EFILive MAF Worksheet

[SSpdDmon]

Since EFILive works a little different than HPTuners, I decided to create an Excel sheet to calibrate the MAF in EFILive. Given we can't see the raw data (as far as I know we can't - the export feature didn't work on mine), I set it up to use the weighted average of up to four different logs. You would have to log DYNAIR_DMA (grams/sec) and MAF Freq (Hz) in SD, plug them into the sheet and it would generate a new table for you. Would anyone here be interested in it? I'd need a host...

[Tordne]

Someone (can't find the post again to give credit) posted a MAP that may help you with this.

I have attached it here for you, and others.

Basically you put you car into SD mode, with the MAF still connected, just set it to fail, set {C2901} to 0. Also set {P0101}, {P0102} & {P0103} to "No MIL". Then go for a drive and try and hit a number of MAF HZ.

I believe you should then be able to go to this MAP and past it back into your MAF calibration. You will likely need to do this a few times...

Cheers

[SSpdDmon]

Yes, that's the MAP I use to gather the data from a logged run. But, that's just a part of it. What my spreadsheet does is take that data (plus up to four other logs) and creates a weighted average for the values returned based on the cell counts (n). This cuts down on the time it takes to log data, paste it into a tune, and repeat. It also only applies a 66% correction factor to avoid overshooting your target. Finally, it creates a simulated value for cells you did not hit based on how far the nearby cells differ from stock. For example, if you missed hitting 9000Hz enough times (set at a minimum of 10 currently), it would take the average of the difference between 8875Hz & stock and 8750Hz & stock to get you in the ballpark. It's nothing crazy. Just a tool to help make things easier.


**NOTE** If you try the calculator, you'll get a couple messages that you need to disregard. The first is a message about a reference to another workbook. Just click no. The other is when you delete the data out of the light-blue fields. It says something about a circular reference. Just click cancel. The required fields will be filled when you paste in your new data.

[Tordne]

That certainly does sound comprehensive!!! Will have to check it out.

[dbaxter_ss]

whats the 66%?

[Black02SS]

I haven't used the MAF worksheet, but I just finished up a car that needed to retain his MAF. I used the MAF MAP posted above and had no troubles at all. I calibrated it just as I would the VE table with BEN. I used the VE table to hit all the cells and paid no attention to the MAF Frequency table at all. After the second pass, all were aligned where I wanted them.

[SSpdDmon]

whats the 66%?

To avoid overshooting your target. If you keep splitting hairs by 66%, you shouldn't go over. But, the changes will eventually become so small that what you end up trying to control is the variance in your measurements.

[Black02SS]

As I am not putting down your spreadsheet I wonder why would you still use this when you can make a map that will do it all and never have to open excel? I think it is great that we have many options, but if we can utilize the program, why not? Just curious.

[SSpdDmon]

As I am not putting down your spreadsheet I wonder why would you still use this when you can make a map that will do it all and never have to open excel? I think it is great that we have many options, but if we can utilize the program, why not? Just curious.

Because, in the MAP, you won't hit all of the cells or you may not hit the cells you want enough. This takes a weighted average of the ones you do hit in 4 logs. That means if you hit a given cell only 15 times in one log, you have 60 data points to average and not just 15 like you would with the MAP. Also, you can see the trend in the variance from stock. If you recorded values for 10 cells that show roughly a 5% difference from stock and one cell is stuck in the middle that shows a 10% difference, you can make the appropriate adjustments. This can also take that same principle and simulate cell values for areas you did not hit in your logs. It's not the greatest creation when it comes to MAF worksheets, but I feel it does a better job than just the MAPs.

[Black02SS]

Here are my thoughts on this and please, correct/educate me if my theory is wrong. The way I use the MAF Map is as follows. I take my VE Histogram and plot as many cells as possible just as if I was calibrating my VE table with BEN. I figure if I hit all of the MAP values and RPMs just the same as if I was tuning the VE table, then there should be no issues using the MAF Map. I am making sure the every cell for the VE table is accounted for before I copy the MAF MAP % into the tune. In my theory, this eliminates any errors that that may pose seeing how every map/rpm value is accounted for. :?:

[SSpdDmon]

Here are my thoughts on this and please, correct/educate me if my theory is wrong. The way I use the MAF Map is as follows. I take my VE Histogram and plot as many cells as possible just as if I was calibrating my VE table with BEN. I figure if I hit all of the MAP values and RPMs just the same as if I was tuning the VE table, then there should be no issues using the MAF Map. I am making sure the every cell for the VE table is accounted for before I copy the MAF MAP % into the tune. In my theory, this eliminates any errors that that may pose seeing how every map/rpm value is accounted for. :?:

Ok...this to me says the range of data you're gathering is great. If you are able to hit every cell in the MAP, that's awesome. But, the question is how many times did you hit the cells? I've found that many cells only get about 20 or fewer readings even during an extensive log. If you use the average (x) values in the map, it's taking the average of the 20 data points reported during your log. Say you have 3 other logs for the same cell with 50, 30, and 15 data points. You can take just one of the numbers and say that's good enough or you can use all of the data. Let's say for instance these were the readings from the same cell in 4 of your logs:

x = 35.675 g/s n = 20
x = 37.259 g/s n = 50
x = 36.914 g/s n = 30
x = 37.756 g/s n = 15

The variations above don't seem that great, but they can definitely have an effect on the calculations the computer is making. Taking just one of those numbers may get you close or it may not. What the worksheet does is applies weight to the averages. There are 115 total data points. So, why not use all of them? If you did, you'd find the cell value based on all of the data points is 36.958 g/s. Just using the MAP you would have stumbled upon the closest answer by the third log. But, by the 4th one you would have been doing more damage than good by changing it to what the MAP says. What it boils down to is the MAP, on its own for an individual log, doesn't account for prior readings. Therefore, using a log that has 500 hits for a cell, making changes, then using a log that has 50 hits for that same cell, and making changes means you value the average of the 50 just the same as the 500 hits. Sorry for being redundant here. I just want to make sure my perspective is understood.