Doc
May 13th, 2008, 5:53pm
Borrowed (http://www.cobaltss.net/forums/showthread.php?p=2312926#post2312926)
So after 42,363,956 pms, here it is. I was unsure as to whether to put this here or in the tuning section, but the tuning section has one already that has been muddled with learning and questions, etc. I wanted to get all of the info on the front post to make it easier. The mods can (and I am sure will) move this if needed.
First off, I need to thank everyone who helped me to learn this stuff, not the least of whom was Asphalt Assault, who started the whole trend and taught me a bunch. He started the above mentioned thread (http://www.cobaltss.net/forums/showthread.php?t=69454).
***Disclaimer - I am in no way responsible for anything that you do to your car as a direct or indirect result of anything I say or post! There is an inherent risk, any time that you make any changes to a car's computer.***
So here we go:
First of all, forget about all of the advertised stats for the Siemens 60lb injectors. They actually flow at somewhere around 63-65lb/hr at 43.5 psi with a factory tolerance of about 6%. It is VERY important to get your injectors flow matched and to know what the actual flow rate is. The ones I got from NewEra came with a sheet telling me the numbers, and they were flowed at 63.4lb/hr on average. I open up my "LSJ Injectors" spreadsheet and put in the appropriate numbers. This file can be found by searching the HPT forums. I then create a new worksheet and paste the numbers from the LSJ Injectors spreadsheet here. I then create a formula to divide these numbers by 1.3. It should look like this:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/IFRadjust-1.gif
Post those numbers to your IFR table:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/IFR.gif
Now you go to your modifier table and enter 1.3 for all values.
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/modifier.gif
Now, know that there is some controversy over whether this actually works. The results have shown me and others that it does. It is not as exact as having an injector table that doesn't have to use the modifier, but it is FAR better than maxing out 2/3 of your table at 63.5. I am in the process of coming up with other ways to get around these little limit, but for now, this works.
Okay, IFR is set. Next are the offsets. You can actually tune the offsets, but to do it correctly requires a LOT of work. You do it by tuning your MAF at the offsets that you have, then introduce a modifier. Cot is a good one. You can also try other ways, but the purpose is to get your computer to do math based on your PWs. If the changes go rich, your offsets are too high, if go lean, they are too low. Adjust the offsets, retune your MAF and do it again. An easier way is to get them from someone who actually knows them. Every injector is a tad different, but the offsets are pretty close. These are the correct offsets:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/offsets.gif
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/offsets2.gif
Now is the fun part. The adder table. This is where most people completely lose it. It is simple, but incredibly hard to get perfect. The reason is because these injectors just cannot flow properly in the low pulse width areas that we need them to. A return fuel system really helps here, but that is a different how to completely. :)
I start by entering .05 in the 1.000 msec cell of the adder and set the 0ms cell to somewhere above .6. Then interpolate. Set your short pulse limit to 1 and your min PW to .1:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/fuel.gif
The short pulse limit tells your computer when to look at the adder table, so if you need to adjust the 1.125 cell a bit, change your short pulse limit to 1.125.
