Need a lot of power in a little space!

Hi guys,

I'm building a prop, and I've got a problem.

What you see here is an older model, which used a 555 timer and a 4096 chip to power a 7 led display. The 9v battery, which barely fit inside with all the wires once the top was down, was able to power that for a good 15 hours.

I am now building a new model which has 84 leds in it that will draw around 500mA of power. A 9v ain't gonna cut it. So, I've been looking at the possiblity of powering the thing with 6 AAA batteries, which would provide enough juice to get 15 hours out of it.

Problem is, those take up a lot of space. And as you can see, I don't have a lot of space.

The box you see there is around 4.5x3x1.25". Just tall enough to fit a 3x2 configuration of AAA batteries, but probably too narrow to fit 6 side by side, unless I orient them along the length of the thing.

I think one of those configurations will be okay, because the Pro Mini is a lot smaller than my original circuit board, and the wiring this time around should be a bit neater since I will be using ribbon cable to bridge between 10 segment led displays, but I want to be prepared for the worst case scenario where I have to reduce the number of batteries I have in there.

I'm thinking I have one of two options here. Either I can find a way to power the 5v Pro Mini off 6v, or I have to buy a 3v basic breakout board and a 3v Pro Mini, and run it off of that instead. (You do need a 3v basic breakout board to program the 3v Pro Mini, correct?)

Thing is, I'm not sure what the best way to go about dropping the voltage would be. I've looked into switching regulators but those are pretty expensive. I've also considered the possibility of using a diode to drop the voltage to a little over 5v and connect my power source directly to Vcc, but I'm not sure if that would dramatically reduce the amount of time I can run it off the batteries. I may be no better off than using the 9v if I do that. On the other hand, I'm concerned about the 8mhz Pro Mini's ability to drive my array, though since I'll be switching between only 8 columns of LEDs at 60hz, it'll probably be fast enough. (Really didn't want to have to spend money on another breakout board though.)

Anyway, I'm interested to know what you think would be the best way to proceed here.

Hehe... Ghostbusters?

Have you considered using a Arduino Nano instead? It has a regulator on it and so kind of side steps the whole issue.

Also... I'd think a standard LDO Regulator would probably get the job done... switching probably isn't neccesary.
http://www.ti.com/lit/gpn/REG104-5

Another option swould be to use Lithium Ion rechargable batteries. You can get 7.4v out of 2 cells that way.. a 7.4, 2600 mAH packs is about 3" x.75" x1.5" and would go for around 5 hours I think.

Check out RC Hobby stores or sites for Lipo battery packs for ultra micro RC airplanes and helis. The size is relative t the power output, some of the smallest wieghing in around a gram. Pretty amazing. You will have to figure out you exact power requirements but Im guessing you will need one or two about the size of a ketchup packet . I dont think you can go much smaller than this.
You will also need a dedicated LIPO charger that is appropriate for the current output of the batteries you get. Do a little research, Lipo's can be dangerous when damaged or not charged properly. You should be able to include a charge port on your project, then you can plug in the charger (like a cell phone) and not have to remove the batteries. or buy a couple of packs and swap out the dead ones for charging.

Can't use the Nano or LiPoly batteries for this. I sell these, so I need to keep my costs down, and I can't make people buy expensive chargers just to power their toy.

I took a look at that reg104-5, but that's SMT.

Would one of these do instead?

I tried to narrow the search down. I think my output current is gonna be around 500mA. So I think I want something that can handle 1A maybe to have some breathing room? Or do I want more? I'm not sure if those are maximums or not.

And will heat be an issue with these at all when I'm stepping the voltage up, rather than down?

How about using a pair of 9V batteries in parallel. Looks like a modified handle could easily hold the batteries. Should be able to offer the current needed and last a few hours.

Handle is a piece of PVC pipe, and the 9v won't fit in it. There's no real good way to put a battery in there anyway. It's bolted onto the body and the bottom isn't removable.

As for using a couple 9v, Wikipedia says they've only got around 625mAh each, and the circuit will be using around 500mA. So that comes out to around 2 1/2 hours of run time, which isn't a whole lot.

