Pete's Log: Diagnosing Circuits
Entry #1973, (Smokepacking)(posted when I was 43 years old.)
I breadboarded a circuit and it didn't work right. I was able to reduce the malfunctioning part to this sub-circuit. When I press the button, the LED lights up as expected, but only for a second or two and then it goes out again.
I didn't make it any further in my diagnostics before I had to resume chasing JB around. But I kept thinking about it. And after a while it occurred to me that there should probably be a resistor connecting the button to the transistor base. Sure enough, after bed time I was able to test it and that fixed the issue.
So the problem is I still very much think of things in logic levels (i.e. low or high voltage) and often forget to factor in current. So I think the problem here is that without a resistor, the current flowing through the transistor base was enough to glitch out the power supply after a couple seconds.
I'm real excited that the answer came to me while doing other things and I also hope actively thinking and writing about it will ingrain the lesson in my mind. Remember the current!
In other news, I designed and ordered another PCB already without waiting for the first one to get back. Bold, I suppose. But this one is less than a square inch in size, so including shipping it was less than $5 for three boards. So it seemed worth taking the risk. This one has vias and a copper pour so that feels fancy. More details to follow.