Hi, Since current DC adapter we used is 5V +/- 5% 2A rating output, M1 board can not be exempted from the user attempting a reversed DC voltage suddenly to the damage of the product even using official adapter, or even over-voltage condition from others.
I did experiments to verify the functionality while reversed polarity input by different schematics made of variant parts : http://en.qi-hardware.com/wiki/Protection_of_Reversed_Polarity_on_DC_plug-in The wiki page is not only trying to help M1 but also being as a brief introduction on how reversed polarity protection could be taken. Sch. A - RC2 design. Sch. B - added one fuse and one diode(which placed in parallel with load). Sch. B-1 - added one fuse and one Zener diode (which placed in parallel with load). Sch. B-2 - added one fuse and one diode in series. Sch. C - added one fuse and one N-MOSFET. A type of fuse can be one time trip or resettable/recovery fuse depends on what feature the board you would like to have. The performance will also be quite different. A voltage drops on fuse, you can read out from the results. A type of diode can be a traditional diode or Schottky barrier one (which has low forward voltage drops). When using a series loop with diode and fuse. A Schottky diode is better. Especially on limited voltage source like if mm1 uses a 5V +/- 5% adapter. So a 4.75V is experimented. For example Sch. B-2, at best to use a Schottky diode. Sch. C is also have very low voltage drops on its Rds(only 32 mΩ @ VGS = 4.5 V). Sch. B-2, I didn't do experiments because I can guess/imagine the result is quite same as Sch.C. A Zener diode (5.1V / 1W) I just bought from normal shop without datasheet in details. Not sure what its Zener Current to check. As well-known, a limited current resistor is necessary to be as power dissipation purpose. See the results Sch.B-1, the slope on breakdown performance seems that not good(it's 5.314 V when input increasing to 5.5V, it should stay 5.1V roughly). A sharper slope is better. This should do further experiments using 3W with good sharp slope spec. Using this sch, a total and more current goes through PTC fuse to exactly increase its postive resistor to trip whole board no matter using reversed polarity or non-reversed one. When reversed input, the negative current immediately goes high depends on adapter's rating value. Then it later stays at 0.3A roughly by an increased resistor from PTC fuse got warm. A non-reversed input, let Zener diode and PTC fuse be the role even a stupid over-voltage coming. Zener stays 5.1V then immediately trip simultaneously. Value of used part needs to be fine tuned though. When input using a 5.00V, Sch. B-1 have good results better than Sch. B, see the voltage at symbol 5V marked on M1 is : 4.743V v.s. 4.681V. With regard to mm1 solutions: 1, we can not use a one time fuse definitely. 2, voltage input currently from 5.25V to 4.75V, I would like pick Sch. B-1 which is possessed of over-voltage protection as well as the voltage at symbol 5V marked on M1 is acceptable. Actually there's still a way to fix the over-voltage problem: using another regulator to accept a wider range of input, said as defined: Input Voltage (recommended) 7 - 12V Input Voltage(limits) 6 - 20V but through this way, we must run into a messy situation on variant adapters. It's an bad idea though but it's schematic is just made of Sch.B-2 with a traditional diode and a new regulator. Comments are welcome. -adam
_______________________________________________ http://lists.milkymist.org/listinfo.cgi/devel-milkymist.org IRC: #milkymist@Freenode Twitter: www.twitter.com/milkymistvj Ideas? http://milkymist.uservoice.com
