12 volts to 120 volts the easy way?

[BernieM]

Living in a remote desert location, I have to generate my own power from solar. Big loads require big inverters, and for me sine wave is a must, not modifieed sine. Higher voltage inverters allow more batteries in a string meaning more storrage capacity in that string and since multiple parallel strings don't charge well, a large capacity single battery or string is better than smaller parallel arrays.

Big sine wave inverters are expensive. So while considering the problem, I hit upon what at first glance seems to be a simple and effective solution. Namely, put sufficient batteries in series to reach the desired a/c voltage and then using a simple and inexpensive waveform generator, drive a large power transistor or series of them to provide the wattage capacity desired. It should be clean enough as far as power goes to be usable without much waveform conditioning I would think.

With 12 volt batteries, for example, 10 of them when fully charged will be at 126 volts and when drained still be at 108 volts, nearly 110 volts, though one wouldn't want to let them get that far drained. Probably cut out when they reach 11.2 volts or so, or 112 volts on a chain of 10. With my 24 volt inverter, the maximum number of 12 volt batteries I can put in a string is 2, with 6 volt batteries, 4. With this method I can use ten 12 volt batteries or 20 6 volt!

OK so what am I missing? I get a sneaking suspicion that someone is going to mention the down sidde that I'm overlooking. Ok so have at it! Is this a good solid solution or should it be filed in file 13?

 

[nsaspook]

Lets start with a very simplistic circuit.
First to generate a good sine wave directly from the batteries the total voltage will need to be 2* the peak voltage of the output AC waveform. You then need to isolate the DC average voltage from the AC voltage with a large capacitor to get that pure AC sine wave output. This means using a class A power circuit.

It might work but will be very inefficient unless you also need a space heater.

 

[x86]

Beginner engineering student here. Have you thought about current?

Wouldn't a 120V supply supply different current than the current from a 12 volt supply?

 

[phinds]

Beginner engineering student here. Have you thought about current?

Wouldn't a 120V supply supply different current than the current from a 12 volt supply?

Yes, but so what? What's your point? The actual current drawn depends on the voltage source and the load.

 

[BernieM]

Potential is at 120 volts already, ten 12 volt batteries in series, for example. Simply turning an scr on and off 120 times per second reversing polarity every other on cycle gives me a 60 hz square wave, which is what a lot of cheap inverters actually produce. Seems a simple capacitance circuit might shape the square wave into the sine. Or maybe someone makes a cheap square wave to sine wave converter. I mentioned the power transisttors or maybe power FETs as just a way to make the low power of my cheap waveform generator output the power as a sine wave. In other words use the low power sine wave generator to drive the base of the transistor. The load will determine the current drawn through the transistor or whatever.

 

[phinds]

Potential is at 120 volts already, ten 12 volt batteries in series, for example. Simply turning an scr on and off 120 times per second reversing polarity every other on cycle gives me a 60 hz square wave, which is what a lot of cheap inverters actually produce. Seems a simple capacitance circuit might shape the square wave into the sine. Or maybe someone makes a cheap square wave to sine wave converter. I mentioned the power transisttors or maybe power FETs as just a way to make the low power of my cheap waveform generator output the power as a sine wave. In other words use the low power sine wave generator to drive the base of the transistor. The load will determine the current drawn through the transistor or whatever.

This method will not produce what is normally mean by "120V" because that is an RMS value whereas you are talking about peak value. Reread nsaspook's post #2

 

[BernieM]

Sorry I was thinking about 240 when I was writing 120. 120 is 60 volts peak, 120 volts peak to peak. But that's fine then 240 volts a/c from the 10 batteries. I guess I am most interested if there is a relatively simple way to do this that isn't too ineffficient. No I don't need a space heater in the summer in Arizona.

 

[phinds]

Sorry I was thinking about 240 when I was writing 120. 120 is 60 volts peak, 120 volts peak to peak. But that's fine then 240 volts a/c from the 10 batteries. I guess I am most interested if there is a relatively simple way to do this that isn't too ineffficient. No I don't need a space heater in the summer in Arizona.

You need to check your math.

 

[nsaspook]

I guess I am most interested if there is a relatively simple way to do this that isn't too ineffficient. No I don't need a space heater in the summer in Arizona.

There really is no efficient way to directly generate the sine wave directly from battery voltage. Using a push-pull transformer solves the 2* peak voltage problem but if you stick with direct analog generation for the pure sine wave you will be stuck at <50% efficiency. A efficient device is a complex device as this project demonstrates.

https://www.wpi.edu/Pubs/E-project/Available/E-project-042507-092653/unrestricted/MQP_D_1_2.pdf

 

[ibuildstuff]

Simple way to get around the RMS vs. peak voltage problem: add 4 more batteries. That brings you to a total of 168 volts, which is really close to 170 volts, which is the peak voltage out of your standard 120 volt wall socket. With that out of the way, the only other problem I see with this is that you're assuming the transistors will behave in a completely linear manner in response to switching voltage. I have no idea if that's true, but I doubt it. You better check that out before you blow some appliance up with a malformed waveform.

 

[jim hardy]

Big loads require big inverters, and for me sine wave is a must, not modifieed sine.

i don't know of any household appliances that are picky about sinewave purity.
What is your thinking?
I've seen a square wave inverter with just a third harmonic filter, a brute force L-C series resonant circuit tuned to 180 hz , but the waveform is still rough...Mr Spook's link is a most interesting approach . I never saw that triangle wave - comparator trick before in that application.