I just thought of something though. I'd been doing my calculations for run time based on adding the mAh. So for 6 AAA I calculated that I'd get 7500mAh.

But that's wrong, isn't it. You only add the mAh if you connect them in parallel. If you connect them in series, as I would be doing here to get 9v, then the mAh stays the same.

That means connecting two 9v in parralel might be the best solution after all. Two 9v in parallel would give me 1250mAh, which is the same as six AAA in series.

Hm.

I guess that leaves me with three choices:

  1. Use 2 9v or 6 AAA's and accept a 2.5 hour run time, which might just be sufficient for a day's use.

  2. Switch over to the 3v Arduino Pro Mini, which would give me more options, like running two sets of 3 aaa's in parralel.

  3. Find some way to cram six AA's in there.

Ya know... the datasheet lists 6.0v as the max operating voltage. Maybe you could do 4x AA cells.. cheap enough and should fit the case.

With package fresh alkalines you'd be at 6v, and with nIMH you would probably get as low as about 4.5v as a minimum if someone wanted to use those.

You probably want to wait until someone without "Newbie" by thier name verifies or disputes my guess, but I'd bet you could run it from 4xAA cells and it would do just fine.

Or, it could be more voltage sensitive than I think and it would be a horrible idea...

Ohh, and AA Nimh batts are around 2600 mAh so you are back at about 5 hours runtime.

I think connecting 4 AA's directly to Vcc is probably a bad idea.

First of all, I'm pretty sure you don't want to run the device at it's absolute maximum ratings for any length of time.

Second, even though the batteries say 1.5v, I'm guessing it's possible they may put out a little more or a little less than that. Which means you might exceed 6v.

Perhaps if one connected a diode in series to drop the voltage a bit though, one could safely run off 4 AA's?

http://www.kpsec.freeuk.com/components/diode.htm

A silicon diode has a forward voltage drop of 0.7 and can handle 100mA of current. That would be sufficient to run the Arduino, but insufficient to power my circuit which will need to draw much more than that.

Hm.. but then there's that zener diode at the end there. I'm not sure why you are supposed to run them in reverse, but they seem to be designed for just this sort of situation. I guess I should research those somme more.

Thing is, even if I can use the zener diode to drop my votlage from 6v to 5v, and it can handle over 500mA, I won't know what my run time will be like, because a batteries voltage drops as it is used up, and that means if I start out at 5v, by the time I've used the batteries half up maybe it's down to 4.5v and it's no longer enough to power the Arduino. On the other hand, maybe I'm not understanding the Zener diode right, and it will maintain 5v for me so long as the batteries can out put at least 5v, which means they could drop a whole volt before they stop being able to provide a full 5v.

Hmmm, the Diode may be a good answer, just for safety's sake it's not a bad idea to protect against reverse voltage.

I'm pretty sure Alkalines top out at 1.5v a cell... but you are right, I'd put a multimeter on a few to be sure, LOL.

Zener diodes are hooked backwards because they have a "Breakdown" voltage.... meaning like a normal diode they stop reverse current, but only until they reach that magic voltage point, then they start letting the juice flow.

Also... your LED's can run right off the battery voltage, so only the Arduino would see the voltage on the other side of the diode. (And the current would be minimal.. 100 mA is probably plenty if you aren't trying to direct drive the LEDs.

I don't think you can direct drive 500 mA of LED's right off the Arduino so you would have to use a transistor or something to feed that many LEDs anyway.

(The Arduino is rated for 40 mA per pin, but it has a package maximum of either 200 or 300 mA, I don't remember which.)

Yes, I am using a transistor to feed the leds 500mA. But I don't think I can feed the leds directly off the battery, precisely because I'm uisng a transistor.

With the transistor I've got a loop from battery through arduino, through transistor, and back to the battery via the leds. Also from battery through transistor, then back to battery via leds. I don't think one of those loops can have a different voltage than the other. And if they can, I have no idea how to make that work properly.

