Without going way down into the details of the whole project, I've got a ProMini that's sitting on a spinning plate, physically independent of my main Arduino which is a mega. The plate the ProMini is sitting on is Macrolon, and therefore see through.
The ProMini will be using bluetooth so send data back down to the Mega, so the Mega can spin up or spin down that Macrolon plate using a stepper motor (and a driver of course, bought them both from Pololu).
Well actually I guess that is a lot of the details
Any who, I'd like to turn on the ProMini remotely. As in no wires up to it (we though about one of those rotating joints that transfers data, but not reliable enough), and I can't really use bluetooth signal until about 60 seconds before go time. Alright all the darn details. It's going up on a rocket and I want to reliably counteract the spinning of the rocket.
So, I was wondering there is a slick way to turn power on for the ProMini up on the plate? The Mega will get a signal at T-60 seconds to launch, and can turn on, but that doesn't do the ProMini a lick of good since it's physically separate. We also can't really leave the ProMini powered up (weird, NASA doesn't like RF signals near their rocket fuel???!?!?!).
My initial thought, was the Mega gets power, we shine a bright white LED, that hits a photo cell on the ProMini (since that Macrolon plate is see through), and that closes a circuit for the ProMini's power system.
I think that won't work, mostly because of our testing phases.
The test engineers at the launch facility will want us to be able to turn on and control any part of the system at will. We'll get push back if our answer for startup is, "Just spin it!"
You're thinking is spot on though! It's a 30 foot long rocket and it will certainly cause centrifugal force.
They will be able to supply the power up signal to the main Arduino Mega, and if we have a system that is easier for them to test, the more likely we will be to not get cut a couple days before launch.
Try your approach then. Use an optocoupler with NPN output to pull a P-channel FET gate low to let current flow from source, promini starts up and holds the gate low in parallel to keep power on. After some time, the promini can release the gate to cut the power again.
You must find out yourself whether a photo cell can provide enough power for the Mini. Check with different light sources (frequencies), depending on the spectral ranges of the emitter, plate and receiver.
For turning on a separate power source, a photodiode or similar receiver should be sufficient and more efficient. Then the Mini can keep the source locked on as long as required, in parallel to the light receiver.
I don't fully understand the on/off test requirements. Using white light may not be a good idea, unless the test shall be performed in the dark . No problem in the fully assembled rocket, though. A test command can be added to the Mega, causing it to light the LED. This command may have to be restricted to test mode only, when there is a risk that it could turn on the Mini by accident. Ending test mode then also should power off the Mini, somehow. A test without the Mega doesn't make sense to me, it would require another data transceiver, to which a light source can be added as well.
I also wonder why you don't use light (IR?) transmission instead of bluetooth?
If the power source of the mini is on the same side of the plate as the mini and is spinning as well, you could use a transistor and a ldr in a simple voltage divider to detect whether an LED that the mega controls is on or not. A downside to this is that even if the LED is off, some power will still be used through the divider. Also make sure the transistor can take the current load of your setup as it will be controlling the supply of the mini (and possibly the motor(s)).
CrossRoads:
Try your approach then. Use an optocoupler with NPN output to pull a P-channel FET gate low to let current flow from source, promini starts up and holds the gate low in parallel to keep power on. After some time, the promini can release the gate to cut the power again.
I am going to try that, but I may switch to a photodiode like Dr. D suggests.
Paul, I actually did! I was really worried about the Dopler effect causing a ton of packets loss, but in the lab, on a spinning plate (4Hz rotation was the highest we could reach, admittedly, when we finish the build, we'll have to push that 5.5Hz to ensure everything will work on the rocket). So far so good. Granted I'm just the code monkey/worker bee, and would so rather not be using radio signals at all. My original preference was a slip ring.
L3gend, current won't be an issue I don't believe. The mini is just grabbing data from two sensors (gyro and accell), and throwing it through bluetooth back down to the mega.
