AVR boost converter


Over at SpriteMods, [sprite_tm] realized that a microcontroller could be used as a boost converter to power itself. A boost converter steps up voltage from a battery by switching the output of a coil. First, it is tied to ground so a magnetic field can build up in the coil. It is then released as a higher voltage than the input. Normally dedicated chips do this at an incredibly high frequency, but the PWM signal from an AVR works well enough. This can be used in low-power situations where space is an issue.

[via EMSL]

19 thoughts on “AVR boost converter

  1. Have been using simple boost converter made with 555 and comparator to for example provide high enough voltage for relays and other low power stuff. It makes programming easier to have switching with external controlling circuit, but it’s more user friendly to handle it in the software: easier to adjust etc.

    I just got few good sized inductors from some disassembled pcbs and was going to make either: 10v-14v -> 15v boost converter to make car charger for laptop or to make 50v-> 0-50v buck converter. This time i’ll handle controlling with microcontroller so i can try my new programmer/debugger at the same time

  2. I recall this being discussed on an avr-related mailing list (maybe avr-chat?) in the mid 2000s, including nicely-formatted ASCII circuit art. Unfortunately, I am not able to find the message online or in my personal archives. My recollection is that this concept was proposed (including the bootstrap switch) but never tested.

  3. This is great, I’ve thought about this, but I could never figure out how to easily get it jump started. He did is so simply and elegantly, great job! My original idea was have like a super cap, but this is wayyyy better.

  4. There should be a capacitor in series to the switch (with discharge resistor in parallel) so that you can hold the switch for a longer time without running too much current through the coil. Or wouldn’t that work? I am not so sure…

  5. @Lupin, I dont think adding a cap in series will work. The point of pulling the inductor to ground is to allow current to flow through the inductor. Then, when it switches off, the inductor will continue to drive that current to the cap. With a cap in series, the current in the inductor will go to zero once the cap is full. after that, you will need a way to discharge that cap inorder to have current flow through it.

  6. Gosh darnit i’ve done this before with a zilog 8 pin proc outputting 5 khz on a mosfet…I didn’t deem it publish worthy i was able to power a blue led off of a AAA battery

  7. He should replace the diode with a p-type MOSFET and he’ll get much better efficiency. There’s a good bit of power being burned up on the diode. Would probably send him to the mid to high 80’s efficiency wise.

  8. A schottky diode would be a quick way to increase effiency. Yes, a pmos could be used (synchronous rectifier), but you need to add some logic / use extra pins to assure there isn’t cross conduction (both fets on at same time, shorting vout to ground).

    also, you’d do better with a more capable gate driving scheme for the fet than the (probably weak) pwm signal from the ic.

    probably overkill, but that’d likely get the highest efficiency / longest battery life.

  9. yeah, was chatting to Mr Sprite on IRC just now.

    BTW if you need small FETs, take apart old electret mikes as they have one. Its an N channel JFET but for low power applications this doesen’t matter.

    Another source for MOSFETs is the buck circuits from dead PC motheboards- a minute with a heatgun will remove them with no damage if they are surface mount, otherwise use chipquik.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.