Protecting reverse connection of PIC, possible?

I was just wondering if, there is any circuit that can help you protect your PIC chip from getting damaged when you reversely connected the PIC? ( Referring to Vss -> Vdd pin & Vdd -> Vss pin)
If there is a circuit, could you share? Thank you.


P.S: Referring to currents up to 1 Amp.

Simplest way is to add a diode in series with the Vdd pin. Use a Schottky diode to minimize forward voltage drop across the diode, though you may still need to increase your Vdd supply voltage by a few hundred mV to compensate. If you can't do that, you could use a P-channel FET oriented backwards from the usual way (connect the source to your PIC Vdd pin, connect the drain to your Vdd supply). You'll need to add a reverse voltage detector circuit to control the FET gate. With this method you can achieve a lower voltage drop from Vdd to the PIC Vdd pin, at the expense of higher component count, size, & cost.

Reverse voltage detector circuit?

When you want to have a circuit which will protect from switching Vcc with GND on the power supply the simplest way in my opinion (tell me if I'm wrong) is the diode bridge
https://en.wikipedia.org/wiki/Diode_bridge
which will always give good polarity of voltage.
BR

I don't think that might work for the condition I want, ndru_w.

I was hoping more of a circuit where when the PIC is put in reverse connection, something, like a relay, can disconnect the PIC from the PSU (Power supply unit).

Thanks for the suggestions

By reverse voltage detection circuit, I just mean any circuit that will pull down the gate of the P-ch FET I mentioned when Vdd exceeds some threshold, and release the gate when Vdd goes below the threshold, as in the case of a reversed supply. The specifics would depend on the value of Vdd, the threshold voltage and Vgs rating of the FET you choose, the accuracy and precision with which you want to control the turnon/turnoff voltage, etc. The simplest method, if the FET ratings are suitable, is just to connect the FET gate to your circuit common. You could connect it directly, but I would do it through a resistor. Connect the FET drain to Vdd, and the FET source to the PIC Vdd pin. In this configuration when normal power is applied, the FET is initially off but its body diode conducts current into the FET source, pulling the load up. Since the gate is fixed at circuit common, the source rises relative to the gate. Once this voltage exceeds the FET threshold, the MOSFET becomes enhanced and shorts out the body diode, thus conducting with voltage drop Iload*Rdson. You should be able to choose a FET with an Rdson low enough that this voltage drop will be less than you'd get with a Schottky diode.
If you power up with the supply reversed, the body diode is reverse-biased, the FET channel never conducts, and the load remains isolated from the supply.

Hi,
-I think the bridge should be as universal solution the best:but in that case is not practicable, than it does GND separation...
-The relay isnt practicable as semiconductor protection; it has too slow reaction time...
-The FET-Protector; I dont no-whats the lowest working voltage for that pls?
The most PICs are working not with +5V, but fine lower as 3.3V or even less as 2.5V!....
-Also, the simplest way is in my opinion the best too_& practically the only way; a SIMPLE SERIAL DIODE. Or do you have better ideas pls?
K.

I agree that a series diode is the simplest, but you pay a penalty with the forward voltage drop. And at 1A current, this is likely to be a significant percentage of the supply voltage, especially if the supply voltage is low as you suggest. If the supply can be increased to compensate for that loss, then I agree this is probably the best solution. If not, the FET solution can be adapted to work with an arbitarily low supply voltages by introducing an auxiliary supply voltage if necessary and/or increasing the complexity of the gate control circuit.

**broken link removed**
http://www.irf.com/technical-info/designtp/dt94-8.pdf
**broken link removed**
http://focus.ti.com/lit/an/slva139/slva139.pdf
**broken link removed**

Thank you all for sharing.
I'll try it out with a PIC.

Hi Shinnster,

Is it not the best way to design in a polarized connector to use!?
Is it a problem pls from a serie product or from your experimental unit?
By serie product I dont understand the problem, than for that is all fixed-or not?
Than, a PIC dont need 1A, but if you have a better Schottky selected it will have max at 0.3-0.5V by tenth of A currents...
Than thee most wall adapter can not deliver very exact voltages :-(
K.

Diode, most people recommended that, that would be my first try.
P-Fet would be my second. I am just doing this out of interest. This could be beneficial for some people who are really clumsy in dealing with PIC and sometimes connected it reversely.
I'll try whatever suggestion people give me, if deemed reasonable.