How-to: Rescale Load Axis (percent) Values in Evo VIII ECU
Evolved Member
Joined: Jan 2006
Posts: 953
Likes: 2
From: Oxfordshire
I have a Buschur hard MAF pipe and an open element Buschur filter and am seeing my calculated loads in EvoScan go past 260 consistently from the high 3000 rpm range to around 6000 rpm at 20 psi. This surprises me. I thought that with an open filter the MAF reads low compared to what the engine sees, and therefore I would definitely not have a rescaling issue. Yet it seems like I do. I don't have a crazy setup. I haven't dynoed recently, but I'd expect to put down around 330-340 on a Dynojet if I were guessing.
Am I not understanding properly how the MAF pipe and open filter affect the load readings. I would have thought that lower MAF readings would translate to lower calculated loads and more headroom on load ranges, but it seems like the opposite is happening?
I've done a lot of searching and still don't quite understand what's happening.
Thanks,
RichJ
Am I not understanding properly how the MAF pipe and open filter affect the load readings. I would have thought that lower MAF readings would translate to lower calculated loads and more headroom on load ranges, but it seems like the opposite is happening?
I've done a lot of searching and still don't quite understand what's happening.
Thanks,
RichJ
Changes in the inlet, greatly affect the way the MAF reads the airflow. The Evo MAF uses the Karmen vortex method. What this means is the airflow is straightened, Thats what the honeycomb does, and then it is disterbed and the little eddies (vortices) are measured. The more eddies, the more air your flowing.
The problem is if you change part of the inlet, you can get more eddies for the same amount of air as before.
You may also be flowing a little more, so your MAF readings should go up a bit. Calculated load always read a bit high. I think the equation uses a *5 constant, I worked back from my timeing cell and found my constant to be nearer 4.5.
MB
Hi
Changes in the inlet, greatly affect the way the MAF reads the airflow. The Evo MAF uses the Karmen vortex method. What this means is the airflow is straightened, Thats what the honeycomb does, and then it is disterbed and the little eddies (vortices) are measured. The more eddies, the more air your flowing.
The problem is if you change part of the inlet, you can get more eddies for the same amount of air as before.
You may also be flowing a little more, so your MAF readings should go up a bit. Calculated load always read a bit high. I think the equation uses a *5 constant, I worked back from my timeing cell and found my constant to be nearer 4.5.
MB
Changes in the inlet, greatly affect the way the MAF reads the airflow. The Evo MAF uses the Karmen vortex method. What this means is the airflow is straightened, Thats what the honeycomb does, and then it is disterbed and the little eddies (vortices) are measured. The more eddies, the more air your flowing.
The problem is if you change part of the inlet, you can get more eddies for the same amount of air as before.
You may also be flowing a little more, so your MAF readings should go up a bit. Calculated load always read a bit high. I think the equation uses a *5 constant, I worked back from my timeing cell and found my constant to be nearer 4.5.
MB
Does that make sense?
I do have 660 injectors, but my mail-in flash didn't use a different injector constant and I haven't changed the EvoScan data.xml file to change the base load calculation. Could the injectors be affecting the load calculation nonetheless?
Thanks,
RichJ
That doesn't seem right at all. Is this the calculated load? You can find out what load for sure by logging the two-byte load. I'm guessing that you haev different sized injectors or latency values and they weren't compensated for in the calc for load. You just can't hit 300 load at 14 psi.
Eric
Eric
This is going to sound crazy, but when I did the block cut and pastes to do the rescaling, I found that the cut and paste wasn't 100% accurate. Some cells got changed in the block cut and paste. I'm glad I did a cell by cell proofing to fix them. I have no idea why this would happen. Similarly, if you select a block of AFR cells in a fuel table and advance them all at once, they don't all go in sync. Some advance 0.2 at a time and some don't move at all with the increment (+) button unless you hit it twice. Bizarre. Maybe this has happened to others and I just haven't read about it, but EcuFlash doesn't seem to be as reliable with manipulating data as, say, Excel.
I'm not a newbie with spreadsheets. I'm adept with Excel, Access, etc. This is just weird.
I'm not a newbie with spreadsheets. I'm adept with Excel, Access, etc. This is just weird.
ECUFlash is a little wierd, its not really a problem with the program, but how it calculates those values you have put into the table, because there is some "rounding" error, you'll find that some values don't increment immediately and others do, if you were to look at the hex values instead, you'd see it more clearly.
