Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Design for a Pure Sinewave Inverter, what is easiest and cheapest to implement?

Status
Not open for further replies.

David_

Advanced Member level 2
Joined
Dec 6, 2013
Messages
573
Helped
8
Reputation
16
Reaction score
8
Trophy points
1,308
Location
Sweden
Activity points
12,217
Hello.

I have just gotten into sinewave inverters because I want to build such a device to power certain labb equipment in certain situations(and I simply want to learn more about these sort of devices).

I have chosen to skip modified sinewave inverters all together and I understand that there are a few different ways to realize a pure sinewave inverter.
I just finished reading this PDF document in which the go over the design of a pure sinewave inverter(hens forth called a inverter) based on analog circuits. It is designed for creating 120V 60Hz and I need 230V 50Hz, in any case the designers in that document didn't really complete the design because it failed in certain aspects and they didn't have time to prove that there solutions will work but I'm sure they would if one was to follow there recommendations.

But that design completely lacks a circuit for stepping up a say 12V DC source to a 170V DC source which they needed to drive a H-bridge to output a 120V sinewave.

In any case I was thinking that it would be nice if one could enable more than one input DC voltage level, but I am getting ahead of my self.

I have also seen designs using microcontrollers and other fancier ICs, and the design I linked to is very simple but not really as good as I want and the more complex designs I have seen have been good enough but way more complicated than what I like to do.

If I asked you, "what do you think the most straight forward approach to designing a pure sinewave inverter to convert 12VDC into 230VAC 50Hz is?"

What would you answer?

Regards
 
Last edited:

If I asked you, "what do you think the most straight forward approach to designing a pure sinewave inverter to convert 12VDC into 230VAC 50Hz is?"

The most straight forward approach in my opinion:

1. Using a switch, convert 12V DC to 12V square pulses at high frequency;

2. Using a conventional transformer, convert the 12V square pulses to 300V pulses;

3. Using rectifiers, convert 300V pulses to 300V DC;

4. Using a switch, convert 300V DC to 50Hz pulses;

5. Using a bandpass filter, get the final output at 230V AC 50Hz

Unfortunately life is not so simple. You did not mention power needed because if you want any significant power, you need to have both feedback and regulators at every stage.

Getting a bandpass filter for 50Hz that can handle significant power levels is not trivial.

The "pdf document" you have referred to also contains many inaccuracies. Better to refer to "application notes"

"pure sine wave output" coming out of an inverter always will have some harmonics because they are also synthesized using a PWM approach which is usually very satisfactory.

By the way, please tell some instruments that need "pure sine wave" to function well.
 

I've been thinking about the power needed, and I don't have any particular value but when I search for pure sine wave inverter designs online I only find examples of really hefty designs with hundreds and thousands of watts. I don't need more than 460W and that is a conservative estimation.

The instruments I will use this for don't particularly need a inverter but I want to enable a isolated power solution, sure a isolation transformer is as simple as it can get in order to accomplish that but I also would like to be able to power instruments wherever I am.
But one of the main motivations for this is simply that I find that I am able to feel interested by this sort of design and as such want to learn more about it by doing a design, these devices are after all very necessary in some instances.

The pdf I linked to as well as application notes from TI has told me that pure sinewave inverters can power devices as a better than mains quality power source. But I guess that depends on the particular design and probably it's down to the output filtering.

So without saying that this can't be lowered if it would serve some good enough motivated purpose lets set the power level as 460W output power, maybe it's more straight forward to simple say 500W then.

Then I would like to ask:

What amounts of current does a 500W pure sinewave inverter demand from a 12V battery, if we pretend there are no losses in the system and that the inverter is used to deliver the maximum power, would it be as simple as 500W / 12V = 41,6A?
 

Most modern inverters are pretty efficient these days. So if you want 500W out you need to feed about 600W in. That will be just about 80% overall efficiency and perhaps you can do better than 80% but is does not harm being a bit conservative.

As you yourself has estimated, a 12V battery will deliver 600W/12V or about 50A. That is a lot of current. And your battery will run down quickly- depending on the AH capacity. And you will also need a battery charger.

Often it is more convenient to use a 24V battery or even 48V battery to reduce the current demand. You will understand when you will see the cabling needs.

By the way, AC power systems are rarely isolated (theoretically speaking) because there are always some capacitative or inductive coupling. Physical DC isolation is often good to prevent electrical shock but as the power levels increase, the effective AC isolation gets lower.
 

The step-up transformer is ideal if we want ease and simple construction. Use an H-bridge to apply true AC 50Hz square waves. Shape them into a sine by installing a series capacitor. Thus you reduce problematic spikes which tend to happen when switching inductors.

This is a basic schematic showing waveforms.



