If you want to be the best engineering student then figure out why they are WRONG trying to use just one NPN transistor. I already told you the answer in the first post I made to your thread here...

http://www.8052.com/forum/read.phtml?id=64165

...you can choose to believe your teachers and ignore my comment BUT you said your transistor did not work!!

It is possible that you may get the transistor in your simple minded circuit to work with a standard 8051's type of quasi-bidirectional type I/O pin if you lowered the base resistor to a much smaller value. However this design technique would not be best because the pullup current ( I_OH ) on these parts is rather low and is not normally able to bias an NPN bipolar transistor fully ON, particularly if it has a low BETA and has a relatively high collector load current. The PNP approach takes specific advantage that a standard 8051 output can drive to GND much much much harder and is fully able to bias the PNP transistor on, even guarenteed, whereas the NPN transistor is may be marginal at best. (Do you want a marginal drive to the transistor relay driver that controls the air bag deployment in your car? I would guess not!! In schools, one time designed, crappy circuits promoted by the professors, are not a good way to gain experience to being a great engineer able to design a motor driver for a Martian Rover!! Have these same said teschers ever in their lives designed production volume circuit that had to meet worst case design parameter analysis, reliability and safety guidelines? Probably not!!)

The simplistic NPN design also has the issue in that from power-on through the RESET pulse time and till the code in the microcontroller starts up and initializes the NPN base driving pin to a low level to set the relay coil OFF.....during that time the relay coil will be ON. For some designs this initial click of the relay can be detrimental to the product operation or at least annoying.

Michael Karas