Now, remember your excel spreadsheet? Add a new worksheet. Go to your adder table, highlight it and copy with labels. Paste to your new worksheet. Next, enter a formula to add the 2 together. Should look like this:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/addertotals.gif
Now, set up your AFR error histogram like this:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/MAFhisto.gif
This will filter out any data received when the PW is below your filter limit - in this case 1.3ms. It will only show data from 1.3ms up. Now, at this point, I recommend you set up a MAF trim histo and run in CL for a bit, to get your AFRs close. You just made some pretty drastic changes and I want your car to actually start. Once you get fairly close with the trims on, switch to open loop and tune your MAF using AFR error. You should now have a perfect MAF from 1.3 ms on up, which will include everything from the 2400 hz cell of your MAF table to RL. Now, the PW is calculated by the computer figuring out how much fuel is needed + the offset needed to deliver that amount of fuel. Because the offsets change with manvac, you will command a different PW at 2400 during accell than you will at 2400 during decel. This is going to help you. Create another histogram.:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/PWs.gif
This will show your PWs. What I do here is, I will go through my graph and see where I am running really lean or really rich. Not in spikes, but for durations. Set the view mode to "L" for last value. As you can see from the picture above, I am really lean at .763ms. I see the lean spot on the graph, my PW histo says .763, and by following that whole "lean" spot on my graph, the PW is a steady .763. This is good info. I look at my table and see that I am at 2659hz. Well, I know that the 2400 and 2700 hz cells of my maf are tuned (above 1.3ms) so this is not a MAF error. It has to be a fueling error. I go through and make some notes at which PWs under 1.3ms gives me what errors, and make sure there are definite trends here. Forget flukes, look for consistent data. Now gather your information and go to your excel spreadsheet to see where the changes need to be made. To use the example above: I know I have a lean fueling error at .763ms. By looking at my excel spreadsheet, I see that my total at .5ms is actually .679 and my total at .625 is actually .843. Here's where it gets tough, because I need to figure out which one to change. By further studying the data, I see that everything in the low.7xxms range is lean and the low to mid .8xxx is a little rich. The lean is far more pronounced than the rich, so I bump the .5ms cell of the adder to remove the lean errors. I'll leave the rich error alone for now. You'll find that your injectors hate PWs in the low .9s. This is an example of a 60s adder table.
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/adder.gif
You may find you are having some issues with the lower MAF cells. One thing that I have found that helps is to zero out the cells from 4200 on up, or 5400 on up, just to see what the lower cells look like. MAF is a 3rd order polynomial, so it should be smooth. If your MAF is all full of peaks and valleys, you're doing something wrong.
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/lowmaf.gif
Oh, and once you are done looking at your lower MAF cells, don't forget to put it back to normal. I left half of it zeroed once and my car didn't like it much.
So are we having fun yet, or what? :twothumbs
So after 42,363,956 pms, here it is. I was unsure as to whether to put this here or in the tuning section, but the tuning section has one already that has been muddled with learning and questions, etc. I wanted to get all of the info on the front post to make it easier. The mods can (and I am sure will) move this if needed.
First off, I need to thank everyone who helped me to learn this stuff, not the least of whom was Asphalt Assault, who started the whole trend and taught me a bunch. He started the above mentioned thread (http://www.cobaltss.net/forums/showthread.php?t=69454).
***Disclaimer - I am in no way responsible for anything that you do to your car as a direct or indirect result of anything I say or post! There is an inherent risk, any time that you make any changes to a car's computer.***
So here we go:
First of all, forget about all of the advertised stats for the Siemens 60lb injectors. They actually flow at somewhere around 63-65lb/hr at 43.5 psi with a factory tolerance of about 6%. It is VERY important to get your injectors flow matched and to know what the actual flow rate is. The ones I got from NewEra came with a sheet telling me the numbers, and they were flowed at 63.4lb/hr on average. I open up my "LSJ Injectors" spreadsheet and put in the appropriate numbers. This file can be found by searching the HPT forums. I then create a new worksheet and paste the numbers from the LSJ Injectors spreadsheet here. I then create a formula to divide these numbers by 1.3. It should look like this:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/IFRadjust-1.gif
Post those numbers to your IFR table:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/IFR.gif
Now you go to your modifier table and enter 1.3 for all values.
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/modifier.gif
Now, know that there is some controversy over whether this actually works. The results have shown me and others that it does. It is not as exact as having an injector table that doesn't have to use the modifier, but it is FAR better than maxing out 2/3 of your table at 63.5. I am in the process of coming up with other ways to get around these little limit, but for now, this works.
