Sinusoïdal Power Inverter, complicated ?

Status
Not open for further replies.

MarcelM0

Newbie level 6
Joined
May 17, 2017
Messages
11
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
123
Hi,

I'm thinking to build a sinusoidal power inverter (around 3000w) for my home. I am 100% on solar energy.

The principle seems simple ;

First stage to produce a sinus wave form (60Hz) ==>> Amplifier to get 12V (and much amps) ==> Transformer 12V/110V

My question : Is it really so simple ? Why are the comments on internet so weird about building it. Why are the prices so high ?

I'd like to have your opinion before i start.
Thank you for your comments.

Marcel
 

Hi,

What is your input voltage?
What about cos(phi)?
What efficiency do you expect? What no load power do you expect?

12V, 3000W means 250A average current. But what peak current? I'd expect at least twice the average current, three times?
What amplifier do you have, that is able to deliver more than 500 Amps?

I have some experience with electronics... I don't call it "simple".

Klaus
 

Thanks, it was a simple question on the principle, not on details.

All I have for now its a 12V input (my solar batteries)... and a circuit which gave me 2 clean stable sinus signal (60hz) - Produced by an OP-Amp and a Twin-T filter. Why 2 ? They are exactly in phase opposition (180 deg) - with the idea to use them in a push-pull configuration in the future.

At the oscilloscope ; I get a sinus signal around 2.5V (20 mA max). , very steady and simulate perfectly an 110v outlet signal.

Now the next step should be to bring it at 12V - as input for my transformer (12v/110v - +- 30 amps)

How can I do that ?

Transistors yes... In Darlington ? IGBT ?

Thanks for any help.
 

Hi,

Thanks, it was a simple question on the principle, not on details
Short answer:
* the principle is simple
* the details are complicated, especially when you want to achieve some "quality", like efficiency, ruggedness, clean sine wave.

The higher the quality, the higher the effort.

Ther are many different solutions, many of them discussed here in the forum.
Read through them to find out what you need and how much effort you want to spend.

How can I do that ?
Transistors yes... In Darlington ? IGBT ?
Simple answer: yes. Darlington, Mosfet or IGBT is possible.
More detailed: Now you need to decide your quality, decide your specifications.
These parameters not only determine the transistor, but the whole amplifier technology, the circuit.

Klaus
 

Hello,
Here is an LTspice simulaiton of a seceral kW grid tied inverter.
it is in the fre ltspice simualtion from linear.com
just converter .txt to .asc and hit run
 

Attachments

  • GTI _Open Loop_2.txt
    15.8 KB · Views: 89
A pure sinewave inverter does not waste thousands of Watts making heat caused by a linear amplifier, instead it uses Pulse-Width-Modulation (PWM) so that the output devices switch completely on and completely off at a high frequency then their PWM modulation produces the sinewave with a low amount of wasted power making heat. Your huge low frequency transformer can be replaced with a smaller high frequency one. Voltage regulation must be used to keep the voltage constant.
 

As Klaus mentions: the concept is simple and straightforward.

But the actual details of the implementation are very significant. Starting with the PWB layout and the thermal management.
 

See comment #2: it is really messy to handle 250 or 500A currents; you need bus bars!

As your source is the solar panel, it may be better to connect in series and get 48V or even 96V for the input of the inverter. For the same power, you will need to handle less current.

By the way the prices are high because the demand is less. Once solar panels become more common, the prices will fall. But if you want to tie it to the grid, there will additional demands and I do not see the price coming down anytime soon.
 

I'm not sure if post #1 means 100% in favor of solar power or 100% of the power is derived by PV.

They may be looking for a PV to 110V system rather than grid tied.

Brian.
 

If you are just interested in a 3KW inverter from 12V (or higher) to 110V 50/60Hz, it may not be difficult at all. But you will need additional features:

1. If suddenly some clouds appear in the sky, your power output will drop and you will need to shut down the inverter. That may happen with annoying frequency if your load is close to the max.

2. Are you planning some energy storage device (battery, e.g.)?

3. What kind of load you are planning to use?
 

Thanks for your great comments.

c_mitra ; I do have a set of batteries (as buffer). My actual loads are "low" amps devices - Leds (for my light, small kitchen equipments, TV, my PC is the most glutton.
For now - and since 2 years - I'm using simple "car inverter" labeled as 1000W (more 400 w at most).
But my goal is to be able ... one day... to use an small electrical heater in the kitchen - to cook. Or maybe a small router for wood work.

