Simple 3.3V Charge Pump Voltage Booster ...

Beau Schwabe

Serendipitous of something I am putting together for the EXPO, here is a simple charge pump approach to boosting the Propeller 3.3V up to 5V. You could probably drive this circuit directly off of a Propeller I/O pin without any issue, and eliminate both of the transistors in the schematic below, but just to be safe there is a a simple inverter driver. All you need to do is generate a square wave on the I/O pin connected to the inverter at about 5kHz. When the I/O is not oscillating, the output acts as a straight through (with two diode drops ... 2.1V) ... So essentially with a little bit of voltage feedback to the Propeller from the output pin, you could create a 'mini' digital power supply to 'dial in' a specific voltage.

The thread below discusses a similar approach using a buck-boost regulator to increase voltage ... http://forums.parallax.com/showthread.php?p=913571

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 6/10/2010 6:20:21 AM GMT

Peter Jakacki

I might be wrong Beau but running at 5kHz means you have to have big bulky caps. I have used a good ol' 555 for this kind of thing and the advantage is that they are self-oscillating, run at higher voltages, plus they can source/sink 200ma and they are cheap. The CMOS version operates down to 1.5V. Use schottky diodes for the pump to minimize the losses and run it fast enough and you can get away with very small caps. Indeed, I have used a 100nF coupling cap (pump) and a 10uF charge cap (storage) on the 555.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
*Peter*

ElectricAye

Hey, that's cool.
I've seen charge pumps for making high voltage but it never occurred to me you could use the concept to generate slightly higher voltages like this.

Thanks!

Beau Schwabe

Peter Jakacki,

The 5kHz is just something to wiggle the pin and get you started. The benefit of having a micro control the wave, is that a square is not always the optimal choice. Most of the time it works ok, but not always. In applications where you want to optimize the Input to Output impedance for example, it might be better to alter the duty cycle.

The optimization exercises are part of a learning curve I will leave to the user... this is just to provide some possible ideas of what can be done with just a few components placed in the right direction.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Peter Jakacki

I agree that having the micro "wiggle the pin" allows us to do more, even regulate as I have also done. The point was that the caps are huge really, don't you think? Using a 555 as the driver in place of discrete components is also an idea as it can be manually controlled quite easily or left to oscillate. My problem is that I am a minimalist if there is such a word, less is more.

Talking about step-up converters the more usually way is to use an inductor and typically they are part of a step-up converter circuit but the Prop can drive an inductor quite easily with just a simple NPN+R+diode+cap. The advantage of using an inductor is that you can get a much greater range of voltages and/or more current.

Beau, have you thought about setting up a Parallax page devoted to applications and circuit ideas? They could be fed from the forum and suitable entries admitted to the applications page.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
*Peter*

Beau Schwabe

Peter,

"Beau, have you thought about setting up a Parallax page devoted to applications and circuit ideas? They could be fed from the forum and suitable entries admitted to the applications page." - I have thought about it, and it's been suggested several times by others - one of these days, one of these days. Perhaps after the Propeller2 is complete I can focus more on something like that. Right now though, the only reason I'm up for air is for a few ideas I wanted to present at the EXPO. Until then, my capacity is limited to little 'pop-corn' circuits here and there.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Beau Schwabe

Here are some You-Tube videos of a Charge Pump like this in action slightly scaled up. This is part of what I'm working on and I hope to explain it a little better at the EXPO.

The game is really all about impedance matching, and Solar cells are notorious for having horrible characteristics. Truth is ALL power supplies of ANY kind have similar problems, just not as exaggerated as a solar panel.

Here is my analogy:

If Power is equivalent to how much time you put into something, say for instance walking a mile takes about 15 minutes or less. If you were to ride a bike for the same 15 minutes (on a flat road) you would use about the same amount of Power or energy walking a mile in 15 minutes verses riding 2.5 miles in the same 15 minutes.

What's different here between the two? impedance!! or mode of transportation ... The 'impedance' of riding a bike is better matched than walking given the amount of power you put into it.


Videos:
MPP Voltage (1)
MPP Voltage (2)

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Oldbitcollector (Jeff)

Incredible!

Beau, do you have some kind of production planned for this simple circuit?
I'd personally love to have a few tiny PCB's for this.

If not, perhaps this is a project I should put Nick's Diptrace tutorial to.

OBC

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller Feature Projects: PropellerPowered.com
Visit the: PROPELLERPOWERED SIG forum kindly hosted by Savage Circuits.

Beau Schwabe

Oldbitcollector,

Just curious ... So how much would people be willing to pay for something like this? ..Only 3 wires for hookup (Power In, Ground, Power Out)


...See, the way I see it is like those 'dumb' blinky LED lights you see advertized in a magazine that people will pay upwards of $10 for. ...and I keep saying to myself "you missed your opportunity there".

Ok, so I'll put together a kit/assembled version of this It won't be ready for this EXPO (too much going on as it is) but perhaps the next EXPO in Norwalk.


Keep in mind this is no miracle power circuit, it follows Ohms law very closely... it's just efficient about doing it.

Here is another 'real' example:

Our 9 Volt 1 Watt solar panels have an open voltage rating of about 10.5 volts which falls to about 9.5 volts after the sun hits them and they warm up. They have been optimized to operate at 9V and deliver 1 Watt of power.

Plotting and characterizing these cells I have determined that the 'sweet spot' for them to operate is with about a 100 Ohm load dropping the nominal operating voltage to 8V ... 8 Volts across 100 Ohms is 80mA or 640mW.
The other 360mW (remember rated at 1 Watt) is burned up within the solar cells own internal resistance.