Cutting and pasting is a similar thing, sometimes pasting into the cells causes the values to become rounded up or down(especially if you copied it into Excel then back into ECUFlash) if there are more than 2 decimal places.
Cutting and pasting is a similar thing, sometimes pasting into the cells causes the values to become rounded up or down(especially if you copied it into Excel then back into ECUFlash) if there are more than 2 decimal places.
I was about to rescale my evo8 ecu load to 300 using the tutorial in this post (losing 2 columns of resolution) and I thought why can't you just change the number of elements in the load scale to 21 instead of 19 when you go into "edit map definition" and then just fill in the new 2 columns at the right with 280 and 300% load values? I tried adding 21 elements to the load axis and fill in the timing & fuel maps and it works in ecuflash. I am a little hesitant to try it in my car since I dont know if the ecu can only read or work with 19 load columns.
I would have post some screenshots into this reply but I am new on forums and I dont know how to post screenshots yet, just attachments so bear with me.
Any thoughts on this? Has anyone attempted this?
Thanks
I would have post some screenshots into this reply but I am new on forums and I dont know how to post screenshots yet, just attachments so bear with me.
Any thoughts on this? Has anyone attempted this?
Thanks
I was about to rescale my evo8 ecu load to 300 using the tutorial in this post (losing 2 columns of resolution) and I thought why can't you just change the number of elements in the load scale to 21 instead of 19 when you go into "edit map definition" and then just fill in the new 2 columns at the right with 280 and 300% load values? I tried adding 21 elements to the load axis and fill in the timing & fuel maps and it works in ecuflash. I am a little hesitant to try it in my car since I dont know if the ecu can only read or work with 19 load columns.
I would have post some screenshots into this reply but I am new on forums and I dont know how to post screenshots yet, just attachments so bear with me.
Any thoughts on this? Has anyone attempted this?
Thanks
I would have post some screenshots into this reply but I am new on forums and I dont know how to post screenshots yet, just attachments so bear with me.
Any thoughts on this? Has anyone attempted this?
Thanks
You can't do this because you would be overwriting data that is beyond the table in your ROM. There isn't any free space beyond the table's data in the ROM. So, if you added new columns to your tables and flashed your ECU, you would be overwriting some other code that may possibly br very harmful to your ECU.
You can extend your tables if your relocate your maps to an empty space in your ROM and point the ECU code to the new location. People like tephra and jcsbanks have done this in map switching. They create new tables in emprty spaces of the ROM and create code so that the pointers will switch to these new maps. These empty spaces may be large enough to include maps with more resolution.
Eric
ECUFlash is a little wierd, its not really a problem with the program, but how it calculates those values you have put into the table, because there is some "rounding" error, you'll find that some values don't increment immediately and others do, if you were to look at the hex values instead, you'd see it more clearly.
Cutting and pasting is a similar thing, sometimes pasting into the cells causes the values to become rounded up or down(especially if you copied it into Excel then back into ECUFlash) if there are more than 2 decimal places.
Cutting and pasting is a similar thing, sometimes pasting into the cells causes the values to become rounded up or down(especially if you copied it into Excel then back into ECUFlash) if there are more than 2 decimal places.
Thats only for the fuel cells. Timing cells don't take decimals like they do for the subaru's. Would be nice if they did though.
Great Thread guys!
I just wondering, I want to push the limits of the stock MAF sensor. Why can I not scale to like 340?
I'm running a gt35r, Is fuel or boost cut gonna come in before I reach that load ? Is there a way to adjust or "trick" it so I can try and run that high of a load?
Thx!
I just wondering, I want to push the limits of the stock MAF sensor. Why can I not scale to like 340?
I'm running a gt35r, Is fuel or boost cut gonna come in before I reach that load ? Is there a way to adjust or "trick" it so I can try and run that high of a load?
Thx!
Important
When rescaling the load cells for timing/fuel maps, other tables are being changed as well. These tables will need to be adjusted after the fact as well. See this thread for details:
https://www.evolutionm.net/forums/sh...d.php?t=318623
When rescaling the load cells for timing/fuel maps, other tables are being changed as well. These tables will need to be adjusted after the fact as well. See this thread for details:
https://www.evolutionm.net/forums/sh...d.php?t=318623