Another method is here...
http://www.tinaja.com/glib/msinexec.pdf
sort of a a twist on Fourier
synthesize a sinewave as a sum of square wave harmonics

old jim

 

[Jobrag]

How big are your loads and why do you need such a clean voltage, (your not Blofeld setting up a secret base are you). Have you considered an MG setup?

 

[sophiecentaur]

for me sine wave is a must, not modifieed sine.

This is a statement that I have to question. What is so special about a 'pure' sine wave? If you consider that 'the Mains' would do your your purpose then I suggest you look at the actual waveform the supply company gets to your house. It is only, ever, a rough approximation to a sine wave and it varies as other loads are added and removed. So, if you reconsidered your requirement, life could be a lot easier.
Alternatively, you could perhaps modify the equipment you are supplying with the 120V AC and make it less fussy. That could be a much easier way through this problem of yours.

 

[Svein]

http://en.wikipedia.org/wiki/Power_inverter

 

[sophiecentaur]

http://en.wikipedia.org/wiki/Power_inverter

Yo which particular bit of that wide ranging reference were you wanting to draw attention?

 

[rollingstein]

i don't know of any household appliances that are picky about sinewave purity.

Do motors run hotter? I thought they did. Not sure.

 

[Svein]

Yo which particular bit of that wide ranging reference were you wanting to draw attention?

I just wanted to draw attention to it. I am repeatedly getting warnings for being too close to the answers...

 

[Chacabucogod]

What about using a Boost Converter or a cascade of these to get your voltage really high? Then build the inverter.

 

[sophiecentaur]

I just wanted to draw attention to it. I am repeatedly getting warnings for being too close to the answers...

A few words to help direct the reader to the appropriate section can help without 'giving away' anything. But, in any case, it was not a homework type question so it deserves loads of helpful answers.

Do motors run hotter? I thought they did. Not sure.

Possibly, but that's no reason to have to produce a 'perfect' sine wave. (How perfect does that mean, in any case?)
I asked the question about required quality of sine wave because it is the first thing an Engineer should do - i.e. define the specific problem rather than come in half way through with an answer that takes ones fancy.

 

[sophiecentaur]

What about using a Boost Converter or a cascade of these to get your voltage really high? Then build the inverter.

What would be the point of changing the input DC voltage when a transformer will do the job (probably) cheaper and more efficiently, once you have your AC?

 

[rollingstein]

Possibly, but that's no reason to have to produce a 'perfect' sine wave. (How perfect does that mean, in any case?)

Agreed.

I faintly remember a vendor trying to sell us some system to improve the harmonic quality. Is that similar to going closer to the "perfect" sine wave?

As far as I remember the argument was that even at the same power factor rotating machinery would operate better ( cooler?) with a better harmonic structure.

Could have been snake oil.

 

[sophiecentaur]

Could have been snake oil.

Oil in a bottle with no numbers on the side is probably snake oil, for an Engineer. The Power factor thing may or may not be a real consideration here.
The OP may not be referring to motors and efficiency. It would be nice to know the actual numbers involved in his particular requirement. I suspect that the actual requirement hasn't been specified quantitatively and the notion of a pure sine wave may be a lot less important here than he assumes.

 

[Svein]

http://www.instructables.com/id/250-to-5000-watts-PWM-DCAC-220V-Power-Inverter/

 

[sophiecentaur]

We lost the OP somewhere along the way. We are now anchorless and the conversation has no direction to go in. There is no solution to this unspecified problem.

 

[ibuildstuff]

We lost the OP somewhere along the way. We are now anchorless and the conversation has no direction to go in. There is no solution to this unspecified problem.

Seeing as how no one really told him what was wrong with his idea, instead opting to ask him the same questions over and over about sine wave purity, I'd say he's busy building a setup according to his original idea. Seriously, what else did you guys need to know about the problem that he didn't say in his original post?

 

[sophiecentaur]

Seeing as how no one really told him what was wrong with his idea, instead opting to ask him the same questions over and over about sine wave purity, I'd say he's busy building a setup according to his original idea. Seriously, what else did you guys need to know about the problem that he didn't say in his original post?

We need to know his actual requirement. What would be so special about his equipment that would need a clean (unspecified how clean) sine wave supply? That question surely needs to be answered before any solution can be reached. There are various levels of power, purity and cost and why not let us know what's really required? There are some fundamental issues here. Do you not appreciate that?

 

[jim hardy]

Seeing as how no one really told him what was wrong with his idea, instead opting to ask him the same questions over and over about sine wave purity, I'd say he's busy building a setup according to his original idea. Seriously, what else did you guys need to know about the problem that he didn't say in his original post?

How much power he wanted, for starters. His linear approach is practical for maybe fifty watts. I've used it to make 3 phase 120VAC in test equipment.

I for one took this post as evidence of confusion:

Sorry I was thinking about 240 when I was writing 120. 120 is 60 volts peak, 120 volts peak to peak. But that's fine then 240 volts a/c from the 10 batteries. I guess I am most interested if there is a relatively simple way to do this that isn't too ineffficient. No I don't need a space heater in the summer in Arizona.

With a single 120 volt supply and simple linear amplifiers as he proposed, you can only make 84 volts RMS at maybe 30% efficiency.

Since he doesn't appear familiar with sinewaves i asked why he felt he needed one.

 

[dlgoff]

Seeing as how no one really told him what was wrong with his idea, ...

Well BernieM did say,

I get a sneaking suspicion that someone is going to mention the down sidde that I'm overlooking. Ok so have at it!

bold by me