it is rounded to 1decimal place.. 0.05 up and 0.049 down... just normal mathematics. There doesn't appear to be any issues really raised here. If you like, you can present some extraordinary figures and I'll take a look. All the feedback I've had to date is that the Map Tracer calculations are correct so far.

 

Actually I just realised something that you guys might not understand... You need to know that the ecu interpolates the data between load and rpm points in the map. to make an informed decision on what the AFR is at a particular point, you will need to manually calculate what the interpolated target ecu value was at the actual rpm/load point and match that against the same logged Load/RPM AFR value.

 

for example one cell having a range of Load 220-240 and rpm 5000-5500 is a Huge Range! you don't really have the same AFR at 220load/5000rpm as you can at 235load/5400rpm. So even min, max, or averages could be very very skewed to one side of the cell range displayed.. you really really should hover your mouse over each cell and review all the data in each cell. I just had a great idea that I should interpolate the Target AFR from the loaded ROM for you, and display it next to the appropriate Load/RPM in the tooltip against each logged point.

 

The most accurate display method would be to decide on the nearest cell depending on load and RPM. So the 5000 RPM zone would include 4750-5249RPM. 240 load would include 235-244.

Is this what the Evoscan map tracer is doing?

I was going to do this but with uneven spacing between RPM and load zones the code/calculation was getting complicated in my small brain. It would need each logged load or RPM value to be looked up in the load or RPM axis.

Perhaps we could use:

abs(loggedRPM-RPMzone)

Iterate through the list of RPMzones and then pick the cell with the lowest value - so giving the nearest cell.

Is this already covered Hamish or are you doing it some other way?

 

It appears that when you hover over a cell, all of the incremental data is in numerical order, not the order it was logged. If I knew that the AFR went from 11.3 at the first reading in the cell and ended at 11.1 at the end of the cell, I could make a more educated decision on how to make my adjustments accordingly. Even better, if the hovering window showed both the RPM and AFR (or timing) side-by-side increasing in RPM down the list, that would make life a lot simpler.

 

This is what I originally thought too, that the Ecu is doing. A 5000rpm cell would interpolate vertically from 4750-5250rpms, and also have some influence from the horizontal loads nearby.

 

I believe it does too. Bilinear interpolation in other words.

 

does anyone think that rescaling the map so that we have more detailed scaling in the 220 to 300 load range would be a bad thing ? would there be anything wrong with scaling the first 10 - 100 every 20 instead of every 10 and using the places freed up and put them in the higher load cells ?

 

I saw a writeup by mellon about that a while back. I'll try to find it.

 

If you just did WOT pulls all the time, or made a map strictly for the track, I would say its fine but we spend 90% of the time in the lower cells so you would be losing resolution in the area you spend the most time in. The larger the gap is between cells, the greater the chance of knock and irregular results every time those load cells are crossed. You are asking the ECU to interpolate a very wide range which is asking the ECU to do a lot of calculating in a small period of time.

 

How many logs can you compile in the map tracer? Is it unlimited? Or is there a limit?

 

At least you can get your Maptracer to work. I can get the sample logs to work, but can't get any of my previous logs.

evo4mad, I sent you an email per another thread and have had no response.

Anyone have any thoughts?

 

I like this idea. Would you also incorporate this bilinear interpolation in future upgrades?