Backyard Aquaponics
A place to discuss aquaponics

Must be great the first time you get a utility bill that's actually credit

Nah, It's not so bad. The best way is to start with ICs such as those from Linear Technology.
They are VERY easy to get working. You just follow their recommended circuit and layout and thats it. Easy as pie. They use what's called current mode feedback which is much more stable and straighforward than ICs from National Semi (I shudder from pain of memory)
Don't really need a scope unless you are super picky about the output waveform. And if you are charging a battery then it really doesn't matter..

I dont really do sensible.

I'm not so good at at it

I've found a few kits, I might just go down the kit track.

Once I choose a few I'll post them up and ask for some feedback.

Or if anyone knows of a good kit please let me know.

I think I can still solder, but for some reason, now that I'm off the morphine, I cant remember anything I learnt in the last 3 months.

drugs r bad m-kay

Cant I just do this? (I can hear another Nooooooo! coming) This was the kind of thing I was thinking about originally ...



This is just a flow chart.

The switches in the dotted line boxes are controlled by a chip.

The plan is that any number of capacitors (only 3 in this example) would be charged either in series or parallel depending on the supply voltage. Then the chip compares the voltage of the caps or source, and compared to the desired output voltage, it could then mix whatever capacitors it needed to achieve the desired output voltage. With the battery (right hand side of the circuit) disconnected, the capacitors would be charged. Then the source power (left hand side) would be disconnected, the number of capacitors needed would be calculated depending on desired voltage, then they would be opened to the battery. So it might charge all the capacitors in parallel, then discharge the required number into the battery in parallel (or any other combination).

The output could be controlled via a human dialling up whatever voltage was desired, or by code to create a 3 stage battery charger.

Man I wish I knew what I was doing.

Conceptually there is no problem with the circuit above.
Good page here:

http://en.wikipedia.org/wiki/Voltage_doubler
for that kinda stuff.
It would have to be low power, capacitors are generally very lossy as the current drawn increases, so would the ripple.

I was hoping to avoid loss.

What constitutes low power? I'm looking at my 230 watt solar panel, and perhaps some home made wind power added on.... say another 75 watts

Thanks for that link.

I've been looking for ages, but it's tricky when you dont know what these things are called.

If you have a circuit with a diode, I understand that there is a voltage drop across the diode.

If you are using circuits like a voltage doubler to get to the desired voltage, how much actual useful power is lost by the diode? is it proportional to the amount of voltage lost?

You'll never avoid it completely, there will always be some voltage drop across any cap, resistor, transistor or whatever.


As you're only wanting to charge a battery, it's probably not that important to have a ripple free output.
You can go simple and cheap, switched cap.
or (relatively) complex and more expensive with a dedicated DC-DC converter resulting in much better efficiency, around 90% I believe.

300 watts is quite a lot of electrons to be playing with, things can easily go bang.
As you know I'm a digital guy, this is on the edge of my skillset.
SV can probably give some better input, but I'd say you're really looking at a dedicated DC-DC converter, which as discussed earlier, you can get kits for.

Standard diodes, typically 0.7V
LP Schottky ~0.3V

Power is always proportional to voltage P=VI and V=IR

Hey BW how did you draw your "flow chart"?

Did you use a software program?

Looks very neat and professional.

Charge pumps are never used for high power applications. The biggest I have seen is a 30W one (just in an application note for a controller IC). To be honest Im not sure why exactly high power ones are not used, but I would guess it is because the components would be very expensive, far more expensive than just building an ordinary boost switch mode power supply. A rudimentary one would look like this (this is a design im using in something I am designing for a client)

The PWM_Vol comes from the microprocessor. The micro drives a PWM signal to vary the voltage on the power supply.
The orange wire PSU_V goes to the ADC of the micro and provide a voltage feedback to the micro so it can adjust the PWM signal to get the correct voltage. My application requires a variable voltage supply and this is the simplest way to do it. The voltage can vary from maybe 150v down to just a few volts. There are limitations to such a simple boost supply but they are acceptable to my design.



EDIT: forgot attachment

I found this.

https://www.circuitlab.com/

Its actually a simulator, but I dont know how it works, so I just used it as a diagram. It's also a bit limited as there are no things like lights, and a much better simulator is ktechlab. Ktechlab is free, does everything, and really rocks. But I'm struggling to get it to work on my system for some reason.

I use TinyCad.
http://sourceforge.net/apps/mediawiki/t ... le=TinyCAD
Built in libraries are a bit limited, but if you search around a bit, there are others libraries downloadable.
The built in component editor is also very easy to use, takes a few minutes to knock up a "strange" component.