Weatherproof Power Supply
August 21, 2008 5:35 AM   Subscribe

A probably-out-of-his-depth n00b asks: Weatherproof, "high" wattage electronics power supply?

I've got a solar measurement project I'm working on. Up until now, I've put it outside when I have a few hours to tend it, but I'd like to leave it out unattended and full time. I believe I have the mechanics and some of the electronics all worked out (I have a small model more-or-less working in breadboard mode). I originally thought the main problem here was the weatherproofing. But I think if I use a tightly sealing plastic container, well slathered with sealant and put the entire electrical part under the main body of the tracker, I should be good. Motor shaft will have to poke out somewhere, but it can remain sheltered under the tracker.

The real problem is the power supply. For the sensing part, I could just use a battery and then change it on a weekly basis, say. But the motor to move the solar tracker requires more power than that.

The way I've designed it only requires my stepper motor to take a few steps per day and it shouldn't need to be on and "locked" the rest of the time. The total power draw is low. But in order to move the large bulk for those few steps, the max amperage needs to be much higher than a 9V battery can supply. I haven't measured it exactly, but I think 1.5-2A. It's an ~8Ω motor, so the supply needs to be 12-16V.

One idea is to run an outdoor extension cord to a sheltered spot, plug a modified wall wart into it and then go from there. (The Arduino can take a max of 12V, so I might need some kind of voltage regulator in the mix to step down from 16V....?)

Another idea that I had recently was to repurpose a computer power supply. Run the plug into a sheltered spot and then use various combinations of the voltages available to run both motor and sensors. I even got a power supply to hack for this, but it's going to be some work to remove the spaghetti of extra wires and also build a special connector (I guess) to take the 20 pins and distribute them correctly. Messing around last night I also realized that heat may be a problem, at least from a computer PSU.

Thinking this must already be a solved problem, I did some googling this morning. I found a site that sells all manner of power solutions. For instance. But am I going to be able to plug that into the wall myself? And what about the heat problem?

Or maybe I don't need to be plugged in? (I'd sure be happier if that were the case.) With only momentary currents, maybe I can get by with, say, a dozen 9V batteries in two parallel banks of 6? This may especially be the case if I could get the motor amperage requirements down using mechanical advantage. I bought a few small pulleys but then decided that it was easier to "just increase the power" rather than attach all these pulleys, run the cord around them, waterproof all these mechanics, etc. Now I'm thinking that might be wrong after all.
posted by DU to Technology (11 answers total)
 
Best answer: I think 1.5-2A. It's an ~8Ω motor, so the supply needs to be 12-16V. [...] Or maybe I don't need to be plugged in? (I'd sure be happier if that were the case.) With only momentary currents, maybe I can get by with, say, a dozen 9V batteries in two parallel banks of 6?

What about a 12v sealed lead acid car battery? They're designed to provide large, brief currents to turn starter motors, so a normal car battery is good for a few hundred amps for a few tens of seconds.

You can get smaller volume batteries designed for jetskis etc if you want to get it into a tight space.

You can then just use a 5v regulator to power your sensing electronics from the same battery. Remember to put a fuse and a power switch between the battery and everything else.
posted by Mike1024 at 6:12 AM on August 21, 2008


Are you sure you can't drive your motor just with a couple of 9Vs? It'll be a little slower but should work fine, just incorporate the delay into your microcontroller's code. Easy to test, anyway - run a couple of wires from a 9V's terminals to the motor's terminals, see if it spins. Either way, yes, you really have to throw a voltage regulator on before your give the Arduino voltage - a 7805 should do it. That'll work for your sensors etc too. Have you worked out a motor driver circuit yet?
posted by nicolas léonard sadi carnot at 6:18 AM on August 21, 2008


Seconding the lead acid battery.
posted by caddis at 6:35 AM on August 21, 2008


Check out this Texas Instrument 6 amp power regulator, you could use a car battery and adjust it to 9Volts. I use rechargeable portable DVD battery packs, they claim to be 9Volts, but are really 8.5, so I use this regulator. They'll sample you one from the TI website, just by clicking.
posted by StickyCarpet at 6:43 AM on August 21, 2008


Response by poster: Have you worked out a motor driver circuit yet?

Because I am an idiot who reinvents every wheel he can, I have a dual H bridge that I'm driving directly from the Arduino. I should probably have used a stepper controller, but whatevs.

...run a couple of wires from a 9V's terminals to the motor's terminals, see if it spins.

I think even a single 9V will make it spin. In fact, I've made it spin using the power just from the Arduino. The problem is making it move the tracker. It looks like a 9V could provide 1.5A for maybe 5-10 minutes in total, so I guess that isn't totally unreasonable...

What about a 12v sealed lead acid car battery?

I did think of the car battery idea, but rejected it for some reason. I don't recall why now. It will be too large and heavy, but you are right that I could use a smaller one. How about something like a cordless drill battery? I'll have to check the rating.

Check out this Texas Instrument 6 amp power regulator...

I'm all over the idea of free samples, but I don't get what this is buying me. Why am I adjusting things to 9V?
posted by DU at 6:53 AM on August 21, 2008


Motorcycle batteries are 12V, smaller than car batteries, and come as either lead-acid or sealed-cell lead-acid kinds. They can produce Significantly More amps than you'll ever need.
posted by 5MeoCMP at 7:11 AM on August 21, 2008


Oh, sorry I thought 9 volts was your target. You can regulate to 12 volts, from whatever you have. Radio controlled cars have 7.2 or 9.6 battery packs that are readily available.
posted by StickyCarpet at 7:40 AM on August 21, 2008


I'm all over the idea of free samples, but I don't get what this is buying me. Why am I adjusting things to 9V?

This TI power supply has a variable output voltage, you are setting it to 9V (or 5V) to replace the 9V battery you are already using. That way, you can use the motorcycle battery to power the motor and also provide power for the breadboard.
posted by Daddy-O at 10:26 AM on August 21, 2008


Best answer: For less than $20 you can get a 12-volt sealed lead-acid battery used for home security systems. It is less than 6" x 3" x 3", weights about 2 pounds and supplies about 3.3AH. Larger sizes are also available.
posted by JackFlash at 10:32 AM on August 21, 2008


Nthing the sealed-lead-acid approach.

You could also use a stack of NiCd cells, say 10 AA or C cells to give you roughly 12v. NiCds can supply huge surge currents for brief periods. (In most other ways though they're inferior to NiMHs etc.)
posted by hattifattener at 10:50 AM on August 21, 2008


Are you sure you can't drive your motor just with a couple of 9Vs? It'll be a little slower but should work fine, just incorporate the delay into your microcontroller's code.

The thing you have to watch out for in this approach is: You draw a lot of current, your batteries' output voltage drops (due to their internal resistance), and the drop in supply voltage resets your microcontrollers. Problems of this variety can be difficult to diagnose!

Obviously, there are ways around this, but using a battery with a low internal resistance (such as a lead acid battery) is probably the simplest.
posted by Mike1024 at 1:23 PM on August 21, 2008


« Older Why has the dentist left blue marks on my teeth?   |   my friends suck Newer »
This thread is closed to new comments.