I really dont see the point to a crowbar circuit in something that can use 78xx vRegs. Those have a huge input range (up to 36V!) and are damn near indestructible. Keep it simple: 12VAC or 15VDC wall wart, 7812 for the 12V Rail, switching 5V reg for the digital side. Done
I miss to see your point here I think. What do you think happens if the vreg fails because of whatever reason, and it fails in a welded style? Wow, u get unregulated jumpy high voltage on your rail. What do you do? U regret u didn’t throw a crowbar in there, coz that would kill the rail in an instant, and blow the fuse, and hopefully save as much as possible can be saved in both the load and psu. No?
Should I use resettable or regular fuses? Or both?
There are positive and negative with both. If a normal fuse blows, you gotta have it available to change it without having to open the psu case right? - or at least that’s practical way to do it. If you use resettable fuses, if I got it right, they need cooling time after blown. And it’s not really a circuit breaker / fuse, but close to it - when it trips, it makes the resistance through it so high that almost nothing can pass, but some. Someone will hit me on the head if I’m wrong, but this are things to be aware of. But resettable fuses are widely used in all kinds of electronics, so why not? :). Your choice :).
Grounds MUST be connected. If there’s a lot of noise on the 5V rail, you’ll get more noise on the 9V/12V rails.
Great clarification!
(there are better ic-based solutions out there, though)
Yeah pete, I remember I found some IC based solution a while back and fiddled a bit with it, but eventually put it on hold. The idea is that a zener diode solution as a referance voltage used to trip the thyristor is a bit unreliable because of heat and tolerances. It’s kinda the same thing as using vreg ICs instead of a zener voltage regulator and a pass transistor. I suspect (as u say yourself) that both of the solution will work pretty good.
I just finished up my 6582 powersupply - it got +/- 12V, 9V and 5 volt. All rails fused, all positive rails have crowbars. And I have been designing it with 2% tolerance zeners, and the zener voltage about .4 - .5 volts above normal rail voltage - that is the minimum zener voltage in the datasheet. So in reality the zener voltage is more like .4 to .8 volts above normal operational voltage. I have no idea if it works the way it’s supposed to, but I’ll try it out with some load and see what happens if I increase the voltage on the rails. Should give me a nice blown fuse pretty fast hopefully.