Probably really simple, voltage into current..?

[magician13134]

Is it possible to convert excess voltage into current? If so, how? I know they're related with Ohm's Law... but I don't know how to actually do it. Say I have a 12V 70mA power supply and I want 5V and 150mA, is that possible?

[timv]

You didn't say whether you have AC or DC voltage there. It matters a lot.

For alternating current, that's what power transformers do. Feed in voltage and current on one set of windings, get out a different voltage and current based on winding ratio (but about the same power, minus a small loss from less than perfect efficiency) on the other side.

For direct current, it's tougher. Typical (eg, 7805) voltage regulators just burn off the excess voltage as heat. DC-to-DC converters like Mintyboost are more involved and most often used to increase voltage, but there are examples that go the other way. I've seen "buck" circuits that do this. Look for applications for driving high-power LEDs, where a designer needs to drop a 12 volt supply to about 3 volts for the LED and might not want to throw away the surplus power--especially when running off batteries.

[quick google] This looks like it might be along those lines:
http://www.linear.com/pc/productDetail.jsp?navId=H0,C1,C1003,C1094,P13452

In the normal electrical engineering world, I think the typical practice in the situation you described would just be to spec a different power supply with the right voltage!

[magician13134]

Well I found a really good deal on solar panels though, 2 12V, 70mA cells for $3... I only want 5V out of them, but would prefer higher current. So regulators don't increase current at all?

[timv]

Well I found a really good deal on solar panels though, 2 12V, 70mA cells for $3... I only want 5V out of them, but would prefer higher current.

At that price, buy more panels and wire them in parallel!!

So regulators don't increase current at all?

Nope, sorry, at least none that I know of. You'd need to have something inductive inside, either a transformer or a coil, to do that. Regulators are just a few transistors.

Actually, that is an interesting problem just in terms of making efficient use of the power from the panels, since efficiency is such an important issue for them. But the only way I've ever seen to do real DC-to-DC conversion, up or down, is by driving an oscillator and using an inductor to convert the voltage via an electromagnetic field.

I still want to try making a Joule Thief:
http://www.hackaday.com/2007/10/10/joule-thief-led-driver/

That's the simplest DC converter I've seen. It's designed to boost voltage, as most are, but maybe it could be adapted. They also point out that it isn't particularly efficient, but if you're just trying to get an almost-dead AA battery to light a bulb that's not such a worry.

It might or might not do better than the basic 7805 in your case. But I'm sure you'd learn more if you tried it. Then you could tell the rest of us all about it.

[timv]

Oh, just remembered: An individual photovoltaic cell only puts out something less than .5 volts. A 12 volt solar module is made from a bunch of cells that are wired in series. So it's *possible* that you *might* be able to rewire it to make a 6 volt module with twice the current capability.

Depending on how the connections between the cells are made, the odds of that actually working might be pretty small. But it's something to consider anyway, and a good thing to remember about photovoltaics.

[rglenn]

A switching regulator can turn a higher voltage at lower current, into a lower voltage at a higher current (or vice versa). You'll probably be in a "Buck" configuration.

Keep in mind that the efficiency of this kind of regulator isn't perfect - in simpler versions, you might be looking at 75-85% efficiency (so if you put 1 Watt in, you'll get 0.75-0.85 watts out). This is vastly better than normal linear regulators, which have simply awful efficiency (less than 50% in some cases).

[timv]

Saw this while searching for something else. Think it might work?:
http://beta.mouser.com/Search/ProductDetail.aspx?qs=q2eLhcXmQbhidZrYm51C%252bA%3d%3d

(Mouser's part number is 595-TL2575HV-05IKV if that link isn't good.)

Data sheet shows application circuit. Extra components are an inductor, a diode, and two capacitors. And it isn't surface-mounted.

And if I'm reading the specs correctly, it might actually be efficient to feed it 18V with two modules in series.

For what it's worth, and keeping in mind that I do software not hardware...

[Entropy]

Yup, that's a fixed-output switching regulator, which should work.

Linear Technologies also has quite a few products in that category.

National Semiconductor has quite a few switching regulators to choose from.

NatSemi is by far the easiest to get samples from. TI is also pretty easy.

Linear is a bit harder, BUT their products are much easier to simulate since they all have built in models in LTSpice.

[rickyrockrat]

Use a LM2575. It is appropriately called a "simple switcher". All you need is a coil, a diode, the part, and 2 capacitors. The datasheet says you can get to 96% efficiency, but I would figure on 90% at best.

Figuring on 90% eff, at 12V, 70mA, you should be able to get:
12V*.07A*.9=.756W
@5V, you would get .15A (150mA) (7.56W/5).

Part:
LM2675N-5.0:$4.00
http://www.national.com/pf/LM/LM2675.html
http://search.digikey.com/scripts/DkSea ... 75N-5.0-ND

Diode:$0.31
http://search.digikey.com/scripts/DkSea ... F54GICT-ND

Caps (2):$0.38
http://search.digikey.com/scripts/DkSea ... =P10294-ND

coil:$1.68
http://search.digikey.com/scripts/DkSea ... 53-1395-ND

[timv]

lm2575hv == pin-equivalent to tl2575hv. The sample application circuits in the data sheets even give the same part numbers.

Do I get a cookie?

[rglenn]

[timv]

[rglenn]

I don't know... the one I posted looked tastier.

[disinteger(-13)]

Linux gets cookies? hehe