Small 9V -> 3.3V switching regulator, any recommendations

Go To Last Post
27 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm looking to make some "spooky" LED attachments to the kids' Halloween costumes, and I'm wondering about good choices for a small switching regulator to supply a few 10s of mA at around 3.3V from a 9V battery. Any recommendations?

Yeah, I could use a pair of AAAs, but I wouldn't mind playing around with some small high-efficiency regulators just for the heck of it.

Mike

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Linear Tech has some nice, low current switchers. Not too much money, either. None, particularly, to recommend.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

MC34063, easy to get and relatively efficient for step down from 9V.

oddbudman

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

oddbudman wrote:
MC34063, easy to get and relatively efficient for step down from 9V.

Thanks, that one has my mouth watering. Once a geek always a geek. :)

Mike

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Been digging a bit on the MC34063, since I also have a need of a switching psu that is "reusable" in battery powered designs. It looks great!

Anyway, found this, wich makes it even better : http://www.nomad.ee/micros/mc340...

I ordered some of these parts, they look really nice thanks for the tip @ oddbudman

Author of simavr - Follow me on twitter : @buserror

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You will find that the MC34063 can be bitchy at times. It is a rather old design, not doing PWM.

Stealing Proteus doesn't make you an engineer.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ArnoldB wrote:
You will find that the MC34063 can be bitchy at times. It is a rather old design, not doing PWM.

What do you mean, bitchy? And what is the downside to it not doing PWM?

Do you have another suggestion that's low cost and easy to use?

Mike

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You could perhaps put the right number of LEDs in
series so that the voltage is around 8 Volts
and you get away with a small current limiting
resistor.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ossi wrote:
You could perhaps put the right number of LEDs in
series so that the voltage is around 8 Volts
and you get away with a small current limiting
resistor.

Problem is I want to use an AVR to make the LEDs be interesting, and an AVR can't run at 9V nor can it switch 9V signals.

Mike

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A small 5V linear regulator like a 78L05 comes in TO92. You could run the AVR from this, and use external transistors to switch the LEDs. That will be a lot less hassle then messing around with switching regulators.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

jayjay1974 wrote:
A small 5V linear regulator like a 78L05 comes in TO92. You could run the AVR from this, and use external transistors to switch the LEDs. That will be a lot less hassle then messing around with switching regulators.

But I want to mess around with switching regulators. That is to say, I want to enter the 21st century as far as supplying power to my designs, and this little project seemed as good an excuse as any. I'll still keep my bag full of 78L05s, I'll just be a little more discerning when to use them.

Mike

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Buck converters are easy to make: coil, diode, FET, transistor driver, PWM from ATTiny, current shunt with voltage feedback. Atmel's BC100 reference schematic has several to look at. At less than an amp practically anything would work. But you'll have to figure out how to power the Tiny from 9 volts.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

kk6gm wrote:
What do you mean, bitchy? And what is the downside to it not doing PWM?
The regulation is such that a too low output voltage on the ICs feedback pin can turn on the output stage, but not off. The resulting pulse can therefore be to large, you get an overshot at the output. Unless the current limiting kicks in and shortens the pulse. Which can in turn result in a delay of the next pulse.

Since the oscillator and the turning on of the output stage are not synchronized you rarely get a stable duty cycle. The whole thing "wobbles" in continuous mode.

kk6gm wrote:
Do you have another suggestion that's low cost
You get what you pay for.
kk6gm wrote:
and easy to use?
Linear and National have a wide variety of switching ICs. The MC34063 is 20th century, not 21st century. It is probably 20 years old.

Stealing Proteus doesn't make you an engineer.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm with kk6gm; I also want to use switching power supplies, want to make projects that run from months on batteries etc etc and learn by examples; however all the switching PSU I hear from go from :
+ "it's dead easy, use your swiss army knife, a piece of string and a LED -- done thar for years" to
+ "use a $15 Linear blach blah that you can never find 'the' inductor it really needs to go from 3% efficiency to the graaal 99.7%" or even :
+ "here, here's my design that converts from -15.8V to +360 triphase from just one nuclear cell!"

Result is that no "hobby" project uses switchers -- I haven't seen a single one to be lifted from a project, and to learn from...

I don't mind a wobblier-than-straight-catholic power rail, AVR are sturdy as hell anyway :D

I've ben using a now discontinuted LM2798MM for extra low power need; it's a capacitor based switcher (look mom, no inductors!) that has been fantastic for the 2.0V projects - just 4 capacitors.. however it's discontinued, and I can only source the 2.0V anyway, no 3.3 or any other voltage. If anyone know of a substitute that is as easy to use...

Author of simavr - Follow me on twitter : @buserror

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

why convert 9V to 3V if you can convert 1.2V to 3V instead?

look http://www.austriamicrosystems.com/
I've used them and they really are simple. whole circuit took about 10x10mm space on the PCB. Inductors are readily available and cost around 1$ each.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Also the 7805 will not have the high efficiency that you want.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you use a 78L05 just for the AVR alone, I think a switcher will not add that much battery life as most of the battery capacity will be consumed by the LEDs anyway.

