For fun & profit, I will like to design a cheap "droppable" VPN server that will be placed a mile or so away from a a known wifi hotspot, where the portable device would serve as a portable vpn server on the internet.
My prototype budget is about $150 US Dollars.
I do not see this has been done before, so here my idea for your review. Your experience & knowledge would be highly appreciated. I'm thinking of selling these when I can make them more cheap. I have not built it yet so I only are in the start of prototype idea stages.
PROTOTYPE:
RADIO.
COMPUTER.
POWER.
RADIO: a. Any small 12VDC wifi radio & router should work. b. Maybe the Ubiquiti NanoStation Loco M2 perhaps? c. Prototype cost is near $50 US Dollars.
COMPUTER: a. Any cheap computer board with Ethernet & USB should work. b. Maybe the Raspberry Pi Model B should work? c. Prototype cost is near $35 US Dollars.
POWER: a. Need only charges & supplies 12 volts should work. b. Maybe this Amazon 6W solar panel should work?
formatting link
c. Maybe a 12VCD gel battery should also work. d. Prototype cost is near $40 US Dollars for the solar panel. e. Prototype cost is near $15 US Dollars for the 12VCD battery.
I have not built this yet, before I go in wrong directions, I would enjoy to ask if you have better ideas for parts to improve specifications or to keep the prototype costs down.
Your helping ideas are much welcome ahead of time!
I don't know if this project makes much sense. Just what is on this server that you need to use from a mile away? Or is the goal to make a repeater and provide internet service? That is a totally difference project.
I would avoid the R Pi at all costs. They simply fail. Hand on two weeks and get a Beagle Bone Black.
When you design for solar power, you need to consider it gets dark at night. ;-) Further, unless you want to install a buck converter, I'd shoot for 6V operation.
Battery operation means you need a DC/DC regulator to drive the SBC. The most efficient way of doing this from a 6VDC cell is when the voltage is too high, you buck it down to 5DVC. When it is just right you use a pass gate and feed the raw battery to the board. A 6V gell cell should not be discharge any lower than 5V, so you will not need a buck-boost or sepic. An undervoltage lockout scheme is required to disconnect the battery from the load. Be sure to use some hysteresis since once the load is removed, the battery voltage will jump.
You need a charge controller. Potentially this could be bootsraped, i.e. you use the SBC to control charging the battery. But it is better that the charge controller work independently of the SBC.
Note your solar cells needs enough power to run the SBC AND charge the battery. The battery needs to be able to run all night plus twilight/early AM, and in periods when it is cloudy. Your code on the SBC needs to be rugged. Embedded linux is a bit more complicated than you think. Your code needs to start up by itself should the system barf. Better yet, you have a watchdog timer keep the system running. If you design the charge controller yourself, you can incorporate the watchdog in the charge controller. If the charge controller is uP based,, you need the SBC to perform a watchdog on the charge controller. Your code needsd to be a "service" as far as linux is concerned, so you better be good with systemd.
When you do a proper job, these projects are vastly more complicated than the back of the envelope. Further, you need to consider overload conditions and possibly an over temperature lockout. You wouldn't want to set your property on fire. Worse yet,it is not your property, you are kind of f***ed should anything go wrong. Usually anything built into an ammo cano or NEMA box is safe to deploy. Generally if the area is prone to freezing, the battery is put in a box and burried beneath grade, but with a hump in the soil to keep water to a minimum.
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.