Question : I'm not clear at all , how a PWM can finally produce a Sinusoidal wave ?
Because at the output of my transformer I will get those PWM variations (Right ?)


audioguru : "Your huge low frequency transformer can be replaced with a smaller high frequency one. Voltage regulation must be used to keep the voltage constant."

Very interesting. But how could I get a 60Hz out of this kind of transformer ?.

For info : Solar option is the only one i've got. No standard 110v accessible where I live. That's why i am trying hard to solve my problem (at a reasonable cost)
 

c_mitra ; I do have a set of batteries (as buffer). My actual loads are "low" amps devices - Leds (for my light, small kitchen equipments, TV, my PC is the most glutton.
It does not matter as long as the total load is less than 3kW. I guess as a domestic user you do not have heavy inductive loads. Keep some headroom and try to keep the total load around 2.5kW and a decent phase factor.

Question : I'm not clear at all , how a PWM can finally produce a Sinusoidal wave ?
Because at the output of my transformer I will get those PWM variations (Right ?)

Just like a radio (or a TV) has a carrier frequency (that can be rather high) which is modulated to carry the audio signal. Finally we discard the carrier because we are just interested in the audio. In this case too, we use a high frequency (just to convert the low voltage to high) that will be modulated with a 50/60Hz sine function so that at the end we can forget about the high frequency and just get back our sine wave. It is that simple.

My original suggestion still holds: seriously consider using a 48V system (you will just need to reconnect the batteries and the solar panels) because anyway you want the output at 60Hz 110V, right?

To vaguely understand how this process works, draw a sine wave on paper. Shift the graph 1 unit up so that all the points are on positive side.

Now draw a high frequency square wave (for this purpose make it about 10 times but in reality it can be much much more). Superimpose on the same graph.

Next multiply point wise; where the positive pulse is 1, keep the sine wave and where the pulse is at 0, delete the sine wave.

Now draw another set of square waves (same frequency and phase) but this will have const amplitude but the width will be proportional to the sine functions.

This will be the PWM signal that will be used to excite the 3kW transformer. You will need only a small transformer because the frequency will be around 100kHz range (just saying).

The output too will look like the PWM- const amplitude but variable width. How to recover the modulation back?
 
A few hours ago I tried to post a graph of PWM but this site was down. Here it is:
 

Attachments

  • PWM.png
    47.7 KB · Views: 101
A few hours ago I tried to post a graph of PWM but this site was down. Here it is:

Whereas it is possible to use negative going pulses (as appears from the graph), in practice it only makes life miserable. We usually stick to positive potentials and all the pulses are positive going. That is why the H-bridge has become rather popular because we can get currents both ways (positive and negative) through the load.
 
Here it is with all pulses positive going:
 

Attachments

  • PWM 2.png
    49.6 KB · Views: 98
After your comments I saw the urge to take a brake and get a serious course in power electronics.
I'm finishing a first part (DR Ferdowsi's Videos with his reference book ("Power Electronics by Daniel W.Hart")
Great course and Great Book. Unfortunately it does not cover the DC/AC multilevel inverters (just touch slightly the subject).
And it was my whole point :-/

Anyway it was a great journey in the fascinating P-E world.

Can you recommend me videos courses or books to go further and finally touch a real AC/DC converter explanation ?

Thank you again.
 

A modern inverter will have a microprocessor to control the output (PWM) and also sense the load (current sense) and the input voltage (PV devices are basically voltage sources but their current delivery will depend on the ambient light).

It will also help implement several safety features. It will also drive the indicator lights (they make life little better). Plus some fancy stuff...

I also suggest to get some basic ideas on microprocessors and programming. As with all kinds of technology, the basic idea is simple but once you go into details, the problems come into the picture...
 

"
I also suggest to get some basic ideas on microprocessors and programming. As with all kinds of technology, the basic idea is simple but once you go into details, the problems come into the picture...
" yes it is :-(

As far as i am concerned , Small signal Electronics and microprocessors/programming are +/-ok (as my profession is software engineer).
But P-E. is a kind of new land for me.

As you point it out (and as we say in german too) "the devil is hidden in the details".

My question was : Do you know a good set of videos or/and books for advanced P-E. About DC/AC converters. ?
(More specifically for Solar installations)
 

Status
Not open for further replies.

Similar threads

Cookies are required to use this site. You must accept them to continue using the site. Learn more…