Ok im sure we are all aware that the load calculated is not accurate.

now from my experience the load displayed in evoscan tends to be 15-20 load% higher then the actual cell the ECU follows. ( based on ignition timing)

so what i want to know when you guys tune do you just change the cells that are exactly displayed or do you some how compensate? do you highlight the adjacent cells as well? and do 3 cells at the same time

example of what im asking

evo scan load at a given rpm 4500 rpm says your at 240 % your picking up some knock in the 4500 rpm range and your not comfortable with it

when you go to pull a degree back do you higlight the 220 240 260 load cells at 4500 rpm and pull a degree from all cells, or do you swear on evoscan load calc and only change that one load cell being 4500 rpms 240%.

 

The load calc, like your mentioned, may be off. It is just a guide to let you know the general area for the cells that you have to adjust.

If you want to know what cell your are in exactly, just use the rule that 1* of timing will be pulled for every 3 counts of knock. Then all you need is your RPM and ignition timing to know what load cell you are in.

Or, if you want to know the exact load number, without a doubt, read up on the two-byte load logging. You can log the actual load number that is being used by the ECU.

As far as which cells to change, you should generally always change an area of cells, depending on what the neighboring cells look like. This is especially true for your ignition map. Rapid changes in ignition timing can cause the ECU to think it is knock and it may pull timing. You always want your ignition timing cells as smooth as possible from transitions from one cell to the next. For example, you don't want one cell at 5 and then one right next to it at 12.


Eric

 

first off, you should never be tuning one cell at a time. you need to know that when your evoscan says 230 load calc, that on the actual table, it is using both the 220 cell and the 240 cell and interpolating. it also interpolates up or down based off of rpm. so yes, you should be adjusting the adjacent cells as well when you see knock. the adjacent cells may not have to be adjusted as much (and in some cases not at all), but you need to take them into consideration.

 

This guy is right, and then you smooth it out so there are no big jumps in timing. It should be smooth enough to where if you were to put it in 2d or 3d graph you will see a smooth transition in timing.

 

After a while you get a pattern for the amount of taper you need on the timing for each 20 increase in load. You can tell this by running in different gears and at different boost levels and datalogging. Depending on how it is set you might have LESS knock at the higher boost/load levels because the timing is very conservative there. The load calc was only ever a rough estimate, you can log the real load if you really want, but looking at the timing works just as well practically.

Once I've set the boost level I want I tend to adjust the timing all around a problem area, smoothing in the steps that might have caused the knock as I go. Subtle changes to the ramp rate of timing as you increase the revs can help - eg if you have a problem at 6000 RPM, then don't ramp much over the 5000-6000 RPM cells and then bring it in more at 6500 RPM if it will take it. When it is nice you should find that the last degree of adjustment makes it all silkly smooth and predictable in every driving situation. Takes a lot of datalogging and testing to get a powerful and smooth knock free map. I always work from conservative upwards. Add a degree, then try to provoke it in a variety of gears and then read the logs and go back and improve it.

Same as what the other posters are saying really.