expansion pipe anomoly

i am a hobbyist playing with small mopeds just for fun.
i  have just built two pipes and on the FOS inspired pipe i noticed a strange dip in the power delivery curve just after peak power and have seen this on another pipe made by a well known uk pipe maker on the same dyno fitted to another bike. fitting 2 different pipes to my motor with same porting and carb ignition etc dont display this strange dip.

i realised due to changes in attaching to the cylinder ie making it longer by 23mm that the pipe deviated from the FOS single diffuser layout.
i wonder if this 23mm is enough to cause my problem.
my percentages were as follows
header plus duct 0.37
diffuser   0.31
belly .077
baffle 0.255

but by simply deducting 23mm from the start of the header i am now back at  header=0.34 diff 0.32 belly 0.08 and baffle 0.26 ie the FOS recommendations.
i am yet to dyno to see if coming back to the FOS percentages will remedy this.
is 23mm enough to throw the power delivery out that much?
the pipe curve is the blue one.
[Vous devez être inscrit et connecté pour voir cette image]