Boost Converter/Voltage Multiplier

I've looked at the MAX756 but it's a bit pricey. I have a basic circuit for a boost converter and was just wondering if it would meet the requirements to power an atmega 328 and some basic LEDs.

Circuit from: http://www.ladyada.net/library/diyboostcalc.html

Thanks for the reply. I think the MC34063A is what I need.

I was planning on using 2 AA batteries (or 1 if possible) to boost to 3.3V or 5V. I'm not sure how to properly calculate total current requirement, but I'm using a standalone Atmel Atmega 328 chip (not the Duemilanove), and a shift register (74HC595) to drive 8 7-segment display (LTC5653G).

To accurately measure total current, do you merely hook up an ammeter while the circuit is running?

I just tested my circuit and it's using up 40mA @ 5V (LEDs at a fairly dim level).

Suppose I had a 5V battery rated at 2500mAh, does that mean my circuit would last ~62.5 hours? (assuming 5V is constant) Would I double the circuit life (to 125 hours) by adding two 5V, 2500mAh batteries in parallel to get 5000mAh?

Switch-mode power supplies (SMPS) are rather sophisticated circuits and no place for the beginner or amateur with no experience. I would NOT recommend something that crude to power a sophisticated load like an Arduino. First off, that circuit lacks any sort of feedback. It appears to operate open-loop ("full-blast")

Would using the previous circuit I showed, along with two AA batteries as the input voltage, meet my circuit's power requirements? Or would it be best to just go with the MAX756? (It seems like the MC34063A needs a minimum input voltage of 3V while the MAX756 only needs 0.7V)

What is your desired run-time before the batteries must be replaced? Minutes? Hours? How important is running on two AA cells?

My project is a simple LED wall clock. I know I can just design it to use a wall-wort, but if I could get it to run at least 30 days on batteries it would be nice.

@ 30 days: 720 hours total circuit current: 40mA

Required mAh = 40mA * 720h = 28,800mAh

Assuming 80% efficiency of boost regulator:

Actual Required mAh = 28,800mAh/0.80 = 36,000mAh

I don't think I have a choice but to use a wall-wort.

I wonder if electric cars force you to drive economically; otherwise, your car battery could drain within a few hours.

I've decided to power up the device using a wall-wart, however, I'm not sure if it's better to use a 3V 200mA wall-wart and use a boost converter to get it to 5V, or use a 9V 200mA wall-wart and use a 5V voltage regulator.

I would think using the 3V 200mA and a boost regulator would be more efficient, however, the Duemilanove is designed with voltage regulators instead. Why is this?

the Duemilanove is designed with voltage regulators instead. Why is this?

Cheaper and less complicated. Always a good combination.

I'm not sure if it's better to use a 3V 200mA wall-wart and use a boost converter to get it to 5V, or use a 9V 200mA wall-wart and use a 5V voltage regulator.

What's more important to you, efficiency or simplicity of implementation? If you draw 40mA then a 9V-to-5V regulator will constantly dissipate (waste) 0.04*4=0.16W of power. An 80%-efficient switching regulator delivering 5V*40mA=0.2W of power will require 0.2W/0.8=0.25W and will waste 0.05W of power.

So you save 0.11W of power dissipation (or about 3 LED's turned on) at the expense of costlier components and more complicated implementation.

Of course, a lot depends on the efficiency of the wall warts -- which are generally not very efficient (does yours warm up even when it's not connected to anything?)

-- Check out our new shield: http://www.ruggedcircuits.com/html/gadget_shield.html

Why would you want to fool around with making a boost converter when you can just use a 9V wall-wart and rely on the Duemilanove's built-in regulator? It's no contest IMHO. Never use a boost converter if you can avoid it. Extra work, extra complexity, only worth doing if you have no other options.

Sorry, I forgot to mention that I'm working on etching my own pcb of this project, so I won't be using the actual Duemilanove to drive the LEDs. (I just realized that I should probably be posting in Sparkfun's forums as the Arduino forums pertain mostly to its boards.)

What's more important to you, efficiency or simplicity of implementation?

I would think, from a circuit designer's standpoint: simplicity of implementation (time saved); and from a consumer's: efficiency (money saved).

As both a consumer and aspiring circuit designer, and your given information, a design with a switching regulator looks like the best bet. In the long run, the power saved using a switching regulator will overcome its initial cost.

Just for calculations sake: @ $0.12/kWH

Voltage Regulator (9V to 5V):

Total Power Consumption = 9V * 40mA = 360mW Power Dissipated = (9V - 5V) * 40mA = 160mW

Cost Calculation:

360mW * 24H/Day = 8,640 mWH/Day (or 0.00864 kWH/Day)

0.00864kWH/Day * $0.12/kWH = $0.0010368/Day (or ~$0.0311/Month, $0.378/Year)

So running this device for 1 year would cost: $0.378

Switching Regulator (@80% efficiency):

Circuit Power Consumption = 5V * 40mA = 200mW Total Power Consumption (@80% eff.) = 200mW / 0.80 = 250mW Powe Dissipated = 250mW = 200mW = 50mW

Cost Calculation:

250mW * 24H/Day = 6,000 mWH/Day (or 0.006 kWH/Day)

0.006kWH/Day * $0.12/kWH = $0.00072/Day (or ~$0.0216/Month, $0.2628/year)

So running this device for 1 year would cost: $0.2628

Difference:

You would save $0.115/year ($0.378/year - $0.2628/year) using a switching regulator over a linear regulator.

If the switching regulator costs $5 more:

Time to overcome cost of switching regulator = $5 / ($0.115/year) = 43.5 years