Re the question on maximum currents: the safest thing is certainly to go by the published tables that say things like an 80 mil outside trace is good to 4A. But that's actually based on some very conservative assumptions...
To be precise, what they say about an 80 mil trace at 4A, for example, is that it'll be in thermal equilibrium at 10 C higher than ambient (which they call delta-T or dT). If you figure an ambient room temperature of 75 F, that means the traces will be at about 95 F. I imagine things start getting warm to the touch at maybe 100 F (maybe a little higher for such a small surface area as an 80 mil wide trace). So keeping to the 80 mil = 4A from the tables should keep the board from ever feeling warm, which is a reasonable benchmark, but it's extremely conservative in terms of actual safety. Copper's melting point is about 2000 F, so we have a bit of headroom above 100 F before the trace starts acting like a fuse filament. Although something else would probably fail before the copper melts, probably either the glue holding the copper to the board, or the solder connecting one of the components. But there's a lot of room between 100F and 2000 F. You could probably be perfectly safe pushing the design benchmark up to, say, 125 F, which would give you a dT of 25 C, which would let you carry 6A on those same 80 mil traces.
The other interesting thing to note is the tables are all about "thermal equilibrium". The tables aren't saying that the trace will immediately heat up to such and such a temperature; what they're saying is that it'll be in thermal equilibrium at that temperature over infinite time. That's actually the key to why (say) 80 mil at 4A is "safe": you know that the temperature will reach equilibrium at dT = 10 C no matter how long you leave the power on at that current, so you don't have to worry about transient temperature characteristics; it will always be at the equilibrium temperature or cooler. What's never been clear to me is what the time scale is before you reach equilibrium; the published tables don't include that information. One interesting paper I found* suggests that the time scale is on the order of minutes - maybe 2 minutes to reach 75% of the full dT rise. At that kind of time scale, I think the sort activity you'd see in, say, a pinball chime unit, slingshot, or bumper, could be treated as though it were a sort of PWM signal, by which I mean that you could treat a 50% duty cycle as 50% of the peak current, for the purposes of calculating trace heating. So if you had a 6A bumper solenoid that you wanted to run on one of the 80 mil traces, you'd be within the original conservative 10 C heating profile as long as the bumper wasn't switched on more than 67% of the time.
* https://www.ultracad...es/pcbtempr.pdf
Edited by mjr, 15 August 2019 - 05:36 PM.




Top













are all trademarks of VPFORUMS.