I think it would be better to cut your product with a suitable restriction enzyme ( or use different primers )   then the separation is , for example, a common 200 band and two bands of 130 and 144 which would separate much better. Run gel slowly and in a cold room if possible to stop thermal diffusion broadening the bands. My sizes just for example . New england biolabs have an enzyme cutting application to test many enzymes for suitability

Dear Nguyen

You will be able to differentiate between fragments of the aforementioned sizes on a 3% agarose gel.  Considering the 14bp difference I would run at low voltage for a longer period of time to ensure better separation.  If you have a high Amperage reading then running in a cold room would be preferable to ensure your gel doesn't melt or that your samples do not run at an arch (smiley face gels).

That said, with a 14bp difference I would run positive controls for each fragment (330 and 334) in every lane of your gel to ensure an adequate reference and more accurate callings.

Regards,

Polyacrylamide gels (PAGE) are best suited for that kind of work. Considering the size and the difference between fragments I would use a 8% gel and run it at medium voltage for about 4 hours.

If you cannot access PAGE, you can follow above mentioned suggestions. I would try prof. Mcgregor's first, and leave prof. Rutland's as a second choice as it involves adquiring enzymes.

Good luck   

Yes, A base difference of 10 base pair can be visible in 3% agarose gels. but 8% poly acrylamide gels and capillary electrophoresis are good