I was initially looking at a similar circuit where the latch pin would need to stay high. The disadvantage of that is that the momentary switch needs to be held for a couple of seconds to let the arduino turn on and set the pin high. The advantage of the circuit that cozendey posted was that it did not rely on the latch pin to keep it on and so the button press could also be short. I was able to get the circuit working that relied on the latch pin (we'll call that the sghioto circuit), but I couldn't get the cozendey circuit to work. Cozendey had a video demonstrating it and it made sense to me, so I was just assuming that I hooked it up wrong on the breadboard (although I checked it several times over). The sghioto circuit is not bad and I can use it if I need to, but I do think the cozendey circuit is more ideal for my use case. I do appreciate your input and your changes definitely take me from a state of "not working" to "working", which is very helpful.

 

momentary switch needs to be held for a couple of seconds to let the arduino turn on and set the pin high.
[/QUOTE
Why such a long time? The first line of code should set the pin high.
So what was the problem before when you had 5 volts present when the switch wasn't pressed?

 

The first line of code did set it high, but the arduino is slow, so it took a couple of seconds to "boot up". I imagine that a faster MCU would run the first line faster and this would be less of an issue, so I'll probably go this route. The problem I had with the cozendey circuit is that the switch was irrelevent. As soon as I supplied power, the output to the arduino was 5V, but it should have waited for the button to be pressed.

 

As soon as I supplied power, the output to the arduino was 5V,

Try adding a 100nf cap across the 100K resistor on the P channel mosfet.

 

If you remove the Arduino, and leave just the circuit, the n-channel mosfet still conducts as soon as supply power is on?

 

Yes, that is correct. I haven't had a chance to try the capacitor yet.

 

You could also try to put a diode just after the +5V, pointing towards the P-channel mosfet, so that the power supply can never act as a ground activating the p-mosfet, even if there is some transient state when it becomes on.

 

I tested the circuit on my breadboard and it works fine without the cap and the 10K resistor on the N channel mosfet.
Are you sure you have the P channel installed correctly or have the N channel where the P channel goes?
What is the part # of the P channel?

 

I'm fairly confident that I've done something wrong...I just don't know what that is. I'm using these mosfets: https://www.adafruit.com/product/1794
https://www.adafruit.com/product/355

 

Wow! Do you think those will handle an Arduino? LOL Did you get them mixed up?

 

The description on both of them says "...so you can control it directly from a microcontroller running on 5V logic.", so yes, I did think I it would work on a 5V MCU. I realize that their maximums are overkill, but what spec is not going to work with the Arduino? These work with the "sghioto circuit". I've triple checked that the mosfets are in the correct location (not mixed up).

 

Have you tried the capacitor?

 

Should the capacitor be like C4 or C5 in the diagram below?

 

Like C5.

 

The capacitor didn't change anything. Still 5V right away before pressing the button. Pressing the button has no impact.

 

If you have a few minutes and a volt meter I can probably figure out what's wrong.

 

I think mosfets with lower specs would be a good bet, too. These power transistors would be a better fit for the part that there is 12V involved.

 

I have some lower spec ones coming in the mail in the next couple of days. These from adafruit are all that I had on hand.
sghioto - what did you have in mind? Maybe I should wait until I get the others in the mail and if I'm still having issues, I can let you know.

 

sghioto - what did you have in mind?

If I ask you for some voltage readings in real time that would help

 

I was able to get it working using this schematic:

Thank you for everyone's input.