Which regulator for 30V DC to 5V DC

I need run arduino at 30V DC supply..
Which regulator should I use to drop the 30V DC to 5V DC..??

The maximum current drawn by the circuit will be around 500mA - 800mA
I dont want the regulator to get much hot, as I dont have much place for a heatsink..

Use a DC-DC convertor, also known as switching regulator.
Some examples:

The high input voltage to me suggests the use of a switchmode DC-DC power supply. Using linear regulators on a high input like that means a lot of heat dissipation since a linear regulator would literally convert 83% of the incoming power into heat, i.e. about 20 Watts (!!!). A good switchmode power supply will likely achieve around 80+% of efficiency so it'll not only be much more efficient, it'll also run much cooler!

One drop-in solution is the dimension engineering line of power supplies. They have a fixed 5V model, and adjustable-output ones too. Their higher-capacity line (3A-capable) tolerates up to 35V input. I have had good luck with these power supplies.

I would always double-check to make sure that whatever DC-DC solution you're looking at can handle the very high input voltage. Few modules can, though Crossroads linked to a great set at Pololu.

Its such a large voltage ratio that you may find a two-step reduction via 12V is
better (more choice of converters). This gives you a 12V rail should you ever need it
too.... Although there can be issues stacking converters if the last one needs high
start-up current.

Another possibility is DC-DC converter down to 8--12V or so, then a linear regulator to 5V,
which means the 5V rail has no DC-DC converter noise on it. This would be a good strategy
for sensitive analog circuitry, but less efficient than DC-DC conversion the whole way.

Most DC-DC converters leave 50mV or so of switching noise on the output, which for
instance can affect sensitive audio amplifiers.

I'm intrigued with having 30V DC as the available power source - rather exotic.

there are lots of dc to dc boards on ebay like this, very cheap my advice is find what chip it is and make up the circuit you will learn so much from doing it.

http://www.ebay.com/itm/LM2596-DC-DC-3-2-40V-to-1-5-35V-Step-Down-Adjustable-Converter-Power-Module-/271458822487?pt=LH_DefaultDomain_2&hash=item3f3434e957

I've decided I want to try to make most of my future projects capable of 7-32v or so operation. I like the eBay switcher modules. They are fairly efficient but I think there was a bad batch that was rated at 42v but only used 30v input capacitors one time so I choose versions where you can see the rating on the input capacitor in the picture. New ones say 50v which should be perfectly good for 30v use.

Given the prices over at Pololu and the fact they they're a respectable merchant, I'd source their DC-DC power supplies over a 'unknown' merchant on eBay. But a lot has to do with living in the same country as Pololu, which the OP does not. Thus, an Ebay merchant with a very high rating (99%+) may be a better bet and closer too (re: shipping).

jonwhite:
my advice is find what chip it is and make up the circuit you will learn so much from doing it.

No making a stable switching regulator is difficult. The PCB layout is critical, I have never known even an experienced professional, PSU specialist, engineer get it right first or second time.
You have to know what you are doing. If not you are just wasting your time and money, and unless you have proper test equipment you can never succeed.

There goes another Karma point. Cannot agree more. While copying Gerber files, layouts, etc. published by switch mode regulator component manufacturers can yield perhaps the desired results, these types of devices are inherently more difficult to implement than most folk seem to give them credit for. Which brings up a good point, i.e. how confident is anyone re: the layout practices and EMI / EMC compliance of boards sold on Ebay?

While I briefly considered the switch-mode route for a design I've been working on, I ultimately decided to keep it simple, not extract every mWh from the battery pack and simply run my rig straight from a Lipo battery.