Notice the transistors turn fully On & Off, minimizing heat production.
Biasing depends on proper volt and current levels.

It will be necessary to experiment with Farad & Henry values. With lighter load, the system automatically reduces Amperes drawn from the power supply, and the output waveform becomes more squarish.
 

Getting a bandpass filter for 50Hz that can handle significant power levels is not trivial.

The "pdf document" you have referred to also contains many inaccuracies. Better to refer to "application notes"

Agree with both statements. The PDF paper is obviously a school paper on which the project, as long as it is discusses theoretical concepts correctly, does not require to have an actual, fully functional and reliable circuit.

To me what was particularly bothersome in their filter design is that it shows a pair of resonances reaching 60 dB peaks and they did not mention the effects on transient response, for instance.
 

What type of design do you want? Do you want microcontroller based inverter or you want discreet components based inverter also what are your specification.
 

You choose the most suitable design approach depending on the input and output requirements.
The most obvious are input voltage range, output power level, and nature of the output load.

The most difficult to specify is usually the type of output load. Is it constant and never varying, or are there massive occasional load surges, or extremes of power factor ?

Its no good building a 150 watt inverter to drive a refrigerator that requires 150 watts of running power, if the motor pulls a 2Kw surge of current every time it starts up.

Easiest and cheapest to implement might go bang the very first time you plug some really nasty load into it.
Or maybe the very same inverter will run for years trouble free powering a more friendly load of equal average power.

Making something that works is dead easy.
Making something that works reliably for a long time with a bit of occasional abuse is vastly more difficult.
 
  • Like
Reactions: David_

    David_

    Points: 2
    Helpful Answer Positive Rating
Okey maybe we should change the priorities to:
reliable and safe, and as long as those two things are kept in mind then cheap and easy is relevant.

I realize that simply "easy and cheapest" isn't an good description of what I want, the ease of implementation and the cost of the design is only relevant after the requirements that the design is reliable and safe have been fulfilled.

I am prone to wanting to use a XMEGA AVR microcontroller which I have a few of the A1U and A3U varieties, but the choice between a digital MCU based solution and a analog one is probably going to end up in the digital choice since one point of building this is to get experience with the details and if I am not mistaken a MCU based design gives greater and easier access to the possibility to adjust parameters such as switching frequency, duty cycle, dead-time and so on.

Also a microcontroller based design could maybe make the filtering easier to accomplish...?

As for the kinds of loads, I do not foresee any larger transient load situations, the most usual load will probably be a Rigol MSO Oscilloscope or a Arbitrary Function Generator.
 

... if I am not mistaken a MCU based design gives greater and easier access to the possibility to adjust parameters such as switching frequency, duty cycle, dead-time and so on...

The increase in complexity is only moderate but the gains are significant. In addition, the softwares can be modified as needed.
 

No, an MCU does not simplify anything.
A potentiometer is difficult to beat if you want cheap, simple and adjustable.
 

I have found some interesting sort of guides online and will see what I can make out of that information, I'll be back when I have some schematic to show and talk about.

As I have found one MCU based one and one based on UC3525 if I recall correctly I'll haven't decided yet which kind to go for, but I think I am leaning towards the analog solution.

Regards
 

The analog solution for power electronics is more often than not the optimum solution. Analog not only has infinite resolution, it has vastly higher speed. Both of which are critical for regulation and protection.

By all means a microcontroller might be used for setting the operating parameters and supervision of an analog control loop. But its really just bells and whistles and icing on the cake.

For a real cake, a very simple robust analog solution will win hand down for cost, simplicity, and robustness.
 

For a real cake, a very simple robust analog solution will win hand down for cost, simplicity, and robustness.

Plus, when something goes BANG!, with a microcontroller is far more difficult to determine whether it was the hardware or the software that caused the failure.
 

Yup.

When the system is about to violently die in tens of naonoseconds, its not time to start thinking about very carefully saving things on the stack, and holding a committee meeting about some interrupt.
 

... and holding a committee meeting about some interrupt.

You are being rather unkind.

Interrupts are the most misunderstood and misused concepts and handled well, can be quite powerful.

I feel sad when I see the engineer trying to do the impossible within the interrupt service routine.

By the way, programming is an art but so is the circuit design.

I agree that nothing can beat the simplicity of analog designs if and only if you know what you want.
 

Haha, its why I have stuck to analog design and mostly power engineering.

My software buddies are extremely capable and knowledgeable people in their chosen field, all I can do is stand back and marvel at their accomplishments.

But without the hardware to actually make it all go, they are as helpless as newborn babes.

When each of us pitches in, and together as a close knit team, we make some marvellous new product that is always a very satisfying outcome.

I have nothing but respect and admiration for the software wonks, but some things are best left to a bit of really fast and well thought out analog.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top