Hi,
I don't have an EE background and I can't find explanations of MOSFETs that I can understand. But what I've read, it seems like a MOSFET can regulate both voltage and current by taking the input voltage and current, then "using" voltage to increase the current, so the output will be a lower voltage but higher current. And the vice versa can happen - decrease the current to increase the voltage.
Is that true? Could I send in 5V @ 40mA to output 3.3V @ 60mA?
dlee83:
Hi,I don't have an EE background and I can't find explanations of MOSFETs that I can understand. But what I've read, it seems like a MOSFET can regulate both voltage and current by taking the input voltage and current, then "using" voltage to increase the current, so the output will be a lower voltage but higher current. And the vice versa can happen - decrease the current to increase the voltage.
Is that true? Could I send in 5V @ 40mA to output 3.3V @ 60mA?
Not with a MOSFET on it's own, and you couldn't get 60mA. More like 55mA due to losses in the process. (An ideal converter, with no loss, would output almost exactly 3.3V at 60mA. But nothing is perfect.)
You're getting the idea though. Output power is 85% to 90% of input power in a well-designed DC-DC converter. Maybe even a little better.
The MOSFET would be switched on and off at high frequency, with an inductor and diode to smooth the output by keeping the output current flowing at those times when the MOSFET is switched off.
Google "Buck Converter", and you'll find lots of info on the subject.
Edit: Also Google "Boost Converter", for step-up DC-DC converters. They use a similar technique in a 'flyback' arrangement to output a higher voltage than the input at a lower current.
Desigining something like that yourself from scratch is NOT simple. But, you can buy a step-up voltage regulator chip and build the manufacturer's recommended circuit. Unless you are a chip designer, that's how it's usually done. (Before those special chips were introduced, switching regulators were complicated and expensive.)
In this kind of circuit, the MOSFET is "just a switch". (You could also use a transistor.) The "trick" is the inductor which stores the energy, and switching the current through the inductor on & off precisely (usually at a high frequency) with feedback to maintain a constant voltage.
When you run current through an inductor, the magnetic field builds-up. When you remove the power source, the magnetic field collapses and the coil generates electricity. If there's a (relatively) high resistance load connected when the coil is generating/releasing energy, you'll get high voltage and low current. The hard part is keeping all of this under control... You usually don't want a "high" voltage.
If you've ever seen one of those "pranks" that uses a small battery to shock someone, those use a coil (inductor). The spark plug coil in your car works the same way to generate thousands of volts from a 12V battery.
Why not just buy an LDO 3.3v regulator ?
Could I send in 5V @ 40mA to output 3.3V @ 60mA?
Excluding conversion losses sure.
example:
A house is supplied with 240 V, the watt meter says 7.2 KW is being used therefore: 7200/240 = 30 amps draw.
However, if each appliance is 120 V that means they are drawing ~60 amps = 7.2 KW.
30 amps form the utility 60 amps to the appliances.
5V x .04A = 200mW
3.3V x .06A = 198mW
198/200 = 99%, I don't think any converter is that efficient.
Check the boost converters at
http://www.pololu,com
they have some that could be 95%.
They have some that boost from a single battery too, let you drain it right down to the bitter end 8)
You can't drain a lipo battery below 3.2V/cell without destroying it.
That's why I said to the bitter end 8)
Probably better to use that one with an alkaline battery where recharging is not planned.
they have some that could be 95%
As Crossroads implies there is a limit to what can be achieved.
I picked up some of these, the best I calculate was 96%.
Here
I don't think that dlee83 was actually planning to design/build his own buck converter. I got the impression that he was just interested in the theory.
But it's not at all hard to build one from scratch with reasonable efficiency, about 85% or so.