[dluke]

8

Originally Posted by BmerMeUp

Hi guys.
I read through the thread but there seems to be a lot more effort being exerted than it took me when I coded mine.

My original notes (to myself) for what worked were as follows:

Feature: Code 4 blinks (from 3 blinks) for turn signal - add Hex 04 for each blink
Module: CAS > NFRM (NETTODAT)
Entry: Location #0341505, 0341507, 0341509, 034150B


So basically, this is what worked for me:

• Set your blinker using iDrive or BC to 3 blinks.
• Open NFRM, (I have a 2008 [11/2007] 328xi E91),
• Open the NETTODAT for it, examine those 4 memory locations as shown above (memory address runs down the left of the NETTODAT data) - you will note those four bytes are all the same.
• Change ALL FOUR locations by adding Hex "04" for each blink you wish to add. So if your byte says "28", add Hex "04", you get: 28-> 29, 2A, 2B, 2C. So your new value is 2C. THE TRICK as I understand, is that byte will be different for each car, depending on various factors sharing that memory location. So the value is NOT THE SAME for everyone. You must do the math yourself. Unless NCSDummy will do it for you?
• Load NFRM back using NETTODAT, it should be good.

EXAMPLE NETTODAT:
The red numbers are memory locations 0341505, 0341507, 0341509, 034150B

B 00341400,0010,0F,FF,3E,60,28,32,00,00,9D,2B,09,88,28,00,14,39
B 00341500,0010,28,27,28,27,28,2C,28,2C,28,2C,28,2C,00,00,00,00
B 00341600,0010,00,02,00,02,00,02,00,02,00,02,00,02,00,00,00,00

well with my ncsdummy method you dont need to deal with any confusing hex numbers.. you just input the number of blinks you want.. also not only can the hex values differ from model to model but the location of the values that need to be changed can also differ, so that means in order to be sure of the location of the values that need to be changed you need to read the module once with triple blink off and once with triple blink on and then compare the data sets to get the locations.. then figure out the hex value you need for the number of blinks you want.. and then code it via nettodata which my system was not doing correctly and the reason i came up with the new method which you just code via normal fsw_psw.. and i dont mean any offense or anything.. theres nothing wrong with the nettodat method its just everyone that has tried it both ways has said the ncsdummy route is much easier