5 and 3 volt from 12

Hi All just a generic question… I did a lot of googling and I have to admit I am a beginner amateur when it comes to actual electrical engineering… What I’m trying to do is have the standard 3 and 5 volt from a car /motorbike battery. Now I read a lot of bad things about peaks etc. so I would like to show sketch in Fritzing to see if I can get any feedback concerning this there are 2 circuits and I was wondering what the best approach would be and why. bare in mind I’m learning this stuff so don’t be too hard on me :confused:

How much current (Amps or milliamps) do you need?

Your general approach(s) are good. I didn’t totally analyze your circuits. A voltage regulator is the normal way to step-down voltage.

But, you’ll need to get rid of the series resistors at the 12V input. The diode on the input is only good for reverse-voltage, and if that’s going to work you need a fuse. Without a fuse the diode will get fried with reverse voltage.

Linear voltage regulators (like the ones in your schematic) are only good for low current less than 1A. And, it you approach 1A you’ll need a heatsink. (The current ratings on the parts don’t tell the whole story.) A switching (AKA switchmode) voltage regulator is more efficient and they can (generally) handle higher current. But, switching regulator circuits are not as easy to build so if you go that way it might be best to buy a pre-built module.

With the voltage regulators in series, the 1st one has to handle the current of both circuits, so it will get hotter, but the 2nd regulator doesn’t have to “drop” as much voltage so it doesn’t get as hot. So, the preferred method depends on how the current & power (power is related to heat) is distributed,

Suggest you purchase a ‘Buck converter’ from eBay or Pololu



Cheers doug I will keep all of that in mind. The whole design is basically to power a ATMEGA328 and have modules or parts working on 3 volts as well. so small example have an ATMEGA328 with a relay and a 433Mhz button to switch the relay on/off (Simple example but you get the drift. This is also the reason I want to do this myself. You can get breakout modules for everything but A. it takes the fun out of making it yourself and learning more in the process. And b. you end up with a bucket of modules tied together for something quite simple. I Don't use arduino's for the final project or when I implement the project. I simply use the arduino 328 which costs me £1 and make a simple PCB Advantages are that it costs a fraction of what an arduino would cost every time, Your design is a lot smaller and I made a dedicated board with a USB serial module and a ATMEGA328 socket to load the arduino sketches onto the chip. Works like a charm. I reckon thats where I differ from a lot of people I see on several boards they want it to work ASAP and if it does the job is does the job. I want to refine and make it better first to get the optimal as the I consider it as a project on its own to make a good workable PSU for that with all the safety features. This will be used for many projects to come.

Hello, It looks pretty good. + what DVDdoug said. However the 3V output gnd connection is wrong. It should not be connected between the resistors.

Good luck!

R1 @ 1k is a problem.

The input diode is a problem, add an input fuse :wink: .

Dropping 12V to 3.3V by linear regulator is a huge step: your current limit without heat sink is about 20 mA (175 mW dissipation), with big heat sink about 200 mA (1.7 W dissipation!). The 5V regulator has the same problem.

I've done quite some 12V/5V/3.3V projects, what has worked great for me so far is a 2-3A 12V-5V buck converter, and a linear regulator for 5V to 3.3V. Typical current draw of those projects is no more than 200-300 mA at 5V, 50-100 mA at 3.3V. Works like a charm, no power supply issues provided there's enough 12V on the input (for projects where I suffered voltage drops I added a diode in the input and a 2200 µF capacitor to bridge that dip).

Attempts of powering a small 3.3V project off 12V by linear regulator has so far only resulted in failure... due to overheating regulators.