Go Down

Topic: Help (Read 891 times) previous topic - next topic


TomGeorge

Hi, sorry it won't work, it relies on all the regulators to be exactly 3.3V. That is not possible.
If you look online at the major electronics suppliers I am sure that you will be able to find what you are looking for.

Tom.... :)
Everything runs on smoke, let the smoke out, it stops running......VK3DMK

macegr

#2
Oct 10, 2013, 10:21 am Last Edit: Oct 10, 2013, 10:25 am by macegr Reason: 1
Could your question be phrased as below?

Quote
I need 3.3V at 3A, and have a 12V/3A power supply. I have some fixed-voltage 3.3V/1A linear regulators. Can I put them in parallel and get up to 3A at 3.3V from the combination?


If so, then it's highly unlikely. The first problem is that your regulators will not output exactly the same voltage; whichever one has the lowest voltage (even a small amount is enough) will attempt to supply all of the required 3A current. It will overheat and fail quickly.

The second problem is that your regulators are rated for 3.3V/1A under great conditions. First, that they don't have to dissipate too much power, and second that they are heatsinked enough to keep them near room temperature. It's unlikely that you'd even achieve 3.3V/1A from the top circuit. Your regulator will attempt to burn off (12V-3.3V) * 1A = 8.7 watts as heat. In a SOT-223, that's a lot of heat to get rid of somehow. If you look in the datasheet, you should find a table or graph that indicates what current ratings are possible at various junction temperatures. And you should also find a case dissipation rating (degrees per watt) that will help you find the difference between ambient temperature (what your heatsink can manage) and junction temperature at a given amount of dissipated watts. Then, if you're specifying a heatsink, you'll need to find out how many watts it can transfer to the air depending on various ambient temperatures.

Sound like a tough problem? It is...it's an engineering problem, not just plugging parts together and expecting them to work. You can significantly reduce the amount of risk/worry by choosing a different regulation solution:

* Find a premade regulated power supply with the desired amperage. Everything is already handled for you.

* Change the power supply to 3.3V + regulator overhead and not much else, and use a linear regulator that can supply at least 3.3V. This will greatly reduce the amount of heat dissipated from the regulator.

* Instead of linear regulators, choose a switching buck regulator design. This will have much greater efficiency, so not only is heat not a big problem, but you also aren't uselessly wasting twice the power your circuit actually needs. There are simple switching regulator designs like the LM2576 that don't require many external parts.

* Find a premade switching step-down regulator board or DC-DC converter that can supply the voltage and current you need.
Unique RGB LED Modules and Arduino shields: http://www.macetech.com/store

cjdelphi

A drop from 12v to 3v is a bit much to begin with,  use a switching regulator to drop it to 5v

Now if you wish to continue to use this method , you'd need  a low ohm resistor value on the end of each output, making sure 1 will not take all the current over the others.

Another method would be to use  12v, buy the adjustable LDO of the AMS117 range, set it .6v higher than 3.3v to accommodate for the voltage drop, and have it output around 3.9v, you then pass it to a beefy transistor where the base gets 3.9v (using a 1k resistor) this keeps the current right down and the AMS adj regulator very happy.

Now, the bad ass transistor takes on the full 12v onto it's collector and on the emitter out pops around about 3.3-3.4v depending, if it's not correct adjust the regulator via a potentiometer, now we got the correct voltage on the emitter you can then have it control several amps,  and the heat dissipation problem is solved.

Failing that, buy a switching regulator. step it down to 5v then supply it to  AMS117.


Go Up