Okay, IFR is set. Next are the offsets. You can actually tune the offsets, but to do it correctly requires a LOT of work. You do it by tuning your MAF at the offsets that you have, then introduce a modifier. Cot is a good one. You can also try other ways, but the purpose is to get your computer to do math based on your PWs. If the changes go rich, your offsets are too high, if go lean, they are too low. Adjust the offsets, retune your MAF and do it again. An easier way is to get them from someone who actually knows them. Every injector is a tad different, but the offsets are pretty close. These are the correct offsets:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/offsets.gif
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/offsets2.gif
Now is the fun part. The adder table. This is where most people completely lose it. It is simple, but incredibly hard to get perfect. The reason is because these injectors just cannot flow properly in the low pulse width areas that we need them to. A return fuel system really helps here, but that is a different how to completely. :)
I start by entering .05 in the 1.000 msec cell of the adder and set the 0ms cell to somewhere above .6. Then interpolate. Set your short pulse limit to 1 and your min PW to .1:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/fuel.gif
The short pulse limit tells your computer when to look at the adder table, so if you need to adjust the 1.125 cell a bit, change your short pulse limit to 1.125.
Now, remember your excel spreadsheet? Add a new worksheet. Go to your adder table, highlight it and copy with labels. Paste to your new worksheet. Next, enter a formula to add the 2 together. Should look like this:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/addertotals.gif
Now, set up your AFR error histogram like this:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/MAFhisto.gif
This will filter out any data received when the PW is below your filter limit - in this case 1.3ms. It will only show data from 1.3ms up. Now, at this point, I recommend you set up a MAF trim histo and run in CL for a bit, to get your AFRs close. You just made some pretty drastic changes and I want your car to actually start. Once you get fairly close with the trims on, switch to open loop and tune your MAF using AFR error. You should now have a perfect MAF from 1.3 ms on up, which will include everything from the 2400 hz cell of your MAF table to RL. Now, the PW is calculated by the computer figuring out how much fuel is needed + the offset needed to deliver that amount of fuel. Because the offsets change with manvac, you will command a different PW at 2400 during accell than you will at 2400 during decel. This is going to help you. Create another histogram.:
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/PWs.gif
This will show your PWs. What I do here is, I will go through my graph and see where I am running really lean or really rich. Not in spikes, but for durations. Set the view mode to "L" for last value. As you can see from the picture above, I am really lean at .763ms. I see the lean spot on the graph, my PW histo says .763, and by following that whole "lean" spot on my graph, the PW is a steady .763. This is good info. I look at my table and see that I am at 2659hz. Well, I know that the 2400 and 2700 hz cells of my maf are tuned (above 1.3ms) so this is not a MAF error. It has to be a fueling error. I go through and make some notes at which PWs under 1.3ms gives me what errors, and make sure there are definite trends here. Forget flukes, look for consistent data. Now gather your information and go to your excel spreadsheet to see where the changes need to be made. To use the example above: I know I have a lean fueling error at .763ms. By looking at my excel spreadsheet, I see that my total at .5ms is actually .679 and my total at .625 is actually .843. Here's where it gets tough, because I need to figure out which one to change. By further studying the data, I see that everything in the low.7xxms range is lean and the low to mid .8xxx is a little rich. The lean is far more pronounced than the rich, so I bump the .5ms cell of the adder to remove the lean errors. I'll leave the rich error alone for now. You'll find that your injectors hate PWs in the low .9s. This is an example of a 60s adder table.
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/adder.gif
You may find you are having some issues with the lower MAF cells. One thing that I have found that helps is to zero out the cells from 4200 on up, or 5400 on up, just to see what the lower cells look like. MAF is a 3rd order polynomial, so it should be smooth. If your MAF is all full of peaks and valleys, you're doing something wrong.
http://i182.photobucket.com/albums/x213/SJSchafer/tuning%20stuff/lowmaf.gif
Oh, and once you are done looking at your lower MAF cells, don't forget to put it back to normal. I left half of it zeroed once and my car didn't like it much.
So are we having fun yet, or what? :twothumbs