Power Controller v1 Assembly

Those with a keen eye may notice that this is now called the power controller v1. The boards arrived a few days ago and I've been working slowly through populating the first test board and testing as I go. The trip-reset logic worked will despite some teething (that's on me for trying to experiment with different component values). The master control relays and associated logic was installed and also worked well. Then came the override and bus control logic.

When testing the override logic it was immediately tripping the board when the bus was put in the ovrd ON position. After some pondering I realized that I rely on the sim controls to pull-down the signal to ground. When I designed it I thought it made sense, let the sim controller do the pull-down to simplify the board, after all most boards have integrated pulldown resistors. I forgot that if the sim isn't connected at all you lose that functionality. Anyways no big deal, just make a dummy jumper or something that you plug in if the sim isn't connected for maintainence or whatever, right? Well, yes actually. But then I realized that wasn't the only issue with this setup. Not only would it trip if the sim control input was floating, but if you drive it high at the same time you are in the ovrd ON state it also trips, and that is not quite as easy to solve.

So in short, the issue is that the L293D (the relay DIP used to control all of this) doesn't really do reverse-biasing protection. In this one configuration you are "closing" the relay and applying power to the output while the input floats. It turns out this makes the L293 unhappy and causes the board to trip. Adding a diode to the output would fix this, but the trace is integrated into the PCB and there's no place to intercept it, so there's no way to bodge the board and fix this short of scraping traces, which kinda defeats the point of having a nice robust PCB.

So where are we now? Remember that v1 I mentioned? Well that's because I'm now working on v2. The v2 will fix this issue along with a bunch of other minor things that I found when assembling the board. A diode is the easy fix, but I'm also working on a different design which instead uses individual TO-92 package transistors, probably BJTs because I don't have any PNP MOSFETs on hand and don't feel like buying them just for this.

The current state of the project is working on prototyping this new logic and then will update the board design and incorperate a dozen other QoL things along with it. My desire is that I can get this turned around and re-tested in about 2 weeks. I'm actually a little relieved that a v2 is needed because it gives me more ability to refine my PCB design process and learn more about how to make a PCB that is actually good. My next project is the audio cards which are analog (ugh) and will have dozens of examples built. A minute of optimization now will save me an hour of frustration on that project.

So anyways keep an eye on the power controller v2 progress!