I have a 12V power supply and I need to obtain two poles with positive and negative voltage. The polarity inversion needs to be controlled electronically.

It does not matter what voltage I get, as long as the two poles are symmetrical (+V, -V) and that I can drive high current through them, several amps (currently 4-5).

A simple voltage divider won't do because the resistors involved will limit the current and my load is around 3 Ohm, so even a small resistance added will have impact on the circuit.

The only solution I have so far is an H-Bridge (which works since I don't need both polarities at once), but I am looking for something simpler.

Any ideas?

 

If you want simpler then there is this. But BOY would you arms get tired switching that back and forth. The H bridge sounds good. What makes you think there should be something 'simpler', (beside a half H bridge).

 

Haha awesome!

 

(BTW, how would a half H-bridge reverse polarity?)

 

What is the load?

I don't understand how it doesn't need + and - at the same time, unless it is a Motor.

 

You'd need a full H-bridge (two half-H bridges) in order to reverse the polarity.

You could also use a DPDT toggle switch, but you'd have to switch it manually.

You could also use a DPDT relay, and energize the coil 1/2 the time using a timer.

 

What is the load?

I don't understand how it doesn't need + and - at the same time, unless it is a Motor.

Electrolitic cell.

You'd need a full H-bridge (two half-H bridges) in order to reverse the polarity.

You could also use a DPDT toggle switch, but you'd have to switch it manually.

You could also use a DPDT relay, and energize the coil 1/2 the time using a timer.

Right, I have been working with an H-bridge (MOSFET) but I am not pleased with it. Relays are not an option due to high duty cycle. I am meaning to try solid state relays.

I am controlling the switching with a microcontroller, so the H-bridge needs a driving circuit, which increases the complexity.

While this prototype requires just 4-5 A, larger versions may go to 100A and up to 8000A. No I am not kidding. So I need a scalable solution.

 

Are you sure of the application and the magnitudes of current? The polarity reversal seems to serve no purpose - http://en.wikipedia.org/wiki/Electrolytic_cell

 

Are you sure of the application and the magnitudes of current? The polarity reversal seems to serve no purpose - http://en.wikipedia.org/wiki/Electrolytic_cell

Yes I am absolutely sure This is a specialized application.

 

I can make 3 guesses, and 2 don't count.

 

This is a relatively obscure research related application, having to do with oxidation of organics. What did you have in mind?

 

Organics oxidize in several manners. Electrochemical cells aren't the first thing to come to mind -

 

Is it cow crap? or human?

 

Is it cow crap? or human?

Haha, neither.

 

MOSFETs are quite efficient, but need good drivers to "snap" them on and off quickly.

If you want to pass a large amount of current, you will need very good heat-sinking. Copper is one of the best metals for making heat sinks from; aluminum is only about 57% as good. Water is 1.4 times as good as copper, but corrosion can be a big problem.

 

Is there a reason the purpose of the circuit you would like us to help you make is not disclosed?

 

MOSFETs are quite efficient, but need good drivers to "snap" them on and off quickly.

If you want to pass a large amount of current, you will need very good heat-sinking. Copper is one of the best metals for making heat sinks from; aluminum is only about 57% as good. Water is 1.4 times as good as copper, but corrosion can be a big problem.

Yeah I have considered water cooling, but it seems that if ultimately this goes in the H-Bridge direction the system would have to be massively parallel so there will be lots of area for heat sinks. We'll have to see when the issue comes around. At the 5A level it is not problematic.

One possible simplification to the H-Bridge is to use an integrated circuit. Infineon has a 40A continuous current rating for half-bridge devices, which include charge pumps for the N Channel MOSFETS. But I am wondering if that is a good idea since one can obtain much higher rated individual MOSFETS (100A). And since the H-Bridges will be parallel, maybe a single charge pump could be used to drive several of them from the Microcontroller.

 

Is there a reason the purpose of the circuit you would like us to help you make is not disclosed?

This is for research into the oxidation of individual organic compounds through 3 oxidation mechanisms.

I am just looking for alternatives to what I already have because many reactors of different sizes will be required and we want to save money on both the components and the electric bill.

 

If we know what your trying to do as a final goal, we can think ahead.

Often people come with help requests, and a week and 20 schematics later, they say, Will this nuclear reactor fit in an altoids tin and run off a coin cell?

That kind of stuff.

 

Haha, true. I didn't post more info initially because I didn't think it would ring any familiarity.