support an opinion before you even state it.

As said previously - my graph supports what I wanted to get across. If you have an opinion that you want to get across then post your data according to your own graphing methods. Either put-up or shut-up.

Tim

rooster433

Heres how you can do it

11.5/950mv
12.0/930
12.8/900
13.0/890
14.0/600
14.7/460

BY

your RPM..

Do it on pencil and paper or maybe paint it doesn't matter what you do it in.

**Disclaimer I'm not totally sure about 2-3 of those Y points.. I'm taking a resonable guess.. the important ones between 890-930 13.0-12.0 are correct. BESIDES... what are you tunign for a number or more HP?














I don't have your same datalogging software

rooster433

Why do you get so hostile when someone challenges your idea. If i had the same data laying around I would do it. Doesn't matter u don't seem to listen anyway.

Why don't you just prove me wrong?

TRAXION

This is very similar to things that kvu use to say (no offense Tim - that was a long time ago). What idea of mine have you challenged? What the heck am I suppose to prove wrong? You haven't said a darn thing! Tim and I are the ones who have made conclusions. You're just sitting around asking me to graph stuff for you. So - please tell me what idea of mine that you are challenging AND please tell me what idea of yours is it that I am suppose to be proving wrong????? You haven't said anything!!! I'm not trying to prove anyone wrong I'm just saying that I'm not graphing anything else because the data is on a different computer. If you're so motivated to prove something then please just do it.

Tim

TRAXION

Yea - The TPS% dropped exactly at that point and then went back up. I guess I was worried that it wouldn't shift into 3rd and maybe instinctively did this.

Tim

rooster433

aight

This guy started the post saying the o2's are useless..

Instead of you sharing with the guy, showing him that thats not the case that they are good for something you have to post a hodge podge graph that means nothing towards or against that arguement.. in fact it would just make the normal person believe that is the case, that at wide open they are good for nothing.

Do you understand now?

I've been, the whole time trying to say that they are consistant and pretty good for tuning.

I've been the entire time getting someone who has the data to post a accurate chart that would show what i'm saying..

TRAXION

Well, there ya have it. You are so caught up in your own idea that you not only can't see why I graphed it the way that I did but you also didn't pay attention to what I said. I clearly stated "You can get a really good idea of AFR based on O2 mV so long as you have put a wideband O2 on your own setup and compared that to your own narrowband O2." Ummm - Did you not read that? Isn't that where you are going with what you are saying? And you claim that I am trying to prove you wrong? Stop jumping to conclusions.

I graphed it the way that I did for two major reasons...

1) To show that you can get a really good idea of AFR based on O2mV.
2) To show that a given NB mV doesn't always correspond to the same AFR.

What does that mean? That means that I can use O2mV to get in the ballpark on my setup and that the Wideband will allow me to fine tune and really nail down the fueling.

I already said exactly what you are trying to 'prove' and yet you are poking at me? I don't get it.

Tim

Low C1500

I gotta say that TRAX's graph is pretty clear. A narrow band isn't linear, and a wide band kinda is. So by inversing the NB .mv and plotting it with the WB nothing will be proved other than they went up and down together.

Rooster, I see your point, but man, its pointless.

Although you could probably fudge the NB values and make them just about match the WB, you'd have to use logarithmics though.

JP84Z430HP

Rooster,
If he were to scale the chart differently, it would just make the differences harder to see. Anyone that has any idea about AFR vs. O2 mv knows that they are opposite, lower MV is a higher AFR.

Now, let me throw another idea into the mix here, I may even start a new thread about it. Has anyone messed with the 0-1 volt output on some of the wideband controllers? Has it been more linear, and therefore a more reliable input to the ECM for O2 voltage? I guess I'm just wondering if I get a WB with the 0-1 output, would it's 0-1 be any better than the basic NB O2?

rooster433

Right..

Go check trax's table again

TRAXION

That is exactly what I was trying to say - but you said it so short and sweet Thank you.

Tim

TRAXION

I got access to the computer in question and regraphed using the scales that YOU suggested. Let the fun begin. So tell me rooster .... exactly what does this graph show that my previous one didn't? I'm waiting on your answer. Be specific please.

Tim

11sORbust

rooster, I posted the data to show what I have. I didn't title the thread "narrow band is useless". I tune big cammed,n/a engines. They seem to show the same trend. That the mv is useless for some reason. It could be cam overlap or something. I really can't explain why one can't use o2 mv for the type of engines I work on. For milder apps, the o2 might have some use. But when I see a a/f ratio of 17:1 and a mv readings of .950 then I know that the NB is worth a crap. You can't argue that.

You might have a point BUT you have to have the technical data to back yourself up in this forum. I don't think you have bad intentions. But you are going to get nowhere w/o any hard data. I am the kvu guy tim mentioned. I never had anything in the past to support my theories. So the result was I couldn't PROVE my point. Then the my post turned into bait. So before you start making a point in diy-prom, you better have the data.

Trust me, I wanted to scale the a/f to mv. But I really don't think it's honestly possible. The position of the sensor and a billion other things makes it impossible to make the mv convert to a/f ratio IN GENERAL.

please don't get upset and find the data to support your views. But you really need to take my data into consideration. Cause the factory o2 is far from a a/f ratio gauge.

Tim

rooster433

Pretty consistant from what I've been saying from all along and pretty consistant with every car i've tuned.

They aren't useless, and when graph properly (like the treadstarter can't do) You can see the data you need to make this assumption.

rooster433

and when they are aligned when graphed
noted by my orriginal post XXX/XXX

rooster433

a