Coming back to the request from the OP, I'd buy from a reputable vendor, i.e. one that has put the time and resources in to ensure the switch mode power supply is going to work properly. Or, consider adding an additional power supply that simply takes line power instead. USB chargers are very inexpensive now and reputable brands don't cost an arm and a leg, yet offer 1A+ charge performance at 5VDC.

here is one i knocked up a while ago using the lm2575 ist attempt (reg1), worked really well just followed the data sheet which recommends every track to be as short as possible i then surrounded it with a gnd plain, everything on the board worked.

noise was an obvious concern so i knocked up another 2 identical circuits (lm2575’s) then placed them on top of the processor still everything worked as it should. i have experienced very noisy environments ie equipment that effect the operation of processors you can see straight away as it up sets the functionality of it also another good way to see noise is to monitor serial data from the processor noise really does upset the data.

i then used this board to see if it interferes with other equipment such as mobile phones and even placed on top of a i7 processor under load and didnt effect the operation of it i couldnt get that circuit to effect anything, my version of emc testing :wink: so it can be done, building stuff you learn so much from it.

worked really well

So what tests have you done to confirm this.
Most of the problems with switch mode power supplies come at specific input voltages / current loading values. It takes a lot to test them and you need a good scope. I suspect your tests have not been rigorous but I could be wrong. What tests have you done and how many have you tested?

What if I use a 7812 (12v regulator) on the 30 volt and then again a 7805 (5v regulator ) on the 12 v output after that…??

Will that be too stupid activity…:stuck_out_tongue:
I just need to run a atmega328 and 4nos. of 7segment displays

I hope none of them will get too hot if I am drawing only 400 - 500mA as load

Well, if you already have a bunch of 7812s and 7805s, and heat sinks for them, and you are not concerned with burning off about 12W of power, why not use them. You'll probably need a good heatsink for the 7812. Run the circuit for a minute, if the 7812 gets too hot to touch, you are overloading it. There are also ways to use these regulators as the "brains" for higher current pass regulators - see the application notes on the data sheets.

Otherwise, DC to DC modules are cheap and widely available, and can cost only a little bit more than two 7800's and all the heat sinkage you need for the linear power regulation. They are much more efficient if you are running off batteries.

I'm intrigued with having 30V DC as the available power source - rather exotic.

I have found lots of 48V-secondary transformers in the trash. I have a whole bucket of them. Which is odd because 48V DC is a standard telco voltage, and you'll get less than that when rectifying 48 VAC. But, anyway, they are common. So you get about 30VDC from rectifying that.

I bought a handful of these: http://www.seeedstudio.com/depot/Adjustable-DCDC-Power-Converter-125V-35V3A-p-1534.html?cPath=1. Although I would take the 3A claim on the one above with a grain of salt given the size of the heat sink. It is good for at least 1A without getting noticeably warm.

w

Hi guys, im on another thread wih an essencially the same problem , I have found this Electronica y Robotica Electan, Tienda On Line

I would like to know if this will be able to handle spikes caused by driving dc servos motors, and if I can stack various to power varios devices without any problems. All will be powered from a 16.8v 5ah battery

Send e-mail to pololu.com and ask.
With 700 KHz switching frequency, I would think any spikes would be rapidly taken care of.
Not sure what you mean by stacking.

What do you all think about the LM2596S-5.0

Here is the datasheet
http://www.farnell.com/datasheets/1768026.pdf

Dear Joy,

Use LM317T .. A very good regulator I.C. with thermal overload protection.
No need to use 2 muliple regs.

Dear Joy,

25965 is great.. but study your current consumption.
Do you really require this IC ? Also a good PCB design is must for this IC.
On other hand, 317 is easy to use and no special PCB layout needed ...if your current requirement is not going beyond 1.5A

Amarinf:
Dear Joy,

25965 is great.. but study your current consumption.
Do you really require this IC ? Also a good PCB design is must for this IC.
On other hand, 317 is easy to use and no special PCB layout needed ...if your current requirement is not going beyond 1.5A

Using a linear regulator for 1.5A with a 25V voltage drop is sheer madness, that's 37 watts
of wasted power to dissipate as heat, a DC-DC converter will be 85% efficient or so.

Use a switching converter, not a linear regulator.