you must first determine the final desired outcome, i.e., I want to record an adult/child female/male with a deep/high/thin/resonant/raspy/shrill voice who is singing/talking for a song/video/movie/game that will be in the front/background of the final mix. Once that is clearly determined, THEN you can make some easier decisions on how to achieve that.
From your OP, I assume that you want to record your own voice and that it's part of dialog, so it will be up front in the final mix and that you are on an extremely tight budget. Your mantra in this should be "Borrow, Borrow, Borrow!!!" :)
The key to voice-overs is to remove ALL room sound from the input. To do this, you will need to have the mic in an extremely silent space and that space should be free from any echos and constant air-flows, i.e. air-conditioning vents not blowing down or across the signal path. I would recommend a walk-in clothes closet with carpet on the floor and lots of hanging clothes on all walls leaving barely any surface without anything on it - even the ceiling, if you can tack up a thick blanket on the ceiling over the top of where your mic is placed.
I would place a decent cardioid mic (SM-57/58) with a quality XLR cable on a sturdy mic stand with a pop filter backed up to some thick hanging clothes to absorb any early reflections. I would place something about 8 to 10 inches away from the end of the microphone to be a guide as to where to have your mouth. The business end of the mic will need to be pointing absolutely directly at your mouth so that it's on-axis. The cardioid pattern will help to give you some decent resonant timbre to your voice depending on how close you are to the mic and the degree of the proximity effect based on that closeness.
To test if the sound of the room has been removed sufficiently from the equation, set the gain for your voice performance at unity and then record just the ambient sound of the room for a full minute. Listen back on a pair of quality headphones and if you can hear anything other than utter silence, set about the job of eliminating whatever it is that you heard. This is easier said than done, but do it you must.
I would try and enlist the help of someone from an actual recording studio or someone that knows, at the very least, something about achieving unity gain and signal routing and achieving a healthy signal response.
You will need a pair of closed-ear, or circumaural, headphones. Sony MDR-7506's are great and even some budget AKG K-55's will suffice. There should be no leakage of the sound from the headphones back into the mic. While speaking, you should be able to hear your voice in the headphones in real time with NO effects. If you use extension cables for any of the audio, use the shortest cables possible that will allow you to do what you need and make sure they are of quality construction. Remember, borrow, borrow, borrow!
Record your voice and when adjusting the setup, be VERY patient and try to adjust only ONE aspect or parameter of the setup at a time and then test it and listen back. If you have the option, try to play the part back in the context in which it will be heard in the final mix. Most of the time, this is not possible, so don't stress over it. Mainly, you want to be able to hear just the voice and for that voice to be agnostic in its context so that it can be placed in a huge variety of settings and that when it comes time, it can be effected properly to make it sit right in the final mix and in the context that it should be in.
Because you're doing this for the first time, this is going to be a time-consuming endeavor, but if you take the time to "get it right", then you can document what you've done and the next time you need to do this, you can get to the good, working setup faster and more confidently.
If you make friends with someone at a recording studio, you might be able to come in and use a studio B or C, or they might have a voice-over suite, during off-peak or even overnight hours for very cheap. Don't do this without trying the at-home setup first and practicing, practicing, practicing the dialog and getting your inflections, delivery and all other aspects of your performance down so that it's second nature to you.
Doing any of that while you're paying for studio time is incredibly stupid and a waste of time and money. Avoid that at all costs - literally.
Finally, remember that talking or singing into a microphone is a learned skill and one that needs plenty of experience and practice. The microphone and signal path become part of the instrument of your voice and the voice PLUS the equipment = the full instrument. One is not separate from the other. Take the time and learn the skill or use someone else that has already taken the time and has the experience. Now, break a leg, as they say!
