Sometimes buying a ready made power supply is just not good enough, especially if we want something special. I needed a super efficient 1 Amp power supply for my autonomous weather station that lives in the middle of a field on a small island in the middle of the Irish sea. There is no power option other than using solar panels, which really struggle in these northerly latitudes during the darker depths of winter. A standard linear regulator is not a good option as it burns off power to reduce voltage from 12V to 5V (but great for warming cold fingers).

Being of a basically lazy nature, I did try a few other options first, including a generic adjustable power supply from China. The really great thing about this device is that, other than actually arriving in the post, it made lots of very interesting buzzing, whining and whirring noises that made me feel like I was getting really good value for my money. But alas, it consumed a huge amount of power (50mA) and is now in the bin :(

I then went on a mission to find the ultimate component and, thanks to Texas Instruments, eventually stumbled across the LM2675N-ADJ/NOPB (Which will now be referred to as 'Bob', to avoid the spam filters). And even more praise for TI, the data sheet provided suggested circuits and components. What could go wrong?

Nothing went wrong, except that the other very similar TI components I tried were no good. Conclusion: The supply MUST be adjustable.

Step 1: Component Specifications

The main attraction of this component was it's quiescent current - the weather station operates on a tiny Arduino Nano most of the time which uses about 30mA with the sensors attached and powered up. It would be no good to me if the quiescent current was 50mA like with the dodgey gadget from China as that would be more than the device current. Looking at the datasheet, Bob has a quiescent current of about 5mA, which is totally acceptable.

Every 10 minutes or so, the Nano in the weather station switches on a whole array of other gadgets, which require at least 500mA, so a 1 amp ceiling was selected for effective operation.

The other thing to look at is the actual efficiency under load. According to TI, it should be 90% when stepping down from 12V to 5V.

So what's the downside? Only that the device can be noisy, in the digital sense. Some applications such as radios are probably not appropriate as there would be a lot of acoustic interference.

Step 2: How It Works

Bob creates a high frequency square wave pulse that creates an electromagnetic field in the inductor that effectively stores voltage that's then available to whatever load is applied. It is possible that the input to the inductor can go to values less than zero, so the diode is there to draw current from the ground rail.

Effectively, the circuit can operate like a kind of flywheel, depending on the operating frequency and the value of the inductor. The capacitor helps smooth out the voltage changes across the load - it can be noisy!

Bob is actually a bit more complicated than a simple transistor as shown in the diagram and has a feedback function from the load so that it can stabilise the voltage a bit better, but this only works properly if the feedback wires and resistors are kept well away from the inductor.

Step 3: Building the Circuit

The good news is that the circuit is dead easy to build and requires no special skills. I used a piece of hackable prototyping board to build the permanent version for bolting into the weather station control panel.


  • C7 Electrolytic Capacitor package 100 mil [THT, electrolytic]; capacitance 100µF; voltage 16V
  • C8 Ceramic Capacitor package 100 mil [THT, multilayer]; capacitance 10nF; voltage 6.3V
  • C9 Electrolytic Capacitor package 100 mil [THT, electrolytic]; capacitance 68µF; voltage 16V
  • D3 Schottky Diode package 300 mil [THT]; type Schottky; part # 1N5817
  • J6 Screw terminal - 3 pins package THT; hole size 1.5mm,0.508mm; pins 3; pin spacing 0.2in (5.08mm)
  • L1 Inductor package 400 mil [THT]; inductance 47µH; current 1.5A
  • R12 Trimmer Potentiometer package THT; track Linear; size Trimmer - 12mm; type Trimmer Potentiometer; maximum resistance 10kΩ
  • R13 1kΩ Resistor package THT; bands 4; resistance 1kΩ; pin spacing 400 mil; tolerance ±5%
  • U9 LM2675N ADJ/NOP8 package THT; hole size 1.5mm,0.508mm; pins 8; pin spacing 400mil; label LM2675N ADJ/NOP8; true; layout Double Row

The actual circuit is pretty much taken from the datasheet for Bob except that I have used a 1K resistor for the FB to ground connection and a variable 10K pot instead of the 3k resistor specified from the datasheet formulae.

PS. The LED is optional but should be easily removable as it will waste valuable energy during actual operation 'in the field'.

Step 4: Final

After exhaustive bench testing, this gadget is now ready to be installed in the Weather Station control panel, hopefully before the darkness starts to destroy it's lead acid battery too much!

Oh .... And special thanks to Bob at Texas Instruments .... You are such a great guy .... If I was in the USA I would buy you a drink!

Please in the competitions - top right - Thanks!
Hi. What about TSR 1-2450 modules. Do you find any problems with them?
<p>Does this project require programming?</p>
<p>No programming required!</p>
<p>Thanks for good design!</p>
can I use it for a long-term ?
<p>Yes, I would think so.</p>
<p>This is great! I will probably use this in some of my future arduino projects.</p>
<p>Thanks! Should help prevent burning out the native power supplies on arduinos.</p>

About This Instructable




Bio: I live on an island in the Irish sea called Ynys Mon which was once inhabited by the Romans, the Vikings and is still inhabited ... More »
More by Tecwyn Twmffat:Full English Breakfast Using Fruity Loops Music Production Software Part 1: Introduction Arduino GPRS Weather Station - Part 3: Hardware Upgrade 
Add instructable to: