1.5V to 5V

[Sambuchi]

I need a solution to go 1.5V to 5V

I dont have a pwr requirement as of yet... but I would guess the load would be around 100mA at 5V

I've found one solution that was a switching regulator from TI.

Just wanted to hear your thoughts.

Thanks

Tony

[jwax]

How about a nice little motor-generator set?

[haklesup]

Clearly you will need some sort of boost switching regulator. a good number of IC companies offer regulator chips. Maxim-ic has a good listing of such devices. National Semiconductor has the WEBENCH application whereby you can design an entire switching regulator on their web pages. That may be the easiest route.

Of course there are likely off the shelf DC to DC converters and they may cost the same as DIY depending on if you have to make a PCB and buy more than Qty 1 on some parts.

Remember, power laws apply, Pin = Pout +conversion losses. Most of the ones listed on Digikey have 1A to 10A inputs. You'll want to use D cells for this one.

In most cases it is better to design 3.3V circuits and use a 3.7V lithium pack if its for portable devices.

[Bigglez]

I need a solution to go 1.5V to 5V

I dont have a pwr requirement as of yet... but I would guess the load would be around 100mA at 5V

I've found one solution that was a switching regulator from TI.

What is the 1.5V source? If its a cell you will need to
define the end-of-life voltage (i.e. dead cell) and use
that for the step up stage. Unless size is critical why not
just install four 1.2V (NiCad) cells?

[Sambuchi]

my design does call for a AA size cell.

I am aware that I could use a 3.7 AA cell but would like to keep it at 1.5

haklesup thanks for the chip manufactures. Its been awhile and I new I was forgetting some. National Semiconductor does make some nice parts and I ordered some to test.

[Engineer3104]

Check out this unit from Dimension Engineering. I think it will do exactly what you want.

http://www.dimensionengineering.com/lvboost.htm

[haklesup]

That is a nice little off the shelf converter. Using their example of 5V,300mA output, it needs 1.5V, 1.33A at the input. With a typical 1500mAh NiMh battery, you would get approximately 1 hour of use. Not sure if that's sufficient for your application.

If a particular battery life is required, you will also need to do a little research to select the right battery. Look for discharge curves in the datasheets. The lower the V the more I the converter will need so depletion will accelerate dramatically as soon as terminal voltage starts to go down. Just dividing the batteries capacity by the current used may yeild an expected life longer than you actually get.

[jollyrgr]

Check out my reply to a posting for a similar project. Mine is the 8th reply and has a schematic posted.

http://forum.servomagazine.com/viewtopi ... led+single


Replace the LED with your circuit. Possibly add a electrolytic cap on the output as well.

[Bigglez]

Check out my reply to a posting for a similar project. Mine is the 8th reply and has a schematic posted.

Replace the LED with your circuit. Possibly add a electrolytic cap on the output as well.

That's a clever circuit but it has some limitations. As shown
it only delivers 14mA average to the LED, the OP is looking
for 100mA at 5V (500mW). Energy for the output is limited
by how much is stored in the inductor on the 'charging'
half of the cycle. During the 'discharging' half of the
same cycle the output voltage is limited by the LED Vf
(typically 3.6V for a white LED).

Key to operation is the ratio of battery voltage to Vbe
of the transistors, and while the circuit can start at
Vbe (of Q2) plus a little more to get base current flowing,
it can't run at higher input voltages without over
driving Q2 base.

When L1 is fully charged the Q2 current will shoot up,
and that allows the Q2 Vce voltage to rise enough to
turn on Q1 (via R1,C1) and turn off Q2. Q1 is starved
of base current as the LED voltage falls while the inductor
dumps current into it. The key to starting another cycle
is that Q1 over powers Q2 and conducts when there
is no stored energy in L1.

There is no regulation - The LED will Dim as the battery
dies, until the battery can no longer forward bias Q2.

[MrAl]

Hi Tony,


In case you are interested, Zetex makes a line of ic chips that
were designed for use with higher power LEDs. They also make
excellent dc to dc boost converters, with low input voltages like
1v and even a little lower up to maybe 3v. That means you
can use these chips with one or two 1.2v or 1.5v cells easily.

We had a HUGE discussion on these chips way back in 2002 on
another forum and one person was even nice enough to fund
complete testing on one of the chips in the typical application.
Various people worked on the project and in the end we got
the board size down to the size of a Minimag flashlight head
diameter. The parts are all surface mount, but are rather easy
to deal with because at least the one we tested had only
5 pins, and the transistor only 3 pins, and the special Schottky
we used only 2, plus two SMD caps and one home made
resistor made from a length of copper wire.
The part number for the chip was ZXSC300, and the diode
was ZNCS2000 if i rem right (both Zetex parts). Dont be
fooled by other Schottky's either, this one is the best!
There are design examples on the Zetex site.
Also, since then Zetex came out with more advanced versions
of the same part that have voltage feedback for regulating
the output to a set point fairly well. One part is 310 (same prefix
as above) and they also came out with the '400' family, same
prefix but 400,410 something like that.

One of the nice things about these chips is that they have
peak output current control to protect the output transistor.
Also, the sense resistor can be made out of a short length
of copper wire (i think we used about 2 inches) or you can
buy special very low value resistors like 0.020 or a little
lower or higher, but the copper wire is the cheapest.
Sense resistors that low keep efficiency way up.

[Engineer1138]

Hey, thanks! I was the originator of that old thread and I had forgotten all about it. As luck would have it, I have been thinking of building an LED flashlight, so that circuit should come in handy

Check out my reply to a posting for a similar project. Mine is the 8th reply and has a schematic posted.

http://forum.servomagazine.com/viewtopi ... led+single


Replace the LED with your circuit. Possibly add a electrolytic cap on the output as well.

[Bigglez]

Check out my reply to a posting for a similar project.

Here's a SPICE run for that original circuit:

[Bigglez]

Check out my reply to a posting for a similar project. Mine is the 8th reply and has a schematic posted.
...
Replace the LED with your circuit. Possibly add a electrolytic cap on the output as well.

I tried it and here's the SPICE results:

[Bigglez]

Replace the LED with your circuit. Possibly add a electrolytic cap on the output as well.

I did some tweaking and got the following results:

[MrAl]

Hi there,


Believe it or not a circuit very much like that two transistor circuit
was used in a commercial product. I assisted with the analysis
and development and it turned out to be fairly interesting for a
little circuit like this. I can not disclose too much information about
it because of a privacy agreement (as i am sure anyone else would
want too if they were using their circuit in some product) but i can say
that it is possible to get a decently high voltage out of this thing, like
10v or higher, just in case anyone needs a little higher output.

Engineer, are you talking about the thread that started on
another forum about the Zetex 300 chip back in 2002?