Raising Ni-MH Voltage

[JamesBourne]

I'm currently building a project that generates a frequency using a 555 timer.  Since the power requirements are low, I thought of using a small, rechargeable battery.  After dismantling one of those "battery-less" shake flashlights, I found a Ni-MH 40mAh 3.6V battery inside.

My problem comes with the voltage.  The 555 requires at least 5V to operate and the battery usually gives 3.1V after a quick shake and 2V after leaving the light on for a few hours.  I have read that you can use a charge pump to raise the voltage, but the IC used to do this also requires 5V!

So, if possible, how can you raise the voltage about 3 times without using IC's?  Schematics will be much appreciated.

Thanks

[Hero999]

You're not doing very well if your 555 only works down to 5V. The datasheet says it should work down to 4.5V, perhaps you've not measured the voltage properly or have a fake, although I don't know why anyone would bother faking such a crappy IC.

Discrete design isn't only for very low voltages, there are plenty of ICs which will work down to 2V and some which will work below 1V.

The old TTL NE555 is an obsolete component which has very few uses in real circuits and is probably produced to supply educators any hobbyists rather than real electronic manufactures.

There are better versions of the 555 which work down to lower voltages than 4.5V but they're more expensive:

• CMOS 555s will down to 2V: ICM7555, TS555, ILC555, TLC555 etc.
• There's a version which will work down to 1.5V: LMC555
• The award for the lowest voltage 555 goes to the ZSCT1555 which will work all the way down to 0.9V.
  

But why bother? The only real need for any of the above ICs is if you've already designed your circuit for the 555. You can get cheaper ICs which can be used to do similar things to a 555 and are often cheaper:

• Old 4000 CMOS logic, i.e. CD4001, CD40011 etc. which works down to 3V and is cheap.
• Slightly more modern CMOS 74HC logic, i.e. 74HC00, 74HC14 etc.  which works down to 2V and is cheap.
• Modern low voltage CMOS logic i.e 74LV, 74AU, 7SV etc. which all works down to 1V (some down to 0.8V) but it's slightly more pricey, harder to get hold of and you may only be able to get SMT for some components.

What are you triyng to do?

If it's light a white or blue LED then you're using the circuit, you need a Joule thief.

It's possible to build a white/blue LED flasher with a 74HC IC, a Schottky diode, some resistors and capacitors that works between 2V and 3.3V and doesn't require an inductor or a transformer.

[JamesBourne]

Let me rather say what I'm trying to do:

I want to create a sine wave of 140Hz with the frequency as accurately as possible.  This will be played through on a 3.5mm earphone.  My current idea is to use a 555 to create 140Hz as accurately as possible and filter the square wave to sine.  I'll be happy to use another oscillator as long as the frequency can be accurately controlled.
I want to run the circuit off the small rechargeable battery so that it's small and portable.

If you have any other suggestions or ideas on achieving this, I will be happy to listen.

[Hero999]

This really belongs in the other thread, as it's related to it.
http://www.electronics-lab.com/forum/index.php?topic=20816.msg94020#msg94020

Creating circuit will will run from 2V, produce a reasonably clean sinewave with a frequency of 140Hz and 5% tolerance is not going to be hard, rather than smoothing a squarewave, it's probably better to use a phase shift or a Wein bridge oscillator.

What's the impedance of the earhone?

Your problem is going to be amplifying it to drive the earphone.

A square wave would be easier but you already said it has to be sinusoidal.