Hey, everyone on this wonderful forum! 
I was analyzing RKTI section fr diagrams and found some interesting things like:
Why the hell ZTSPEV is there at all, when in diagram I see that it makes no sense (attached image 1)
val(new) = val(old) + (in - val(old) ) * dZTSPEV / ZTSPEV
Because IV and IN are equal we get same number repeated for whole ZTSPEV time unless B_stend interupts this op.

TEST:
T = 1
val(new) = 90 + (90 - 90) * 0 / 1 = 90
val(new) = 90 + (90 - 90) * 1 / 1 = 90
T = 2
val(new) = 90 + (90 - 90) * 0 / 1 = 90
val(new) = 90 + (90 - 90) * 1 / 2 = 90
val(new) = 90 + (90 - 90) * 2 / 2 = 90
....

Or maybe I don't understand Lowpass function?

Also could someone check TVTSPEV map in their damoses? In my definition file it's axis is wrong...  (image2)

Additionally I have attached my bin maybe someone will be able to find correct axis address. Right now map is @ 1CDDB, AXIS @ 1CDD7

Thank you all

Does everything = 90? Maybe your maths is wrong.

I think I have cracked TVTSPEV. Looks like ori definition was wrong. Correct TVTSPEV address is 1CDE0. Axis is at 1CDD8 and it is 16bit. Now only problem is axis factor...