Conserving Batteries - Voltage Reg. vs. step up

Forum 2005-2010 (read only) Hardware Interfacing
1

Background:
I am building a wireless garage door monitoring system. Although simple in concept, I would like the end result to be elegant / well thought out and professional. This project will consist of a transmitter and receiver. The tx unit will be mounted inside my garage, where it will sense the doors position and tx this. The rx unit will pick up on this, and simply display the status of the door.

Implementation:
The tx/rx units will consist of atmega 328's soldered on perfboards. The boards will simply contain a power supply circuit, the u-controller circuit and connectors for I/O. In the case of the tx unit, it will have connectors for the door sensors, and the rx unit will have connectors for the output circuit (which will simply be a few led's).

For wireless transmission, I plan on using RF communications at 315 Mhz - 2400 bps.

The rx unit will always be powered on and listening for the command byte.

The tx unit will be sleeping (lowest power mode), and will awaken upon interrupts (when the door moves). When interrupted, it will sense the doors position and tx this data to the rx. (I haven't figured out the comm protocol yet, but I'm thinking, command byte, data byte, and parity)

The rx unit will be powered by a wall wart or pre-existing 5/9v source

My Question:
The tx unit will be powered by batteries. I am unsure of whether to use a fixed voltage regulator (such as something in the 78xx family), an LM317 (which will both be driven by a 9v battery), or use a step up converter (which I hope to be able to power with 1 or 2 AA/AAA batteries)

Please advise of which power supply circuit should be used
-78xx
-LM317
-step up converter.

I would like to maximize my battery life, hence the sleeping scheme on the tx unit. Considering that the door might toggle state 4 times in a day, is it reasonable to expect battery life of atleast 6 months on the tx unit?

Additional Info
If you suggest a step up converter, I plan on building spark-fun's commercially available version. Here's a schematic:

Is the 22uH inductor a widely available component?

2

For the Tx unit I would consider skipping the voltage regulator alltogether and source power directly from 2xAA or 2xAAA cells. At 3V you could then use the ATMega internal osscillator at 8MHz (lower voltage and lower frequency will reduce power consumption and increase battery life).

If you need a higher voltage (more than 3V) for the 315MHz transmitter, you can generate this with an MCU driven step-up converter.

3

I question the integrity of the serial library with the internal oscillator at 8Mhz. Doesn't this library depend on the 16Mhz clock to accurately send/receive data?

4

You are better off with the boost converter and 1 or 2AA cells. If
you use a 9V battery you waste 4/9 of the energy.

If you can run your ATmega at 3.3V you could use a 3.7V Li-ion battery.
A 3.7V source and an 3.3V LDO regulator gives you almost 90%
efficiency.

(* jcl *)

www: http://www.wiblocks.com
twitter: http://twitter.com/wiblocks
blog: http://luciani.org

5

If you can run your ATmega at 3.3V you could use a 3.7V Li-ion battery.
A 3.7V source and an 3.3V LDO regulator gives you almost 90%
efficiency.

That would make a nice efficient voltage source. However one important thing when using Li batteries is not to allow the cell to over discharge, that will destroy it (I know firsthand :-[ ).

Some kind of automatic voltage disconnect circuitry would be required to protect the battery when it's voltage drops.

Lefty

6

I question the integrity of the serial library with the internal oscillator at 8Mhz. Doesn't this library depend on the 16Mhz clock to accurately send/receive data?

The serial library works well both for 8MHz and 16MHz. Baudrate is calculated at compile time (using the F_CPU constant) based on the board you choose.

7

I am having trouble sourcing the IC for the boost converter.

I will be ordering parts from this supplier: Electronic components - dipmicro electronics

In addition, even if i were to find the ncp1400 IC, my supplier does not carry a 22uH inductor.

I think the easier route will simply be a voltage regulator. Any suggestions for conserving batteries?

8

Using http://www.findchips.com shows the device in stock at a number of
places. A 22uH inductor is also quite common.

IIRC the NCP1400 is only available in SMD. Is that OK?

For currents of 50-100mA you could also use a switched capacitor
converter like the ADM660 (which is available in TH). There
is boost converter schematic, using the ADM660, in my LCD interface
datasheet at http://tinyurl.com/9mv2cs I use the ADM660 to generate
a 5V supply for the LCD.

(* jcl *)

www: http://www.wiblocks.com
twitter: http://twitter.com/wiblocks
blog: http://luciani.org

9

Some kind of automatic voltage disconnect circuitry would be required to protect the battery when it's voltage drops

Will the built-in Brown Out Detector work?

10

Nice simple solution that doesn't break the bank or cook sensitive cells and is efficient .

Hook 3 alkaline cells in series and wire them directly to the 5V pin. The voltage will vary a bit as the cells discharge, but if its just running simple sensors (like reed switches) it will be quite happy down to at least 3volts. No inefficiencies with voltage regulators or step up voltage convertors and no tetchy high tech batteries to nurse.......

11

No inefficiencies with voltage regulators or step up voltage convertors and no tetchy high tech batteries to nurse.......

I agree, this is probably the simplest and cheapest way to get the project off the ground and proven out. Later if deemed needed a different battery system can be played with

Lefty

12

You could always look into Switching Voltage Regulators, much more efficient than the regular Linear regulators.

They're quite a bit more than the linear regulators as well.. a 5v switching regulator is about $10 USD, while the regular, linear 5v regulator (for example, 7805) is what, 5 for $2 depending on where you buy them.

I haven't used one personally, but I hear they do wonders for battery powered supplies.. I'm just starting to play with some old 12v batteries from a drill, they're only at like 11.48 volts fully charged.

Doesn't seem to discharge too fast, but I don't like the fact that the regulator gets as warm as it does. Not exactly "hot", but warm enough to say there's too much energy being wasted. :confused:

Here's an example of how to make your own DC-DC converter, with a tool to help you calculate everything.
http://letsmakerobots.com/node/3906
I think this will be one of my next projects. (god I have alot of those):smiley:

13

Thanks for all the replies.

I am going to simply connect 3 batteries in series and be done with it. Simple and efficient.

What size capacitor would you recommend in order to smooth the voltage/current? I presume that I will simply place this in parallel with the battery pack.