charge pump to boost 3.7v to 5v?

[jamie70]

Hi,

Just wondering if anyone could help me out with building an analogue charge pump circuit for a voltage boost from a 3.7v lipo to 5v? I know there are chips and dc-dc converters out there but I want to give this a try with all the resistors, diodes and capacitors I have lying around. I found this description surfing google:


voltage doubler can be achieved by connecting the "top" terminal of a capacitor to the + supply via a diode. the "bottom" cap terminal is switched to connected to EITHER + or - input supply.. When this switch (constructed from two transistors?) connects to - the capacitor charges up per normal. But when this switch connects the bottom of the cap to +, the voltage at the top is raised by how much the bottom is raised. If you had a perfect diode/switch this would result in a doubling of the voltage. Have a diode from the top of the capacitor to a 2nd capacitor, and the current wont flow back into the pumping capacitor when it's recharged.
Could use a voltage regulator to drop the voltage, but a better method might be use a voltage reference/ comparator to detect when output voltage is lower than a preset valve. Only when this is the case need the "connections" of the pump capacitor be switched between + and -. If the storage capacitor is bigger than the pump capacitor, the amount of charge each "stroke" of the pump capacitor delivers will only be able to raise it's voltage by a small amount. Several strokes being needed to charge it up fully Even so a transient voltage suppression diode on the output line, might be of some use in ensuring there are no over voltage excursions. There are ICs available which provide the oscilation/voltage test/regulator in a single package. Like the following which also provides a simple breakdown of the concept. It replaces the two diodes, with switches for better efficency...


I found this diagram in a link from that description:

Note that in the circuit the switches could be replaced with diodes.

I also have some RS bits that may help : 555 timer, PNPs and 7805 Voltage regulator My main problem is figuring out the values for caps as well as generally needing some guidance. I figured from some research that c1 doesn't have to be huge, but c2 is "storing" current so it is bigger.

Thanks for any help!

 

[slimest]

If you want a booster with a high current output, you need an inductor for it. Capacitors work only for low current schemes. The best way - using of special chips, such as, for example, MAX1708 or MAX1709. It's easy to google their datasheets.

 

[jamie70]

Hi,


ok thanks. I wasn't sure about current possible from this setup.

 

[breaktru]

Looks like short circuits to me with S1 & S2 closed

 

[bstedh]

I think the physical size of the capacitors you would need to make this work would be bigger than the batteries you would need to power it.

 

[jamie70]

I noticed that inductors in similar circuits on usb chargers, etc are pretty tiny. I found a website with an iductor coil calculator for #windings/guage based on req'd output. I suppose one could use magnet wire wound on a form and make one for this simple circuit.

If the circuit could be built, even with large caps, I'd still like to see it done. I'm a chronic DIYer and terminally stubborn

 

[AttyPops]

They do make a 1/2-way-there version where you add your own inductor and caps/resistors if you are interested. Would save you a lot of work rather than trying to make it from discrete parts. Look at digikey or mouser at the power regulators and see. Some you just add an inductor and some caps/resistors.

Also, you'll need to pay attention, whatever approach you choose, to the amps x volts (watts - power dissipation) for the entire circuit. E-cigs draw a lot more amps than most electronic circuits since they are heaters.... the normal small stuff tends to melt, and you get the magic blue smoke.

EDIT: Kinda duplicate post... see post #2

 

[jamie70]

Hi,

So I found a ic117 v regulator capable of 1.5 amps in an old cd burner.
I'm going to try it out with a darlington pair and maybe an inductor.
Also will replace R off the ADJ with a pot for variable V. If it works nicely I'll post
Details.

Posted from iPhone, excuse the short forms.

 

[CraigHB]

That's a bit low for current output. You might want to look into using a PWM controller with a built-in switch. The MAX1709 mentioned already is a good one, 10 amp switch. I've used it before and it's really easy to wire up. Nice part, voltage mode control, no external compensation network required. It has a good application example in the data sheet. I used a Bournes SRP7030-1R0FM inductor and a couple Kemet T520D337M006ATE010 capacitors. Used a Diodes Inc PDS1040L-13 rectifier. Worked really well, quiet and stable. Got close to 90% efficiency out of it at 5V/2A output off a single Li-Ion battery. Relatively expensive though. Would be a far cry from scavenging parts.

Here's a link

 

[jamie70]

Hi

CraigHB, that sounds like a nice build. Did you ever post details of it here on ECF?

I'd love to see photos or a circuit drawing/schematic of what you did.

Right now i'm playing with an 8 pin dip mc3406 I got from a buck circuit in a 12v charger. When I looked it up on alldatasheets it is multi purposed for step-up as well. Its rated 1.5a but has a schematic for higher A output with a darlington configuration. I'm trying it out on my breadboard right now, but its hurting my brain a little since i'm pretty new to electronics.

Thanks for you leads. What you built is exactly what I'm working toward.

 

[CraigHB]

I never did post anything here. I'm actually going with a better part right now that uses a synchronous rectifier with external FETs. It's more efficient. Don't have any photos though. Not much to it, just based on the application example in the data sheet. Though it requires an external compensation network which can be pretty difficult to engineer, even for a professional. Definitely not something a layman could do.

The MAX1709 is a nice part, ideal input voltage range and power output for a single battery mod, but it's asynchronous, less efficient. Though it does have a nice energizing FET built in and that's a big plus. Also, the built-in compensation network makes it real easy to wire up, a layman could do it.

My assemblies look similar to the off-the-shelf switching converter modules you can buy. I haven't found any with the specs I want so I just build my own. I also want them in a thinner one-sided profile and all the off-the-shelf ones are two-sided and too thick for my form factor preference.