Boost converter using 555

[HooRide]

A friend asked me to build him a circuit and package it nicely in a project box. The project calls for 3 9V batteries, I have several cases that fit a single 9V. I'd like to get 27V @ 4mA out of a single 9V battery.

I've never worked with switching power supplies but this seems like a great place to start.

I whipped up this schematic with the help of another friend and a few circuits I found after a quick google search but my question to you all is, how do I improve efficiency? How do I choose duty cycle and frequency of the 555 and how do I choose the inductor value? Also, instead of the zener to clamp the voltage, would a resistor divider make more sense?

Thank you!

[Neilm]

Have a look on line at some dedicated boost converter chips. They will quite often walk you though working out the values to use.

If using a 555 timer, you will have to consider start up. If you don't consider that you could get some very high currents flowing.

[Paul Price]

I would encourage you to try using a 555, you might enjoy the circuit design experience.

I've built very reliable CMOS 555 based boost convertors using I7555C and the BJT NE555.
Requires two external small signal BJT transistors, use one to sense current limit and one to sense output regulation voltage.
My circuit designs will start up at battery voltages as low as 1.75V and deliver 28V and once started will function with battery voltage down to .9V approx with a low threshold N-Chan MOSFET. All together about approx. 14 parts, including 2 small electrolytic caps, 6 resistors, 1 47pf cap, one .1 uf cap, 1 Schottky diode, something between a 10-330 uH inductor.

I use the resulting 28-V power supply to light up a long multiple-string array of 148 or so ultra-efficient white X-Mas LED's on the back porch.  In the darker dark of night, viewing up at the 4th floor and with a slowly randomly varying voltage around close to the LED's almost not being lit voltage, the light show is somewhat like a glance at twinkling stars in the heavens above.

The same design works well with NE555 but will not start at voltages <2.7V which can be considered a feature: UVLO protection.

Why use a 555?? Might take a few more components, but it is so common and available and cheap.

[TerminalJack505]

Here are some notes that I made working from equations that I found in a book that I have.

Note that this really isn't a good circuit for battery operation due to the fact that the battery's voltage will vary over its lifetime.  Because of this, the duty cycle will need to vary to compensate.

You can find 555 boost circuits that use two 555s (or a 556) that will work much better.  One of the 555s is used to set the switching frequency and the other 555 modulates the duty cycle (using the 555's Control Voltage pin) based on feedback from the output voltage.

[HooRide]

Sorry for the delay guys, I was out of town. So I decided to branch off and try a 556 just for educational purposes; seems like a good challenge.

For the actual build, I did some research and found, like Wilfred suggested, the TL499, which seems like a good solution since I need something reliable and need to deliver the project sooner than later.

Like some of you have suggested, the continuously dropping battery voltage is a problem and I'm not entirely sure what to do about it.

I have the project on a breadboard with the TL499 and it works really well down to 4V with 9V nominal but below 4V, things get out of hand very quickly (huge spikes in current draw and a very hot inductor).

So far I was thinking of a comparator that triggers a latching relay; though, with that addition, the project has now doubled in complexity.

Any suggestions on a simple way to kill power due to low battery voltage (without drawing much extra power)?