This was the programmer I almost made exactly.

so I took it upon myself to remove the resistors connected to the base of the NPN transistors, and tested my circuit with 5 AT89C2051's. It turns out when I first attempted to program each one of them, the programmer seemed to work to my satisfaction. BUT when i tested my programmed AT89C2051 in another circuit, it fails.

One thing that I did notice when I tried some troubleshooting is that one of the transistors may be screwed up. the connections to my P3.3 through P3.7 are configured exactly the same way as in the above diagram except that my transistor is a 2N2222 and the 10K resistor connecting to a base is replaced with a short.

So I manually tested the pin with an LED. I notice that P3.4 P3.5 and P3.7 never go low. I think they should when P3.3 is high (as set by the port).

I think I'm too much in the analog world, but I will right now see if transistor biasing can help me make a better trasnsistor inverter.

 

lol. I just discovered that I destroyed a transistor by applying a full 5V to the base of an NPN. so that means I do need a series resistor connected to the base of the NPN. What is the absolute maximum resistor value I can use?

From my own experimentation, I connected the collector of a 2N2222 to +5V through a 12K resistor. the collector is also connected to an LED, a 220 ohm resistor, and ground (both in series). I tried 820K resistor (for the series base resistance) and the transistor seemed to work as an inverter.