You circuit has many limitations and may well not be suitable at all, but to begin to see the limitations you need to draw the circuit in a more understandable format.
simulate this circuit – Schematic created using CircuitLab
The problems with this type of circuit:
- Vout from Q1 will always be less than the supply voltage. If your load requires 5V, then this circuit cannot achieve it with a 5V supply and Emitter follower.
- If V1 is the same as the Arduino supply, then Q1 can never quite saturate.
- If the load includes any capacitance, when you turn off Q1 the Base-Emitter junction is reverse biased. Even in this case (a 5V supply) the NTE85 is at its minimum reverse bias limit. You would normally NOT design to the limit and would seek to clamp the reverse bias to LESS then 5V.
You could use a PNP BJT (as suggested in another answer) but this may not improve your situation greatly.
- The saturation voltage of most BJT devices is in the 0.2 - 0.6V range so you still lose a significant voltage across the device. In the other answer for example, the suggested 2N3906, the saturation voltage at 50mA is about 0.4V. This means that your load voltage is likely at best about 4.6V and at worst about 4.4V.
- The logic is inverted if you use a single PNP device, although it should be easy to invert your driving logic in the Arduino code.
A better solution overall would be to use a P-Channel FET:
simulate this circuit
NOTE: Using a single P-channel FET inverts the logic, so High = OFF, Low = ON but I'm assuming this would be easy for you to handle in software.
The two most important parameters for the P-FET for this application are a low VGS(th) and a low RDS(ON). Here I've shown an AO3415 to fit this bill but it is in a SOT-23 package. Unfortunately getting extremely low RDS(ON) in a TO-92 package is a real challenge.
If you assess something like a 2N7000 or 2N7002 you can see that RDS(ON) could be as high as 12 - 13 Ohms. This means that even at a moderate 100mA load you could see an on voltage of over 1V which would be unacceptable in most cases. The AO3415 with an RDS(ON) of about 50 mOhms would drop only a few mV.