Just a few WBO2 sensor questions

[91dime]

First im running a 7427 on my s10 with a stock 5.7. For the files, I use the Advanced $0D TP5 v251.xdf and the BJYL 350 TBI 700r4.bin along with the MTX-l WBO2 sensor.

1st o2 question: I followed the instructions in the bin/xdf on setting up data logging of the WB and it works but the readings are off by around +1 in tunerpro compared to the gauge. As I type this it comes to mind that I have my gauge programmed to the E10 blend so 14.1 instead of 14.7. Would this throw it off in TP? If so, how would I come up with the formula to fix that?

2nd: I have the WB on the driver side stream and the NB on the passenger stream. Iv got a leak at the collector on the Pass side throwing off my readings pretty bad so I unplugged it and run OL for now. Iv read a lot on the simulated NB from the WB but saw no exact numbers to make sure the PCM is getting the correct voltage. Will the factory setting on the mtxl work?

3rd: If the WB simulating NB is as bad as I hear, is there a page to explain how to tune only using the WB? Iv searched google and haven't had any luck.

side note: it took over 2 hours to get the 3 collector bolts on the pass side tightened and it still leaks, so im really wanting to avoid changing the gasket. The reason i think it has a leak, with throttle applied it drives fine, when i let off it starts surging way lean then recovers then lean again. It has stalled but is rare. If i unplug the sensor it runs fine. Also this is a heated o2 that has worked fine over the past year.

That's the main stuff for now until I can think of more. Any help is greatly appreciated!! Thank you!

[91dime]

A small update, I set my WB back to 14.7 to see if the XDF will show proper readings and its around +/- .5. Not sure if it was intended to be exact or not.

Also had the simulated NB going at at idle the truck would run fine for 10-15 sec then start swinging between rich and lean for 5 sec or so then stall, I unplugged the wire and the issue went away.

Another thing I noticed in the BLM tables, data was being populated with no o2 connected other then the WB signal. Is this XDF set up to populate BLM data using the WB signal?

[steveo]

is there a page to explain how to tune only using the WB?

first set your open loop AFR to a constant value (say X=14.0:1) within the range you want to tune (the entire range is not a bad choice, except for warm-up, which you could just leave alone...)

log the wideband for quite a long time, trying to flex your engine and try to reach as many operating conditions as possible, and chart the results, filtering out any data that you find unreasonable (such as power enrichment, hard decel, or cold coolant temperature below the afr value you just set above).

figure the AVERAGE wideband AFR (since it will fluctuate) in the log for target RPM and MAP (for VE) or per AFGS (for a MAF car) where wideband_afr/X=multiplier. you can use a spreadsheet for this, or you can try to use tunerpro's built in tools, or you can do what i did if you have any coding skills and create a static data analyzer to do the work for you. interpolate the results (saying 'this curve must continue here' or 'this afr point must be wrong') to construct a best-guess modified VE or MAF table.

then go back to the beginning and do it over. if you did well, this round wont require much adjustmetnt at all. until your wideband reads X +/= 0.2 on average, don't stop.

after that, just set your open loop afr table, pe table, and all taht to the AFRs you want and they will be very close to reality.

[91dime]

Forgive me as im fairly new to this. When you say log the wideband, will that be with the logging software the WB came with or the numbers in TP since I have the WB connected to the PCM.
In my Open Loop AFR vs. Temperature vs. Vacuum table, above 44 Celsius I filled the table with 14.0:1. Side note, in the table since it says vacuum, does that also mean KPA? I see vacuum, kpa and map and it all confuses me a bit on if its all the same or not.

For the formula wideband_afr/X=multiplier. I assume the WB part is what the WB shows on the dash/log, and X=multiplier is the data from the table?
so if I i was running at 16:1 at a particular spot above 44C, the formula would be 16/14 to equal 1.143, if this is correct would I add that to the VE table in that block?

Forgive me if im all wrong, the last time I did the stuff by had was with my 7747 and winaldl to log. After the 7427 upgrade iv been using blm tables in TP and spreadsheets to calculate before my exhaust leak. I have redone my fuel system to output 20 psi which is why I need to redo my fueling then timing sounds like it will be a blast to do.

[steveo]

you'll need a log with rpm, map, and wideband all together.

I see vacuum, kpa and map and it all confuses me a bit on if its all the same or not.

kpa is pressure, atmospheric pressure is around 100kpa, vacuum is the lack of pressure, more vacuum is less pressure. an engine produces more intake manifold vacuum at lower loads.

you might want to google it, as these are not just tuning concepts but also basic engine concepts, so you should read up on how pressure/vacuum works as it relates to engines, as you are tuning a gigantic piston driven air pump, you are doomed without it.

so if I i was running at 16:1 at a particular spot above 44C, the formula would be 16/14 to equal 1.143

pretty much, but you will need to do some averaging or whatever at that 'particular spot' because your wideband likely will fluctuate in many ranges due to the fact that exhaust gasses and combustion aren't really a linear thing. tuning with single data points will give you a ve table that looks like bart simpson's head. that's why tunerpro has history tables (histograms). if you have a dozen data points in a particular region and you take their average, the data will become fairly reliable, but with one or two points, who knows.

if this is correct would I add that to the VE table in that block?

no, you'd multiply it.

[91dime]

Awesome, thank you,now to start seeing how it goes.

you say multiply it, what if it were at 13:1 when its supposed to be 14:1 does it still go 13/14=.929 then multiply that cell? or is this where you do a best estimate and start averaging out and smoothing the table?

And last question for now, does the stoich AFR parameter effect open loop at all or only closed loop? I cant seems to find much info on this.

[lionelhutz]

Be careful when playing with AFR numbers. Remember, the sensor is really measuring lambda or the ratio as a percentage and not AFR as a ratio. The sensor will read a lambda of 1 for pure fuel, E10 or even E85. You are applying a multiplier/math of some value to create an analog output from the sensor and then more math to display the AFR on both the gauge and in Tunerpro. So, anything displaying a ratio can be misleading due to the math conversions from lambda to an arbitrary AFR.

A narrowband O2 sensor works the same way. It doesn't switch at an AFR of 14.7:1 but rather at a lambda of 1.

Right or wrong, I would test the wideband by running the engine closed loop at a little higher rpm than idle (maybe double idle or so) and watch if it displays a lambda of 1. If it's setup to display as AFR then it should display the "ideal" AFR that it was programmed to display.

Say you set it to display 14.7:1 as the ideal ratio. Then, at 13.2:1 you are running at a lambda of 0.9 or 10% rich. Doing the math to get lambda means it doesn't matter if you're running E10 and the ideal AFR is really 14.2:1 or even E85 and it's the 8:1 or whatever that fuel is. The meter when programmed to the ideal ratio of 14.1:1 and displaying 13.2:1 means you're at a lambda of 0.9 which just happens to be a decent starting point for the PE mode.

The other part with the test I do is that if it's AFR is out just a little then I can call the value it does display the lambda = 1 value, assuming I'm confident the sensor and controls are working and the math is programmed/matching in the sensor and gauge. If I see a reading out say 0.2:1 on the high side then I'd likely call the lambda = 1 value 14.9 and use that in the math even though in theory the gauge should be displaying 14.7:1.

[91dime]

With all these numbers I think Im going to set my gauge to use lambda instead to make it simpler, well hopefully.

Side thought, if i log the lambda and its off 10% from what i want, would I increase that cell by 10%? or is it not that simple?