Slight correction, "enable smart tune" does not change the ve tables, the closer your initial ve tablesare to the "truth", the fewer smart tune runs you need. Afaik, the calculation the "smart tuning" does is simply to do a statistical analysis on looking through the sample data comparing how many percent the o2 sensor (integrator) output differs from stoich for each cell on average and the new cell value is calculated from the original cell value based on that percentage. As the sensors are narrowband, that analysis is not exact but it in general corrects the ve's to the right direction. And now the new sw version can even show you real time when there are enough samples for a cell. So you update the ve's and do another smart tune run. And again. As said in the manual, once your smart tune changes the ve's less than 5%, you call it good. And after you get the ve's to the right ballpark, you can start to optimize the afr table to your liking. But there is so much going on in the engine that the delphi does not measure nor model so do not expect it to be exact science.

 

Right. And it would be nice to know what that "sample data" is exactly and how it's used to suggest a change in VE table cell values.

 

My guess would be that it's a pretty simple formula. I know on the dyno 1 whole VE number equals approximately .1 A/F. For instance if I have a value of 92.5 in a certain VE cell and I want to richen up the A/F in that cell by .2 (example I read 14.2 A/F and I want 14.0 in that cell) I would change the VE value to 94.5. So that would tell me that Smarttune just does the math.

 

Inputs:

cubic feet per minute (CFM)
cubic inch displacement (CID)
revolutions per minute (RPM)

The most common and effective way to enhance volumetric efficiency in an engine is to increase the size or number of valves in the engine. While larger valves increase flow, they can have the negative effect of causing the engine to weigh more. Using multiple valves can increase the flow of an engine without adversely affecting the weight. By streamlining ports, a technique known as "porting," one can increase flow capability and, by proxy, the volumetric efficiency of the engine.

 

"ball park". thats about right

 

I see the values in the VE tables as representing the actual true VE's. They may not be super accurate from the factory, or after me making mods. But the basis for good tuning depends on them being true. The ECU relies on that as well. That's why I use the AFR table for A/F adjustments in open loop. Which brings me back to why I want to know more about how Smart Tune creates them.

 

The AFR table controls both cylinders, when you tune the VE tables you are basically calibrating them. When I tune a bike on a dyno I set the AFR tables to 13.0 . Then I do a set up runs sampling the A/F at eavery throttle position (or kpa on newer bikes). Then I adjust the VE's to scheive 13.0 . I do this for each TP (or Map kpa) and I do this for each cylinder (front and rear). When I am finished the VE's are now calibrated to the build and I can set the AFR tables to where I actually want it to be. SMarttune essentially does the same thing, only it is limited by the scope of the narrow band o2s. It has to set the AFR's to 14.6 because that is the setting it takes to turn the o2's on. After you smarttune and "calibrate" your bike to the 14.6 you can adjust your AFR tables to where you want them and put the MAP in the bike without the enable smarttune box checked. Similar to what we do on the dyno.

 

harleytuner:.
out of curiosity, when you tune with the samplers, does the software give you feedback of AFR numbers as you see the VE's change, or is it a different method.

 

I suspect that Smart Tune does not rely on O2 sensor voltages to calculate new VE's. Their accuracy just isn't close enough. I'm going to try my own experiment. I suspect, and I may be way off, but I think it's entirely possible that Smart Tune does this for each cell that it has the data for:

First, it sums up the fuel pulses and calculates the number of grams of fuel used in that time slice using information stored with the calibration. The data has to be massaged a little but it's all there.

Next, since it was in closed loop toggeling fuel around a fixed 450mv (14.68 AFR for gasoline) it can compute the total number of grams of air that was most likely used to burn that fuel in that time slice. That's really the actual air flow number in grams. That number should be pretty close. If using E10 fuel the stoic is about 14.1 to get to 450mv. So that calculation might be off slightly.

Now, Smart Tune can also compute the theoretical air flow in grams for that time slice because all the information is available in the downloaded data to do so. The formula is easily available.

So, in conclusion, calculating the ratio of actual to theoretical like I just described will yield a VE value. Is that what Smart Tune is using for VE NEW.

 

Nope . The ve is just a number . Just a scale to use.
Your making it to hard now.
If your ve's never come in , or they just keep changing on you and your chasing your tail? Look at your sample quality ,are the bungs proper? Is there times the sampling just stops? Does the bike do ok for a few days , then start unraveling?
When you do another smart tune run is it always making some good size ve corrections? Most bikes will not be capable of running closed loop everywhere the system lets you run it., and still be capable of feeding good data back to the ECM. This not only effects closed loop , but open loop as well.