Having some trouble with transistor logic.
I have an existing circuit with a 3.3V rail and signal rail that is pulled to 0V when asserted, i.e. active low. I ultimately need to control four 12V LED clusters, which draw 125mA each. From hereon I shall refer to 'four 12V LED clusters' as a single LED for simplicity. I can use PNP transistor to power a 12V LED, However I would like a second LED that is the inverse of the first.
I tried to add NPN, with the base on the signal and a pullup resistor on the base to the 3.3v rail but this causes the PNP to be always on as current can flow from the PNP base through the NPN base to the 0V rail. A diode would stop the flow for the NPN.
After reading CharlieHanson comment got further on but I need the LEDs to be ON or OFF not reduced.
EDIT: To reiterate, Switch 1 does not exist. It's an output of another board that is either high impedence (actually disconnected) or 0V.
Best Answer
It's not clear why you have chosen a PNP transistor to begin with - I assume it's because you had to start with something. Throw it in the bin; get some MOSFETs, Darlington transistors or beefy NPN transistors (in order of current-conducting capacity, highest to lowest, generally speaking).
The following diagram uses IRF510 MOSFETs, which are standard off-the-shelf types and are sold on all manner of hobby electronics sites in one form or another - maybe IRF540 or 530 instead. It can handle currents of 4A or more (read the datasheet!) and requires very little in the way of external components, as you can see:
You don't need the 3.3V rail; it's of no use. If you can provide some specs for the 12V LED cluster (we accept datasheets or cold, hard numbers) then somebody can suggest a suitable device to drive it if you have trouble sourcing certain components.
EDIT:
In the Falstad circuit you had the forward voltage of the LEDs too low (11V). That simulator is very basic at best, and so was only dropping 11V across the diode leaving 1V at the gate of the second MOSFET. The emmended can be viewed here.