keeping BENs consistant

[mr.prick]

how do you keep the BENs inline?
everytime i make a change it usually knocks them richer or leaner
but never perfect.
i have gotten close but it seems with altitude, and ECT and, IAT changes
my AFR fluctuates fairly wildly.
i don`t have my commanded AFR table set to what is used in the tutorial,
and i will start off right on or a bit rich then after awhile
it leans out or fattens up.
is there a parameter that controls fuel with altitude or
barometric pressure changes?
or am i just trying to hard?

[TAQuickness]

Most of us have noticed the BENs will drift with the weather. COS3 & 5 have some nifty tables to help keep things in line (i.e. A0008, & A0014).

[SSpdDmon]

This is part of the reason why I relocated my IAT. I honestly believe there's a legitimate problem with false IAT readings due to heatsoaked components around the sensor.

The other side of it is the lack of resolution in the VE table. I've noticed I can get the BEN's virtually perfect when the engine is cruising along at a steady 1600RPM or 2000RPM. But when the conditions fall between the cells, the PCM has to interpolate the values from above and below to 'guess-timate' what the VE is (i.e. 1800RPM). I assume this interpolation is linear whereas the VE is anything but.

[mr.prick]

in OLSD i try to keep my driving to times when it is not to warm outside ,
IAT average in the 80`s and ect in the 180-190F.
i have also tried turning off {B0109} Blend IAT With ECT.
but did not see much change.

[redhardsupra]

bringing up an old thread, because i think it's a relevant to place my new research here:
Temperature Modeling (http://redhardsupra.blogspot.com/2007/10/temperature-modeling.html)

[stigmundfreud]

I've been noticing this a lot lately. When I got things dialed in last year the weather was a constant 16c with a fairly constant barometric reading. Recently its been all over the place so not been as able to get tuned in. I guess I'm just going to have to do a long night keeping things constant but it is a pain.

I notice someone mentioned COS3 or 5, its bloody tempting to go to COS3 for this reason but then there is the extra tables to fine tune and its a question of am I too lazy or not!

[Gelf VXR]

I was under the impression that EfiLive and the PCM calculations take temperature at atmosperic pressure into account to arrive at the GMVE or VE?

In the tune tool under properties for VE efficiency values

display VE = default units (g*K/kpa)

[redhardsupra]

the PCM does account for temperature and pressure, as long as they're properly measured/estimated. you start changing the thermal properties of the whole system, and the parameters of temp estimators need to get recalibrated. until they're redone, your temp readings are going to be off and the tune is going to be inconsistent across various temperatures.

[BlackGMC]

I would try filtering on your average IAT temp and a range of ECT temp.... I have noticed that when I first go out logging the first 5 mins of the log are crap due to heat soak from the IAT sensor, after the first 5 mins things line up pretty good.... I would also check out what is affecting your final spark value... I noticed that Spark ECT Table (5910) was always pulling timing when the motor reached normal operating temp, if it was just a few degrees below it would not pull timing.... Also check out your IAT spark table (B5911), mine was pulling timing when I first started to long due to the high IAT temperatures then after a little bit of driving it would stop pulling timing........ Filter on that stuff will keep everything consistent....

[dfe1]

Part of this issue is the fact than an engine is an analognot a digital device. As has been mentioned, there's also a good bit of interpolation going on as rpm, MAP and temperatures move around. An engine may do something different from one minute to the next, even though all relevant variables appear to be identical. I've been running open loop speed density for over a year and BENs are all within 1-2% most of the time. But I've done back to back logs with dramatically different results on some occasions, and with nearly identical results on others. You also have to consider that sensor output may not be consistent enough to produce identical results when you're looking at things under a magnifying glass like an EFILive data log. There IS a reason that car manufacturers go to the expense of using mass air sensors and implementing closed loop systems.

[2002_z28_six_speed]

how do you keep the BENs inline?
everytime i make a change it usually knocks them richer or leaner
but never perfect.
i have gotten close but it seems with altitude, and ECT and, IAT changes
my AFR fluctuates fairly wildly.
i don`t have my commanded AFR table set to what is used in the tutorial,
and i will start off right on or a bit rich then after awhile
it leans out or fattens up.
is there a parameter that controls fuel with altitude or
barometric pressure changes?
or am i just trying to hard?


That is your limit of resolution. You are overshooting the real result if you are trying to tune multiple times in a short period of time and it does this.

Try the weighted average sheet where it takes the absolute value of ur ben and multiplies it to lessen the correction. aka damping the response

[hquick]

OK....yeah..another thread revival.
Has anyone managed to solve, partially solve, figure out a way to minimize the rich lean issues described in these posts?
I have exactly the issues mentioned.

[macca_779]

Howard here is how I do it. Basically if I was to start from scratch I would do a normal as advertised AUTO VE in the early hours where it is nice and cold. Take note of the IAT your conducting this at.

Now utilising alot of data and trying to hit at least 50 counts in the cells you need, I would then apply this to the VE table and don't touch it from then on. The important aspect of my method is to pick conditions where ambient temp is consistent. Now the next thing to do is the same kind of logging but you won't need quite as much data, but this time do it in the warmest part of the day. Now from the new data acquired take all the BEN's over to Excel and average them into one value. Use this one BEN and apply it to the temp you logged at in the second session in A0014. From that linear fill down to the first logs temp. If you have trouble getting a consistant IAT, average them too and apply accordingly.

I have had good success using this method in the Territory where temp swing can be quite large from morning to afternoon.

[hquick]

Thanks Macca.....I'll have to read and digest what you've just suggested. I can understand that if this is working for you...it should work for most. Temps don't vary anywhere else, near as much as there in the NT.
Thanks...I'll try it.