Running the same Solar panel through this circuit with a 330 Ohm resistor increases the nominal operating voltage to 14.8V ... but guess what? It's still at the same Power ... 14.8V across a 330 Ohm load is 44.8mA or 664mW ... <-- A little bit better than just the 100 Ohm, but mostly because the internal resistance of the solar panel has shifted a little bit because of a different load.

What does this mean? Sure you have half the current, but your available voltage is higher at very close to the same rated Power of the solar cell. In other words, that 9V 1Watt Solar Cell 'could' adequately charge a 12V battery, otherwise you would need two of them just to get the voltage where you need it and even then (at 18V) the impedance would still be mismatched.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 6/12/2010 4:10:20 AM GMT

Oldbitcollector (Jeff)

I'd gladly pay $3.00 each for a tiny pcb which these parts could be placed. (Put me down for 3 or 4 immediately)
Bet you could put several break-away versions of this on a single PBC run.

I'll bet you could easily get $8-$9 for a kit.

OBC

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller Feature Projects: PropellerPowered.com
Visit the: PROPELLERPOWERED SIG forum kindly hosted by Savage Circuits.

Beau Schwabe

Jeff,

I had to re-read what you said ... you'd pay $3.00 for an unpopulated board? ... I just did a quick BOM of the parts that go into this and the parts alone are almost $6 ($5.83 - According to Digi-Key) ... that doesn't include the board to put the parts on.

In bulk I could probably get the parts cost down to $4 or less, but the main culprits are the MOSFETs and the Caps. The MOSFETs are probably overkill at 50 Amps and cheaper cousins could be substituted, but there is a reason I'm using the ones I have decided to use.

Tonight I'll transfer the design to a Perf board from the SBB to see how much real estate is necessary and then I can see how much a board would cost. $3 might be well within the limitations there.

A realistic cost for me to break even on your request would be $2.833 if I can get 18 boards out of expresspcb. That means 6 breakout units from each of their three 2.5"x3.75" boards they offer for $51.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 6/12/2010 4:03:27 AM GMT

ElectricAye

Oldbitcollector said...
I'd gladly pay $3.00 each for a tiny pcb which these parts could be placed. ....

Just curious: where did you get the $3.00 figure in the first place? wishful thinking???

Beau Schwabe

He might have done what I did in reverse... the $51 boards from ExpressPCB, you get three of them, so that's $17 a board. Since Each board is 2.5"x3.75" ... you look at it and figure yeah, I could get about 6 on there. ($17/6 = $2.83 ... Jeff just rounded up.

That or the stars were aligned and he just had a good guess

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Oldbitcollector (Jeff)

Yes, I was figuring on blank boards, estimating the size.

I've been doing some homework in this area preparing to do some boards of my own sooner or later.

OBC

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller Feature Projects: PropellerPowered.com
Visit the: PROPELLERPOWERED SIG forum kindly hosted by Savage Circuits.

william chan

Hi Beau,

Can this circuit be used to boost 12v to about 15 volts?

If the propeller stops toggling, should it idle high, low or input?

Would both transistors conduct at the same time if the prop pin becomes input?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.fd.com.my
www.mercedes.com.my

hinv

Another way to controll the output of a solar panel is to control it's exposure to the sun. Once on a trip, I had a harbor freight 12V type solar panel in the box and my son's 7.2 volt battery pack for his truck. I connected it up direct while the panel was in the box, and slid the box off until the voltage was about 8V. It worked pretty well, but I essentially wasted most of the solar panel. I also had no way to measure the current, and I didn't watch the voltage so it probably went higher than recomended.
Could you make a simple charge controller that would boost or regulate down by dialing in a voltage?
Circuit Cellar had a good article on inductor based voltage conversion, but some of it is above my head.

Thanks,
Doug

Beau Schwabe

william chan,

"Can this circuit be used to boost 12v to about 15 volts?"

Yes, at a reduced current ... suppose your supply was only capable of delivering 1A at 12V ... to maintain 15V you would only be able to use 800mA. It's still the same Power. (12Watts)


"If the propeller stops toggling, should it idle high, low or input?
Would both transistors conduct at the same time if the prop pin becomes input?"

At this position in the circuit, a Propeller is not involved, so this is not a problem. If the circuit stops toggling, then the output acts as a pass through with two diode drops.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

tonyp12

Would it be safe to use the Prop to do this?


Conduction through the Props "protection" diodes may cause problems?

The PIC used is probably at 5V, so a prop with 3.3v may not have the same boost.
The two-stage charge pump adds nearly 10 volts to the battery voltage Vbatt using a PIC

Using Props NCO/PWM differential, set and forget.

Phil Pilgrim (PhiPi)

As long as the pins are always outputs, there shouldn't be a problem.

-Phil

tonyp12

But if doing a reset and while bootup, the props pins are not outputs.

Should the 12Vbatt be controlled with a mosfet?, Vg is in mA range.
But would use another Prop pin to turn the small signal mosfet on.

Phil Pilgrim (PhiPi)

You neglected to mention that V_BATT was 12V. I assumed 3.3V, which would be the max for that circuit on a Propeller.

There are probably no issues on reset to be concerned about. At least I can't think of any situations where having the drive pins tristated during that time would be a problem. You just want to make sure that both drive pins get set to output simultaneously and stay that way once in operation. Also make sure that the combination of drive frequency, cap values, and load do not cause excessive output currents into or out of the drive pins.

-Phil