So I've looke dup Zener diodes, and I found this page dewcribing their use:
http://knol.google.com/k/electronic-circuits-design-for-beginners-chapter-7

I also found these 5v Zener diodes:
http://mouser.com/Semiconductors/Diodes-Rectifiers/Zener-Diodes/_/N-2xpqt?P=1z0w4d5Z1z0z819

But they can only handle 500mA max, and since I need 500mA, to be safe I think I should get ones which can handle more. So I found these:

4.7 and 5.1v zener diodes which can handle 1000mA. I think the 5.1v ones are the ones I want. I think the Arduino spec sheet indicates up to 5.5 volts is okay, with 6 being the absolute maximum. 5.1 is pretty darn close to 5v, so I think it should be safe and I probably don't want to feed the Arduino less than 5v if I can help it.

Still working on doing the calculations on that page though to see if this will work out or not.

So based on the calculations on that page I come up with:

6v in, need 5.1v out at 500mA, so:

Vr = Vin - Vout
Vr = 6v - 5.1v
Vr = 0.9v

Rl = Vr / Iz
Rl = 0.9 / 500mA
Rl = 0.0018 or 1.8 ohms

But wait...

If I use this diode:

Then the current the Zener needs to maintain the voltage is 49mA, so:

Iout = Ir - Iz
Iout = 500mA - 49mA
Iout = 451mA

So I can't take more than 451mA without the voltage dropping.

If that's the case, why did they have me do all those calculations for 500mA? I mean really what I want to calculate my resistor for is 500mA + the zener current, which here is 49mA.

So redoing the calculation:

Rl = 0.9 / 549mA
Rl = 0.001639 or 1.6 ohms

So if I build this circuit:

With a resistor less than 1.6ohms, (but not less than 0.9 ohms which would put over 1000mA through the Zener) and this zener diode:

And connect that to a 6v AA battery source, then I should be able to get out 5.1v and over 500mA.

Now, for how long I can do that, I dunno...

Also, I feel like I'm missing something here. I thhink one of the links I looked at mentioned zeners wasting power. I wonder if I am not pulling that 500mA constantly if whatever's left over gets turned into heat. :confused:

Heyyy... wait a sec... you are using a Pro Mini?

LOL... it has a Voltage Regulator built in... sorry about that, I should have picked up on that sooner.

Input Voltage 3.35 -12 V (3.3V model) or 5 - 12 V (5V model) <-- Right off the system specs.

There ya go.

You'll still want to run your LEDs with a transistor.

I know the Pro Mini has a voltage regulator built in, but I'm pretty sure that if you use the raw input pin then you need a minimum of 7.5 volts.

I can't rememebr where I got that number though. I think it was from the forum.

Here's a thread with interesting info.. including the datasheet for the Vreg.

http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1265675585/8

First, it's only good for 150 mA, so do be carefull about how you drive LEDs. The good news is it's one of those ultra low dropout models so anything over 5.35v should be OK, the way I read the datasheet, anyway. (Haha... but yet again, pay attention to that "Newbie" tag..)

Good job on the prop by the way.... I'm a bit of a prop junkie myself. (In the midst of 'Duinoing a light saber myself) I can't get over all the possibilities of using the Arduino for props.. it can do so much!

Thanks, I'll take a look at that datasheet.

If you like props, this is my youtube channel, where you can see videos of these meters in action:

Be warned, I kinda ramble on a bit. :wink:

[edit]

Oh, and this video from my favorites shows you what I'm attempting to replicate with this circuit:

So, two illuminated switches on the front panel, a red one which turns it on, and a green one which will change the display from bar to dot mode, a piezo speaker, four 20 led bargraphs, and two green leds on the tips of the antennas which I may do some PWM on. And a knob on the right side adjusts the speed via a pot.

Crap.

I've just noticed on wikipedia's page about batteries some important info I'd missed:

That the mAh they list for each battery, it says above the chart, basically tells you how long you have till the battery's voltage has dropped by 50%.

So, if I start out with 4 AA batteries, which will be around 1.6 volts when fully charged, I will end up with only 3.2v by the time my nearly six hours of drawing 500mA is up.

That means that I won't get 5-6 hours out of them, because at some point, the voltage will drop below 5v. And by my calculations, that will be when the batteries have discharged only 44% of the way to that halfway point. In other words, instead of nearly 6 hours, I'd be lucky to get 2.5.

