You have your Append Checksum boolean switch controlling a case selector. If it is set to True (Append Checksum), it takes two characters (1 byte) away from the checksum calculation and gives the correct result of DF. Since you loop through and begin with i+1 times 2, it begins with the 00 characters (1 byte) just like you want.
If you set it to false, it takes 4 characters away from the checksum calculation, thus the first two characters (OK) but also the last two (not good) and you wind up with the wrong checksum of B. Note you should probably set the width for the last number to hexadecimal string to be 2 so that you get 0B rather than just B.
Verify the integer representations you are using throughout the checksum. Many of your constants and wires are I16, but also U32 and I8. In reality, the calculation of the checksum should probably be just U8 throughout.
It would also be better if you set up your string array control to be in hex display and use the ASCII representions. Rather than having 2 characters represent a byte such as F0, it would be better if it was just a single byte that would show up as F0 if the control was in hex display. That way you don't have to do the 2 character manipulations all over the place.
Also, you can use the boolean functions such as XOR directly on a U8 integer numbers rather than converting them first to boolean arrays and back to numbers again. By handling the strings as ASCII bytes rather than 2character representations of a byte, and proper use of integer representations and boolean math, you could probably knock your code down to a third of what it is.
Attached is a VI with modified code (I can only save it back to LV 😎 along with a screen shot of the code.
![Check%20Sum[1].VI.png 47 KB](/ni/attachments/ni/170/401778/1/Check%2520Sum%5B1%5D.VI.png){kind=link}
