Enhancement suggestion for Lambda overlays

[Gernby]

I have a few suggestions that I think would make the lambda overlay more usefull. I suggested the first one a few months ago on S2KI, and "Hondata" responded that he would add it to the list. I figured I'd mention it again anyway.

1) Use the WOT Lambda adjustment low / high tables within the loaded calibration as the basis for the fuel change suggestions when in open loop. This would make the WOT Lambda tables much more usefull, since the tuners could simply set the values in those tables first to whatever they want to target, then adjust the Fuel low and high tables until the actual AFR's match the WOT Lambda tables. Once the fuel low and high tables are dialed in, it would be very easy to try different AFR targets.

2) Use the overrun manifold pressure tables from the current calibration to filter out data points from the datalog so that the fuel suggestions aren't artificially inflated by injector overrun cutoffs.

3) Use the Closed Loop criteria from the loaded calibration to filter out datapoints from the datalog that don't indicate the correct fuel status for that MAP and RPM. I believe this would greatly improve the suggestions in the high cam fuel map, since columns 1-7 often show huge negative fuel adjustment suggestions due to false rich AF values while fuel status is 1 or 4.

4) Use the VTEC engagement criteria from the loaded calibration to filter out datapoints that don't indicate the correct VTS value for the given MAP and RPM values.

5) Filter out datapoints that occur immediately before or after a change in VTEC state, or give an optional criteria for time before or after change in VTEC.

6) Use the TPlate and TPedal values to filter out datapoints where there is a significant mismatch between TPlate and TPedal indicating a change in state. Obviously, there are lots of conditions where TPedal and TPlate don't coincide at all, like at idle and when the cruise control is enabled, but that seems easy to overcome.

7) Add another overlay that shows the number of datapoints that were used to calculate the suggested fuel adjustment, or allow the user to enter a minimum number of qualifying datapoints in the Options window.

8) Add another overlay to show the deviation between the datapoints used to calculate the suggestions, or allow the user to enter the maximum deviation allowed for the datapoints that are used to calculate the suggestions.

9) Allow users to merge datalogs and eliminate sections of datalogs, so that they can create "superlogs" that are void of datapoints that they know will introduce error.

[Gernby]

I thought I just posted this one, since I forgot to add it in the OP, but it didn't show up...

10) Use the Closed Loop -> Lambda values from the current calibration to calculate the fuel suggestions for closed loop data points.

11) Add an option to import CSV datalogs so that users can apply custom filters or macros for "superlogs" that achieve specific tuning goals.

[phantom_]

wow, gernby, i'm impressed: the above shows that you've thoroughly analyzed numeric datalogs...

there have been several discussions whether to use external wideband and/or log external wideband data like Innovate LC-1.

regarding your statements, if Hondata implements the irrelevant data-filter-out rules, does this mean that the logged data will be much more reliable than any other wideband log?

[Dan1987]

It doesn't mean that the values will be more accurate than a wideband...only a wideband can do that..it means the suggestions will be more consistant with what fuel changes are actually neccesary, rather than flashpro manager calculating a suggestion from a selection of datapoints which have a large deviation for example.

By the way I really like the suggestion of using the WOT lambda tables for the suggested fuel changes and the one about using MAP to determine overrun so that suggestions in those areas of the overlay aren't visible, that would be brilliant.

[Gernby]

Since most of the suggestions above came from a bad eTune from Evan's Tuning, I figured I should post a link to the thread I just created about that eTune.

http://www.s2ki.com/s2000/topic/904368- ... ans-etune/

[Gernby]

I should mention that the issues with part throttle tuning on the high cam are what lead to this thread. Basically, there was too much reliance on the fuel adjustment suggestions in FPM, which lead to AFR's in the 16+ range at part throttle with S.Trim at +35 while in VTEC. Once the ECU went to open loop, those AFR's would shoot up into the 18's!