And though I can't understand the datasheet for the 5V pro mini's linear regulator (kb50):
http://www.micrel.com/_PDF/mic5205.pdf

This wikipedia page:

...states that a linear regulator can't provide more voltage out than it gets in, and that the amount in must be the amount out plus the regulators "drop out voltage".

Which, for the pro-mini's regulator, looks like it is, at most, 0.35v; though there may be a diode or something in the circuit dropping it more, because I heard you needed a minimum of 7.5v to power it.

So where does that leave me?

Well, I now know I definitely need to feed the voltage regulator on the raw pin more than 5v to power the pro mini. Though I still don't know exactly what the minimum I need to feed it is. Without that minimum, I can't cacluate how long my batteries will last.

I also now know that if I try to feed 6v directly to Vcc through a Zener diode to drop it to 5.1v, that I'll only get around 40% og the mAh which the wikipedia page indicates the battery can give, because after that point the batteries will no longer be able to supply the minimum of 5.1 volts.

Lastly, I know that 9v is an ideal voltage to supply the 5v pro mini with if I want to get the most out of those mAh, because I'll be able to use almost all of a 9v's sources mAh before it drops below 5v.

But I also know that 6v is fine for a 3v pro mini for the same reason, though whether 9v is a whole lot better is a question still to be answered.

So... hm.

Based on this, I guess I should forget the whole zener diode thing, and forget running the 5v pro mini off 4 AA cells.

So with that out of the way, my options are:

  • Use 1 9v battery for a 1.25 hour run time.
  • Use 2 9v batteries for a 2.5 hour run time.
  • Use 6 AAA batteries for a 2.5 hour run time.
  • Use 6 AA batteries for a 5 hour run time.
  • Switch to the 3v pro mini, which will give me the same run times listed above, but allow me to run off 4 AA or AAA cells instead of 6.

Oh, and I've also determined that while 9v seem expensive, when you consider the number of AAA's you need to get 9v, the costs are pretty much the same, so there would be no real benefit to using AAA's, and in fact, they would actually take up more room.

So really my choices are:

  • Use 2 9v batteries for a 2.5 hour run time.
  • Use 6 AA batteries for a 5 hour run time.
  • Switch to the 3v pro mini, and run off 4 AA cells for a 5 hour run time.

I guess that simplifies things.

I would go with the 4 AA's, probably. It has the safest "overhead" voltage, cheaper for users to fill up and will run for 5 hours. I doubt you will need the processing power of the 5V version anyway so it's not like it will work any better.

It will take a real world test to figure out how long it will really work. The 5v/3.3v ratings probably have little bearing on real world "how long it will work" calculations. The 5v version will probably keep chugging along at 3.3v, and the 3.3v volt version may keep going as low 1.5v.... I think the spec sheet says the chip is rated at a minimum voltage of 1.8v and you can generally fudge any spec by about 20%.

I know I sat there and tried to calc out how long my lightsaber would work (10W Green LED) and got way more time than expected. All kinds of strange factors involved... LOL.

I finally gave up, marked it as voodoo and wired it up and timed it.

Here's what I'm trying to do with the Arduino on my project... full RGB blending lightsaber. Any color I want. :slight_smile:

Cool lightsaber. :slight_smile: Are you gonna be able to fit all the sound generating stuff and power for the saber inside it?

As for the 3v Pro Mini, few issues with that:

  1. I don't know if I need the processing power or not. I have no idea how to calculate that.

  2. Delay statements take twice as long cause the dev environment isn't designed for 8mhz. Dunno if the milliseconds() functon has a similar issue.

  3. Have to spend $14 on a new basic breakout board and then swap between it and the 5v one as I program the various arduinos. And they look the same, so I might accidentally plug in the 5v one and fry the board.

  4. Both 3v pro minis and 3v basic breakout boards are currenty out of stock at Sparkfun. (But I'm sure they'll get more in stock in a few days.)

  5. With a 3v board, I can't run two leds in series off a single pin... I guess I could run them in parallel though.

Nothing deal-breaking mind you, but a whole lot of little issues to consider.