also getting the car to idle with the isc took oh about 2 years to get right. some of the aem rpm tables mess it up. also the drivers in the aem that move the isc selinoid are not strong enough to move it consistently so the aem tries to move it to position 40% but it is really at only 32% so the car dies.
i also have a modded isc for the aem and with some of my idle settings in my map my car idles correctly 90% of the time. i also made a jumper harness for the a/c idle up so it works correctly now. it used to add idle % when you turned on the a/c fan to lo, med, or high and not take it out when the a/c compressor shut off. so the idle learned value subracted the idle up amount and when the compressor came back on there was no isc % added in anymore. and the car stubles till the idle learned value comes back up.
dont use the rpm offset vs TPS table as it does not work right and messes with the idle target too much and over too much of an area. also my idle target % table is increases to about 1900 rpm then the rest of the rpm targets are the same %. at 1900 rpm i have 60% and it is 60% all the way to 3000 rpm on that table. why would my car need to idle at 3000 rpm? ever? i warm up at 1900 or less. i dont need the isc moving around while i am accerating in city traffic so when i get over 1900 rpm the isc stops in that position. remember it does not move as much as it thinks so the less it moves the longer it takes for the target % errors to accumulate. thats why for the first hour of driving it idles at 1000 rpm at 30% but the second hour it takes 55% to idle at 1000. the isc is still at 30% but the aem thinks its at 55%. that a 25% error of movement. turn the car off. the isc parks at a known number, then crank it up and it idles at 1000 rpm at 30% agian because the isc was reset and the error was removed. that it why i made some changes to the isc valve and tuned my map to keep the isc from moving except when it has to.