Baro+airtemp compensented load is the load that is invariant between MAF-based and SD-based. Its also what's used for fuel, spark, and MIVEC table lookups. Because of this, it seems to me that this is the best load to log under MAF control when attempting to generate data for the SD VE tables.

 

If you do that though, you'd need to make MAT compensated MAP value to calculate VE.

So VE= Load/(MAP*(298/(273+MAT)))

I'd rather keep it simpler, but the above is technically the most correct?

 

The BARO+IAT makes sense as far as being a closer representation to SD estimated Airflow. I thought fuel lookup was done with uncompensated load and timing lookup was done with Baro_IAT compensated load though?

JCSBanks, it's pretty cold here (-5C or so) I'm wondering if extreme temps may cause the same issue that extreme baro does in trying to set these tables up? This is from the same log as above, but using Baro only compensated load. Below 40kPa and then from 0-2000 RPM and 5000-8000 RPM I don't have values because the standard deviation was very high and I feel is unreliable data. The Baro+IAT compensated load however of the same log provided very tight standard deviations. Just to put numbers to the data, for the Baro+IAT comp, I was able to use the numbers that had standard devations of under 0.03 and more then 30 samples. For the baro only compensated load, I was lucky to get the standard deviation under 0.10 and more then 10 samples on a lot of the cells and above a STD of 0.10 I just didn't use the values (hense the missing values in the tables below).

40kPa - 0.84
50kPa - 0.84
60kPa - 0.85
70kPa - 0.88
80kPa - 0.93
90kPa - 0.99
100kPa - 0.99
110kPa - 1.03
120kPa - 1.03
130kPa - 1.04
140kPa - 0.98
150kPa - 0.96
160kPa - 0.96
170kPa - 0.96
180kPa - 0.96
190kPa - 0.96
200kPa - 0.95
210kPa - 0.93


RPM VE:
2000 - 0.99
3000 - 0.99
4000 - 1.00
5000 - 1.00

I just tossed 0.80 in randomly at both of those extremes on the RPM VE in the previous chart. I've got stock cams (Evo VIII non-MIVEC) so I figured it fell off pretty hard up top. I'l bump them both up, as the rich down low will probably help with starting and catching a quickly falling engine speed to prevent stalling.

 

ok, I see the point about calculating an airtemp-compensated MAP value for the plots to generate VE data. Its more technically challenging to do it this way, but would lead to the correct VE values with less guess-work? Can't the temp compensation be done in EvoScan, e.g.,

<DataListItem DataLog="Y" Color="" Display="MAP kPa" LogReference="MAPkPa" RequestID="38" Eval="1.334*x" Unit="kPa" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="500" ChartMin="0" ChartMax="500" ScalingFactor="1" Notes="" Priority="1" Visible="False" />

<DataListItem DataLog="Y" Color="" Display="MAF or MAT Air Temp Centigrade" LogReference="AirTempCentigrade" RequestID="11" Eval="x-40" Unit="deg C" MetricEval="x-40" MetricUnit="deg C" ResponseBytes="1" GaugeMin="-40" GaugeMax="150" ChartMin="-40" ChartMax="150" ScalingFactor="1" Notes="" Priority="1" Visible="False" />

<DataListItem DataLog="Y" Color="" Display="Temp Comp MAP kPa" LogReference="MAPTempCompkPa" RequestID="CALC" Eval="MAPkPa*(298/(273+AirTempCentigrade)" Unit="kPa" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="500" ChartMin="0" ChartMax="500" ScalingFactor="1" Notes="" Priority="1" Visible="False" />

 

Looks good mrfred, once it is done it isn't hard to use, so we may as well do it right.

03white, thanks for refining this in a relatively extreme environment. Yes I think you're right, but...

mrfred and 03white, if we're using air temp compensated load then we need MAT compensated MAP recorded at the same time... will some users have both sensors present at once?

If they don't have both sensors I think they're best comparing baro only compensated load with MAP, since the difference between IAT and MAT (which is the error introduced if you apply no temp comp at all) is much less than the 10% errors you'll get in -5C compared with 25C when you compensated load but not MAP.

 

I checked again on what load variables are used:

spark advance lookup: For air temp below 77F, baro+airtemp compensated load is used for spark advance. For temps above 77F, then baro compensated load is used.

afr lookup: for closed loop conditions when load is < ~20, uncompensated load is used, otherwise, baro+airtemp compensated load is used. This means that baro+airtemp compensated load is used essentially all the time for AFR lookup.

 

Mrfred, thanks for the explanation on which loads are used under which conditions.

I will definitely leave the MAF in long enough to get Baro+IAT Compensated load vs. MAT compensated MAP. If nothing else, hopefully it will put some peace of mind into the best way to set this up.

With my first round of data, it looks like having the Baro+IAT compensated load is having a huge impact on how reliable the data is. I'll post up the comparison in the next day or two to show the differences on the same log. I'll grab a couple more logs tonight too for extra data points.

Would temporarily putting the IAT in the UICP be a waste of time for figuring this out? I could put the sensor in the UICP pretty easily in the next day or two. To get it in the manifold is going to be a week or two out.

 

So i got a log on the way home of the airtemp/baro compensated load that mrfred posted above. What is the equation i should be using when tracing the log? Note: the airtemp comp load is using the maf IAT but the log also includes the MAT temps.

or

 

If you have logs with MAT and want to be very accurate and complex, then the first option is most accurate. If you want quick and dirty, especially if you are using logs and tuning at near to standard temperatures and pressures (but it will still compensate in other conditions) you can do the second option. Second option has worked fine for me, it wouldn't work for 03whitegsr for his initial setup too well.

03white, if you final position of your temperature sensor is going to be UICP, then use this temperature to compensate your MAP when comparing with your baro+IAT compensated load. If UICP is just interim, it will still probably get you near.

For some people that are going to completely rescale their load (eg strokers, cams, big turbos, weird MAF setups etc), then the IPW and timing at various MAP vs RPM levels are probably going to be the most useful, then they can scale their injectors and VE tables to suit a more rational/neat arrangement than the mess of loads you can get on extreme setups with MAF sensors.

 

Temps were above average for the winter but will be more common in about a month so i will use them. I will post the traces and my tables this weekend when i can dig through them a bit. Thanks for your help guys.

 

This is exactly the route I will be going, as I will be building a new fueling and timing map from scratch.

2.3, with 280's, and a 67mm. This should be interesting to try and tune with your SD patch.

 

http://web.archive.org/web/200412091.../sae940759.htm This is the paper from Moskwa that I was talking about earlier re transients between SD and MAF... there is a pdf download as well. Only for tech head interest!

essevo, GotWheelHop, grayw, this 1:1 method of load:MAP would benefit from IPW numbers at your desired WBO2 AFR in midrange on full boost, so I can form a calculation of what injector size you need to start with to get say 12 AFR if you put 12 in your fuel map and run the VE at 100% so you get 1:1 load:MAP. Essevo, can you let me have this when available?

 

Jcs can you mod 88580014 with last spec? I wish to test it on my EVO IX RS (full JDM Spec.).

Thank you

 

See post #92. It has 8 bit MAP scaling as in the original version, but this is fine, and the benefit is tephra at least had it running.

 

Should MAT be in C or F? I assume this matters.

Haha, also, i am assuming the above equation needs a 100* at the beginning?