Current Protection Circuit V.2

Try adding a 10ohm resistor in series with the pushbutton to limit the current through it. I mean, what if there is a short and the circuit trips and you push the button to un-trip it? Either the button will explode injuring your hand or your power supply will explode. Either that, or your power supply is an alkaline battery and thus won’t explode, but will just sit there dying way faster than it should. (Alkaline batteries are very short resistant.)

Well if your worried about that, then I'll just add a little RC circuit instead, so it'll draw little to no current, even if you hold it down

There we go, fixed it! ...hows that?

Perhaps increase the capacitor to 1uF, because a 100nF it can’t un-trip it when it’s 10.1ohms which is still enough resistance to not re-trip it.

I meant to do that earlier, I'll fix it

Do you know of a way to increase the amount of current the transistor(s) can handle while staying fully saturated? ...For example: when you decrease the load resistance, thus increasing the current flowing through the transistor and the load, the voltage across the load begins to drop some (in other words, this is causing a larger voltage drop across the transistor(s), in which, ideally you want it to be as low as possible, and as close to a strait conductor or switch as possible). Increasing base current doesn't work, and you can only add so many transistors in parallel before it really doesn't work, just 2 is enough, much more than 2 doesn't help much any more... any ideas?

And this is with the collector pushing/pulling the current? If so, then it is because of the parasitic resistance of the transistor and you can imagine a 1ohm resistor in series with each of the transistor’s three terminals. You can change that in the transistor settings to model a higher power transistor. I generally use 10mOhm for the emitter, 100mOhm for the collector, and leave the base untouched at 1ohm. Also, high power transistors tend to have lower gain, so maybe a gain of 40. I’m really just copying all of this off of @thebuggers various amps in his series of Hi-Fi amps. Also, if your doing a darlington/sziklai (collector of an NPN pulling from the base of a PNP) pair, the first transistor might be higher gain, so maybe 300 or 500.

If you make the above modifications, then you will need to increase the 1ohm resistor to 2ohms to make up for the missing transistor resistance.

Thank you.

Is there any other way, (other than altering the settings of the transistors)?

By the way, I bread-boarded this circuit and the last one... the last one, I couldn't get to work at all (but I also hardly had any time to troubleshoot it, it could have very well been just a wiring error). As for this circuit, it was already latched on as soon as you powered it on through like a switch for example (unlike the simulation, where you have to press the button first); and then one transistor was getting way too hot, and the circuit wasn't tripping. However, after troubleshooting it for like maybe 10 minutes, I found that: 1. The transistor was getting too hot because the potentiometer I was using must have been turned to its lowest resistance from the last time I used it, so it was biasing the next transistor with way too much current. (This actually wasn't hurting the next transistor because its base was also being pulled to ground by the other end of the pot. But it was causing the first transistor to load down through the base of the other with a resistance of probably less than 1 Ohm through the pot, and then the 0.75 volt drop of the 2nd transistor). 2. After I tuned the pot closer to ground (less resistance to ground/ more through biasing the next transistor), then that fixed the tripping and latching issue. But it's still not the best fix, what it really needs is a better negative feedback network, cause you can re-latch the circuit on, after its already tripped, JUST BY TOUCHING THE BASE of the two PNP transistors. >>> Welcome to real world problems that you don't see in a simulation😅 haha!!! ....Finally, after all that troubleshooting, the circuit was working flawlessly! Though, I still have yet to do much deeper testing and analyzing of my little circuit, but I intend to, then I'll see if it satisfies my expectations, if not I'll have to make more improvements. Also, I'll update this circuit to include the fixes I had to incorporate when I bread-borded it.

So, you mentioned one transistor getting too hot, but there are two power transistors in the circuit. That’s a common problem when using two semiconductor devices in parallel. What happens is if one heats up slightly more than the other for whatever reason, then it’s gain increases causing it to hog the power causing it to get hotter causing it’s gain to increase further causing it to take more power, etc. until it blows. Then, because the other transistor is handling all the power when it was only meant to handle half the power, it has the same fate on it’s hands. One fix is to add a 100mOhm resistor to their emitters to give some negative feedback. If one hogs all the current, it’ll have a lower emitter-base voltage and the current will drop because of that letting the other transistor take its fair share of current. If your using a lower current high voltage application, then you would need a bigger resistor. If you were using a low voltage high current application then you would need to use a smaller resistor. But, 100mOhm should be fine in this case, and most cases. That’s quite interesting about the touching the base of the PNPs to re-latch the circuit. Not sure why, but I think it might have something to do with static electricity and parasitic capacitance powering the PNPs sufficiently to latch the circuit.

No, neither of them are power transistors, they're just for standard applications (the part numbers are 2N3904 and 2N3906 if you want to check the data sheets), which is why I put two of them in parallel, and it wasn't either of the two in parallel that were over-heating, it was the one that's buffering the voltage drop produced by the 1 Ohm resistor, and I fixed the problem by adding another resistor to the center tap of the potentiometer. I get what your saying about the two PNPs in parallel possibly drawing more current than the other and creating a chain reaction of overheating called thermal runaway, but if I really wanted to eliminate that all together, then I would just bias them individually with a separate resistor, rather than trying to purchase a less common 100mOhm resistor online.

I believe the touch sensitive issue was happening because part of the circuit is fed back into itself so that it "latches" on after you press the push button... and after its tripped, it's in a state of equilibrium, where anything you do to interrupt the circuit, knocks it in/out of that state, and that makes it very sensitive.

If you think I'm wrong, let me know, and tell me why... I'm not afraid to take a little advice or criticism where I'm always open to learn something, but I'm not completely clueless here either, and I do know what I'm doing. And I know everybody sometimes makes mistakes

Both latched and unlatched states should be quite stable given the strong positive feedback. So, it shouldn’t be a case of just-barely-stable-but-on-the-brink-of-unstable-equilibrium that is knocked out of balance by simply touching it. In order to make a noticeable effect there has to be a connection between two parts of the circuit via your body. But, since the only part of the circuit touching the outside world is the single wire your touching, there is no path for electricity to flow. That means that either the pressure of your touch is destroying/creating a connection where there should/shouldn’t be, or static electricity combined with body capacitance is able to provide enough energy to power the PNPs sufficiently to re-latch it, similar to the button providing energy through the capacitor, except it gets amplified a hundred fold by the gain of the transistor, so 10nF at the same voltage, or 100pF at 100 times the voltage, which could easily happen via static electricity.

Anyway, that’s just my thoughts on it. Perhaps the feedback isn’t as strong as I think, but I’d imagine it’d be easier to trip it than to latch it. And anyway, 100pF body capacitance at 1kV at the PNP base should be able to latch the circuit assuming the correct polarity of voltage.

Hmm, ok