The easiest way to get 5V-ish with batteries?

[Pax Writer]

Hey people

I'm making some small battery-powered circuits using AVRs and other kinds of logic, and I was wondering: How would you get 5V from standard batteries?
I've been thinking about 3xLR44 to achieve 4.5V, but I think that may just be a little low.
Also I thought about using a simple 7805 with some combination of batteries, like 3xCR3032, but I can't help thinking that there must be a better (and more clever) way.
The circuit in question consists of a single processor and three 7-segments, so the power consumption is limited. No more than ~200mA peak.

Feel free to educate me
Thanks in advance.

 

[picbits]

I've come across a similar problem and I used 3 x SR41s and a LF version of the PIC microcontroller - these are fairly happy running at around 2.5v upwards.

Use the SR44 rather than the LR as they have a lower voltage drop vs drain.

I managed to get a 10F200 pulsing a superbright LED with various blink patterns for a customer and it lasted over 24 hours on a set of 3 SR41 cells.

 

[picbits]

Oh and stick a whacking great capacitor for your 200ma peak surge across the power supply. 200ma out of button cells is stretching it somewhat.

 

[Pax Writer]

May I ask at which duty cycle and frequency you pulsed the LED? And with which current.
Pulsing LEDs is something I haven't yet found very much material about.

 

[picbits]

I used a 1/8 pulse for the continous mode (i.e. LED lit) at around 2khz frequency.

Other modes were a 0.1 second on and 0.9 second off for standard flashing. Curent was just its rated 25ma

With a superbright LED, the 1/8th PWM made the LED around half the intensity (visibly) although this wasn't measured scientifically.

 

[Pax Writer]

Thanks, picbits
More battery vs. 5V tips are welcome.

 

[picbits]

The best way of doing it is to use a switch mode power supply. These can operate with an input voltage down to a volt or so and can make much more use of the battery capacity.

Downside is that they tend to be a bit bulky and are certainly not something for beginners.

Other options are low dropout linear regulators.

 

[Hero999]

I used a 1/8 pulse for the continous mode (i.e. LED lit) at around 2khz frequency.

Other modes were a 0.1 second on and 0.9 second off for standard flashing. Curent was just its rated 25ma

With a superbright LED, the 1/8th PWM made the LED around half the intensity (visibly) although this wasn't measured scientifically.

Pulsing most types of LEDs doesn't improve their efficiency. The eye acts as an integrator so a light thats pulsing at 2kHz 50% duty cycle but twice as bright will appear the same as one continiously illuminated.

Pulsing most kinds of LEDs is counter productive particularly the white, greeen, violet, pink and blue ones which are actually less efficient at higher currents.

However I've heard that some of the superbright red, orange in infra red benifit from high peak current low duty cycle.

 

[audioguru]

Your vision's response to brightmess is logarithmic So you can see in nearly darkness and in sunlight, a huge range.

Half power in an LED is just a little dimmer. One-tenth the power looks like half the intensity.

 

[Hero999]

Your vision's response to brightmess is logarithmic So you can see in nearly darkness and in sunlight, a huge range.

Half power in an LED is just a little dimmer. One-tenth the power looks like half the intensity.

That's true but somehow I don't think nature selected log to the base 10, more like log to the base 2 or natural log.

 

[audioguru]

In addition to the logarithmic response to brightness, your eyes have automatic gain control with the iris reducing very bright continuous light. I think the iris has a 30ms response time.

 

[Space Varmint]

Hey people

I'm making some small battery-powered circuits using AVRs and other kinds of logic, and I was wondering: How would you get 5V from standard batteries?
I've been thinking about 3xLR44 to achieve 4.5V, but I think that may just be a little low.
Also I thought about using a simple 7805 with some combination of batteries, like 3xCR3032, but I can't help thinking that there must be a better (and more clever) way.
The circuit in question consists of a single processor and three 7-segments, so the power consumption is limited. No more than ~200mA peak.

Feel free to educate me
Thanks in advance.

You can use diodes to drop the voltage. I get best results by attaching them to the (-) negative side of the battery. So if you use 4 1.5 volt cells you can attach the anode to the negative of the batteries and the cathode to the ground return of your circuit. This will give you a supply of 5.4 volts due to the .6 volt drop across the diode.

 

[Hero999]

Under most circumstances it doesn't matter but by convention, the diode is normally placed inthe positive side of the incomming supply.

Fresh AA cells can be up to 1.6V each giving a total of 6.4V so the MCU needs to be rated for at least 5.8V for this idea to work. Fortunately most MCUs have an absolute maximum voltage rating of 6.5V so this won't be a problem, even withouth the diode, although it's a good idea to use it anyway.

 

[Space Varmint]

Under most circumstances it doesn't matter but by convention, the diode is normally placed inthe positive side of the incomming supply.

Fresh AA cells can be up to 1.6V each giving a total of 6.4V so the MCU needs to be rated for at least 5.8V for this idea to work. Fortunately most MCUs have an absolute maximum voltage rating of 6.5V so this won't be a problem, even withouth the diode, although it's a good idea to use it anyway.

Sure you can put it on the positive but I do get better results on the negative for some odd reason. Some times I put them on the ground lead of a V regulator to drop the rated voltage by .6V.

 

[picbits]

I thought sticking them in the ground lead of a regulator raised the output voltage by 0.6 volts

 

[3v0]

Sure you can put it on the positive but I do get better results on the negative for some odd reason. Some times I put them on the ground lead of a V regulator to drop the rated voltage by .6V.

What do you mean by better results?

 

[Hero999]

It does and it's a pretty poor way of doing it as it completely ruins the regulation and temperature stability. You're much better off using an LM317.

As far as connecting it in the postive or negative line to the whole circuit is concerned; it doesn't matter because the voltage across the circuit will be the same.

 

[Pax Writer]

Hi guys

Thanks for all your good answers. I thought this thread ended with my last reply, so I wasn't aware of the good posts.

I ended up solving the issue by using a low-drop regulator which had a 0.6V drop and then power the circuit with either 2 LR44s or 2 CR123As. That single IC was smaller and just as cheap as all the other usable solutions, I could find.
However, your debate about the pulsed LEDs helped make this thread all worthwhile for me. Thanks

 

[3v0]

Perhaps you misunderstand.

A poster may or may not get an answer to their question. It is good when it happens but in truth it is but a side effect. The original post is in main a topic for discussion or debate.

If you understand that, it is clear that if or when the question is answered has little impact on the lenght of the thread.

Another interesting item is that threads with the words: simple, easy, or basic will generaly be longer then ones without.

Hi guys

Thanks for all your good answers. I thought this thread ended with my last reply, so I wasn't aware of the good posts.