Depending on what you want to do with the LEDs you don't have to run the AVR at full speed either to reduce power.

I made a collision light thingy for a Lego aircraft that runs on a single AA cell and has three LEDs. Every ~15 minutes it wakes up and flashes the LEDs for one minute, each LED flashes once a second with 20ms pulses, with each LED having a slightly different repetition rate. When asleep it blinks one LED every minute. It runs for about 9 months on a single AA battery and it uses a micropower SP6641A-5 boost converter. It's a tiny chip, with the inductor, diode and caps it's about the same size as a TO-220 body :)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Do you need the switcher to be available in a DIP package? Linear makes a ton of very easy to use switchers. The MC34063 appears to be a very slow part - 100KHz - meaning that you will have to use fairly large inductors and capacitors. At your load currents they won't be massive, but still...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm enjoying the discussion and, if I do not fool myself, learning a few things too. I looked at some Linear parts and had some sticker shock. $6-$8 single piece prices! I suppose I might have just been looking at the wrong parts, but...

So now I'm looking at what National has to offer.

Mike

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Are you not giving the MC34063 a go?

Why get bogged down chasing dragons and hard to get parts, when all your wanting to do is expand your horizons with a DC:DC converter?

The MC34063 is still used extensively to this day. Pull apart any car phone charger and you are likely to find the 34063 inside. To some extent its like the 555 of DC:DC converters- well worth learning.

I'm not saying there are not _better_ switches out there. But this one may be adequate for your needs.

oddbudman

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

oddbudman wrote:
Are you not giving the MC34063 a go?

Why get bogged down chasing dragons and hard to get parts, when all your wanting to do is expand your horizons with a DC:DC converter?

The MC34063 is still used extensively to this day. Pull apart any car phone charger and you are likely to find the 34063 inside. To some extent its like the 555 of DC:DC converters- well worth learning.

I'm not saying there are not _better_ switches out there. But this one may be adequate for your needs.

oddbudman


Oh yes, I'm still going to try the MC34063 - got some on order for 48 cents each. Price alone makes it attractive, and the fact that it has many sources tells me it is still a viable part. I'm just looking at what else is out there as well. And so far I'm not seeing anything in the same ballpark.

Mike

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Check out sparkfun.com power regulation tutorials, even with usb even with the LM317 where you can adjust 9V to 5V or 3.3V.

I have a bunch of MC34063 ICs. PM me if you want some. IMO, don't use it, unless you want to pump up the voltage.

But the LM317 voltage regulator, which I have some, too, would solve your problem.

Btw, go USB! Trash the 9V crab! It's a waste money and power!

Google usbtiny and USBASP and see the schematics for how make have USB connector as power provider. Sparkfun.com also has the tutorials!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

By the way, an app note for the MC34063. http://www.intusoft.com/onsemipd...

And see thread https://www.avrfreaks.net/index.p... that I posted.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

another option (which I have gotten a few of for learning and testing) is an LM2576 http://www.national.com/mpf/LM/L...

It is only a step down converter and does require an inductor. They come in adjustable versions and fixed voltages variants.

Edit: just went to nationals web bench and found the LM2576 is no longer supported for new designs, may want to look at something like the lm2676. Evidently "Much Improved Efficiency And Switching Frequency" than the LM2576......

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Edit: just went to nationals web bench and found the LM2576 is no longer supported for new designs,

Good thing you checked. In general, I check for the obsolescence of a chip by going to mouser.com. If they don't sell it or mark it as obsolete, don't bother to use the chip, should you plan to mass produce your product.

The MC34063 is still carried by mouser.com. :-)

In the post https://www.avrfreaks.net/index.p... I also found the Maxim voltage stepper ICs. They are pretty good, even take in 0.8V of input and boost up to 5V. But, the Maxim chips are expensive. The MC34063 has a 3V min input. I guess that's why they so cheap.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you just need a single switching regulator and you don't want to build one yourself, just buy one of these. I have one and they are great.

If you want to design one yourself, it's really not that hard. Linear charges the highest prices for their parts, and if you can afford them, they're awesome. Linear's datasheets are the best, they have Switcher CAD, etc, making the design process easy. However if you need cheaper parts, National and TI make some good stuff.

Selecting the inductor, output cap, and diode is pretty easy. Just look at the example circuit in the datasheet, go to your preferred distributor and find an equivalent part. Sumida is a good manufacturer of inductors and they have a wide range of parts. If they specify a certain kind of output cap (like XR7 ceramic) then stick with that kind of capacitor. Pretty simple really.

Designing the PCB for a switcher is not that hard either. Just do two main things: keep the switch node as small as possible, and keep the negative terminals of the diode and output cap as close together as possible.

Good luck whatever you decide.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Most datasheets of switchers have guidelines on how to route the PCB, it slightly differs for step up and step down regulators. Unfortunately LT usually has example layouts using the smallest possible components, and strictly adhering to the example using bigger components is less easy. It's about containing the fast changing currents in a small loop area.

Farnell nowadays carries a very nice broad selection of power inductors.