Modernized Engine Wiring
So why not just a straight up shunt? Why piss away .7v in a diode that can fail?
B.
Valid concerns, since if the diode opens, all current goes thru the dash and ammeter. However, I fuse those to protect them. Not sure what happens to the alternator if open-circuited, but I suspect it wouldn't hurt it. If it did, an alternator is easier to replace than the instrument cluster. Diodes usually fail in short-circuit, though mine does have a series fuse.
You need to design conservative, which is why I used 170 A diodes. But you have to know what that rating means. At any current flow, you must calculate the power dissipated in the diode and how hot it might get. I think "170 A" means the current at which it will have the "0.7 V standard silicon drop", but someone can research.
A diode is much better than a fixed wire resistance because the diode resistance drops as the current increases. If a wire, the resistance would increase with current, due to the heating. The ideal diode to use would break-down at 0.7 V (where ammeter hits full-scale) and conduct all current beyond that to hold 0.7 V. This would be similar to a "pressure relief valve" in plumbing. Your fixed wire resistor is more like a fixed orifice in plumbing. There are such diodes termed "zener" (used in reverse bias), but I haven't seen any that could handle these currents. That is why I rely on a forward drop.
I wanted to keep my ammeter since I hated losing that in my 80's cars, and got stranded a few times because of the useless battery warning that only lit when too late. I don't care that the ammeter is accurate, I just want to know if charging or discharging, plus I feel better when all gages in the dash are functioning. I do have a voltmeter, a portable one I plug in the cigarette lighter ($15 Amazon). I find those invaluable in my newer vehicles that give no other charging status.
