@hippy: Well, it sound like you were perfectly right!
=> I tried another way to measure the current. I used a 0,01 ohm resistance. The results are quite normal. Here they are :
with OS SD > 8,15mV for 0,02 ohm = 407 mA = 2,03W / Powermeter = 4,9W
4, 01mV for 0,01 ohm = 407 mA = 2,03W
without OS SD > 7,70mV for 0,02 ohm = 385 mA = 1,92W / Powermeter = 4,4W
3,85mV for 0,01 ohm = 385 mA = 1,92W
So to conclude, the problem came from too much parameter and random accuracy for my way to measure with my powermeter. But as i get sometimes 2,9V, i just think my powermeter just have a pb.
By the way, the measure with a 0,01 ohm are incredibly precise ! At first, i thought it was just luck xD
Thx people for your tips and your time
Topic can be closed