I build a circuit that work on breadboard now i try to prototype it,so i start solder it on a prototype board and the arduino is out of sync.So i check the tension and for some reason it's 4.4~4.6 on the Vin pin of Atmega328 so i try to find how to debug.I don't have a lot of theorical knowledge on electronic,but my thougt was as long as everything is plug in parrallel the tension should remain the same ( and i don't have any thing in serie) so for the tension to drop i may have something acting as Resistor, and i try to find how and what. I use 22AWG 0.7mm wire o create a 5V line and my GND line on each side of the Atmega ( roughly 7 cm on each side),i start to wondering if it was not a bad idea. I have picture of my prototype bellow,but it's messy so not sure it help. My question is how to you proceed to troubleshoot this thing ? What are the step you do to find out where is the problem
The yellow-boxed 5V and Gnd buses on the freetronics.com.au ProtoShield Basic for Arduino [like you used] already have copper strips connecting the holes. The strips run on both sides of the board and the holes are plated-through, so there's no need (for this kind of circuit) to have additional bus wire running along above the boxes. Having that extra wire there gets in the way of making changes, whether for debugging the circuit or for adding or replacing components.
From the frontside picture, it appears you may have a voltage regulator on board. (However, I don't see the connections to 5V and Gnd, so might be misinterpreting things.)
If your board input is 5 VDC from an external supply, feeding that 5 V into a regulator will lose half a volt or so, which may account for what you are seeing.
However, let's instead suppose you are putting 7 to 12 VDC into an on-board 5V regulator and then measuring 4.4-4.6 V out. One possibility is the regulator isn't working. This can happen if the regulator's input or output capacitors are the wrong sizes or wrong kinds, and cause oscillation. Diagnosing that kind of problem is straightforward if you look at the 5V line with an oscilloscope. Another possibility is something drawing a lot of current (eg a couple of amps) and taking the 5V line out of regulation. That's usually easy to diagnose because of all the resulting smoke and heat.
Drawing a good schematic is an excellent step to take to allow better problem debugging.
Other steps to take:
Measure the amount of current the whole board is drawing, or that the NRF24L01 is drawing. (For the latter measurement, you might need to lift your ground lead off the NRF24L01 board and put the ammeter into the ground circuit. In general you want all the grounds to be at the same voltage, so the voltage drop across the ammeter is a problem, but not much of one.)
Pull the '328P chip and see if that makes any difference. In general, it's a good idea to work out a prototype board's power supply problems – such as shorts, reversed electrolytics, incorrect resistors – in advance of installing a microcontroller chip onto the board.
Remove the NRF24L01, if feasible, while debugging the power supply. If you can't easily remove it, check your schematic and see if all of its pins except 5V and Gnd are open when the '328P chip is pulled. If so, you can detach the NRF24L01 ground lead to take it out of the circuit. If, during testing, you unsolder and remove the NRF24L01 board, consider making a socket for it from two 5-pin female header strips.