Narrowband targeting 12.5 afr.

[bheckman]

I have a 92 4.3 that I have have been trying to tune. The data logs I read say that my blm is 128 while reading 12.5 afr on my wideband. During highway cruising my narrow and drops mV and will raise the blm all the way to 160. The wideband will read all the way down to 9.5 afr at a cruise. Will adjusting my fuel tables even help if I am running rich but my narrow band thinks it is lean?

[dave w]

In my humble opinion, AFR tuning is more accurate.

I've often noticed the AFR's and BLM will contradict one another, showing lean AFR with rich BLM or vise / versa. In my humble opinion, adjusting the fuel table to the AFR's is a more optimal way of tuning. Parameters for choke, acceleration enrichment, and power enrichment require AFR tuning because the computer will lock the BLM's at 128.

I posted information on AFR tuning here: http://www.gearhead-efi.com/Fuel-Injection/showthread.php?5381-WBO2-Tuning-Spreadsheet

dave w

[WASyL]

OL with AFR tuning is far more superior then just CL (BLM/INT in this case) which just targets 14.7 (exept WOT). if BLM says 128 and AFR says 12.5 then by best bet is that o2 is shot or ECM is shot.

best regards

[steveo]

another issue may be your wideband calibration. are you actually looking at a gauge and your laptop while driving, or are you logging the wideband AFR somehow? if so, are you sure your calibration is correct? ECM inputs, if you're using one, are sometimes horribly non-linear.

another issue may be the analysis method.

are you doing? a histogram average of BLMs while ensuring the closed loop flag is set 100% of that time, and comparing it to wideband AFR in a similar histogram? or are you just doing a point-by-point analysis

the sensors both jump all over the damn place so just comparing a few data points may be misleading.

the wideband reading is not actually an AFR. you have, of course put in some math regarding your fuel type, since the wideband has no way to know that, right? if you are running ethanol blend with your stoich set to 14.7:1, your wideband "AFR" is meaningless.

of course the BLM trim isn't an AFR either, its actually the amount of fuel added/subtracted to attain a stoich AFR, whatever that is. 128 means 'no adjustment was made' which could actually mean no adjustment was necessary, or no adjustment was allowed. 128 trims are pretty rare, so if it's just sitting at 128 all the time, i usually suspect it's not trimming. ensure during all conditions you are monitoring, the o2 is rapidly darting back and forth. if it is not, then your trimming is not complete and the log entries are meaningless.

not saying you don't understand this stuff but just wanting you to ask yourself some questions that may lead to a solution.

[bgott]

The phenomenon you are describing sounds to me like you may be entering PE. I would take a look at the log and see if that flag or another modifier is enabled. From observation of my own logs when you enter PE mode it effectively ignores the fuel trim and locks the BLM at 128. IIRC some people will actually tune out PE (set the temperature or TPS to enter at an extremely high value) to tune some of the heavier load parts of the table because the tendency of the ECU to want to enter PE when under high load.

If this is happening you can actually have sections of your fueling table (roughly from 75-100kPa) that will be completely wrong but will be "hidden" in the history average in TunerPro because it doesn't filter data. This is a problem I have with the classic method of BLM tuning these ECUs using that history average. If you disable PE this average would look closer to the filtered data you get from other means. This is an advantage of the sheet from dave and also the software from steveo as both of there methods filter the data so that the averages only include data from points where PE is inactive if I am not mistaken.

Just my $.02

[bheckman]

Thanks everyone for the advice. The problem turned out to be a bad O2 sensor. A new one fixed the issue I was having.