I built these two circuits. Both of these are Darlington-like circuits, using two NPN transistors in conjuction with each other to generate a much higher gain.
Looking at the circuit on the left, the resistor R1 limits the current in D2 to avoid damage to the LED. The resistor R2 turns on Q2 normally. The combination of Q2, R3, and SW1 act to turn off Q1 by pulling the base low. I don't actually have a switch so I just touched the wires together. The gain of Q1 is 100, which is typical for a 2N3904. But when Q1 and Q2 are considered together the gain is 1000. This circuit seems to work fine.
Here are some scope traces from the base of Q1 when SW1 is open and closed. I am just using a scope here because it has a good display and is much better quality than any other measurement tools I have.
Here is the circuit working. The LED is just some random ugly one I had in my box of parts.
The second circuit was made by rearranging the first. It also seems to work, but R5 means the LED never turns off completely. Here are the measurements from Q3
I couldn't really measure any difference between the circuits.
Is there a name for these circuits? Like an inverted Darlington or something?
This test was performed at DC. Would this work if SW1 was toggled at a high frequency such as 30 kHz or would I run into issues?
The current through Q1 and Q3 was not higher than 5mA. Would this circuit work at currents closer to the maximum current values for a 2N3904? Such as 150 mA.
Is there a name for the circuit on the right hand side with Q3. My idea behind it was if R5 is properly sized, the load (D3 and R6 in this case) dictates the current in the base. For high impedance loads, the current in the base is small. For lower impedance loads the current in the base is larger. This should theoretically prevent a condition where the current in the base-emitter junction of Q3 exceeds that of the collector-emitter. Is this possible or am I just imagining this?