Grid Tie Inverter Schematic

[tcmtech]

Back on the computer again.
Got a good question for any of you out there doing schematics and computer circuit drafting. what are good programs to use?
I do most of my schematics the old fassion way. free hand drafting.. yea the old timers way but hey, I love the feel of a good drafting pencel and the poratbility of my hands. Never had them crash and loose anything I was working on either. Never got that doodle bug out of my system!
Most of what I do Cad wise is Autocad based. I like it but its not the fastest or most efficient for circuit diagraming. plus it can glitch and refuse to convert a file over to a common picture type file anyone can open.

The reason I ask this is I want my DIY GTI manual to be user friendly and easy to understand. I will need some help and some positive critqueing done.
I cant share my schematics efficiently without it! Hope you all will help!

The GTI's I am going to cover in it are not theoretical but ones that I already built, debuged, tested and am actualy using. I actualy have to disasemble them and rebuild them to show the actual components and the construction prosses. Plus I will make them more profesional looking too. No farmer Fred welder/battery charger Frankenstien monsters!
The big 15 kw one is not finished yet. software is done but the actual unit still needs filters and full assembly and a few hard runs to debug and possibly redesign if the calculations are to far off from reality.
In return for your help I will give some of my control circuits and GTI system schematics up to everyone here to use or at least the ones realative to what real life problems someone may be having. No point in giving you the PLC program for the 15kw controler if you cant get a 150 watt basic unit to work!
Hope you will help! I will appreciate it if you do!
Do a good job and this web site and some of you WILL get credit on my CD when its done!

 

[mvocray]

Hi! I'm currently using Catena's Simetrix for building my circuits. It's a free software as long as you don't exceed a certain number of nodes or components.

Could you please continue the lecture on synchronization? Thanks.

 

[mvocray]

Hi again,

I have a few things I'd hope you could clarify re current source inverters vs voltage source inverters. I'm assuming an H-bridge configuration in all cases. Sorry it's quite off the main topic since I'm quite a newb in this.

1. I understand that DC current would be fed in current source inverters, not voltage sources, i.e., voltage sources should first be transformed to fixed current sources right?

2. If instead I use the H-bridge to switch and form a voltage source whose fundamental is greater than the line (with the same phase as the line) and connect it to the line using a impedance (low valued resistor or a low valued inductor perhaps to limit the current amplitude), would that circuit cause troubles?

To sum it up: I'm thinking of getting away with making a grid-tie inverter using an H-bridge that does an in-phase modified sinusoid (PWM) with fundamental's amplitude greater than the line. And connecting this modified sine wave to the line in phase with a small filter inductor. DO you think there will be problems with this setup? Thanks

 

[umer27]

@ Jules_Theone

Hay let me know, how reserve engineering that 150 W Inverter works out, if you have any details pls let me know.. I have one of those too.. All I can tell is, they use a high frequency transformer, to boost it to 350VDC, then use another TL494 for making the mod sine wave..

 

[Jules_Theone]

Yeah. I'm planning to make a DC-DC step-up circuit using either a larger tranformer or several 150w high frequency transformer with their outputs paralleled to get higher power. Eventually i want to make a circuit than can sense how much power is available and increase current drawn from the 12v source accordingly. I have a small solar array, small wind turbine and human powered exercise bike i would like to connect up.

I think the way to increase power flow into the grid is to make the output waveform speed up ever so slightly so the inverter is always trying to lead the grid and therefore current flows onto the grid. Adjusting how much the lead is should adjust the current flow. Obviously, if the circuit lagged the grid it would draw current.

 

[tcmtech]

Super Concepts guys!
Heres a real world approch. dont go constant anything! Thats how you can get nuts with the power and not burn out or get massive harmonics!

I will try to get some schematics made up and post them soon, then everyone will have a visual of what I am trying to discribe.
As you may have seen I am giving out design hints for an all analog GTI setup. Basic but reliable. Its a very solid foundation to build the more compicated digital control systems on.
Now, are we all on the same page with the H bridge with a blocking diode to let the power in from the DC souce but not flow back to it, and the small capacitor to trim the switching spikes?

* Switching components should be capable of handling 4x your calculated max amps and 2x your maximum peak voltages at minimum. *

Step one:
For now lets stay with the simple two transformer concept. its an easy way to understand the voltage to phase relationship. The control transformer can be small like one from a power pack you plug into the wall to run a cell charger
(2-5VA). In reality I have often used four transformers to run the H bridge. One for each IGBT. (I am a IGBT fan myself they are voltage controled like a fet but have a much more forgiving and rugged current and voltage handling ability, more like a big transistor.)
The bottom to IGBT's are dumping thier emitters to the common so they can be ran off a center tap contol transformer. Center tap to common, outputs to the IGBT's. A resistor and volt snubber is still recommended on the gate drives! Since we are at 50 or 60 hz a 1k resistor with a pair of back to back 12 - 15v zeners across the Gate emiter connections works well. The control transformer should be 10 to 12 volt ac output per side. (20 - 24v CT) or 4 to 5 times the IGBT's lower turn on limit. My IGBT's are in the 2.5v area so the 10 to 12vac gives me about the right deadband and turn on turn off ramp rates.
The upper two IGBT's need the control transformer windings isolated from everything. Thats why I often use two independent control transformers here. If you can find a quad wound transformer that has all four outputs electricaly isoated, super! it can run all four IGBT's at once. OR a pair of dual wound, you get the idea!
Dont over look the all important phasing of when each IGBT has to turn on! Real obvious but how many times will someone connect one or more phase reversed and short out the whole thing.
I bet you all will do it at least once. I have!
If you got the concept of the circuit layout you will see the control transformers have the primarys tied directly to each other. But not to the primary of the power transformer just yet. Make sure you get the control ones phased properly to each other! The phasing to the power transformer will then only be right or wrong! 50/50 chance at first hookup!

Step two:
power up the control transformers! but have no input DC. And dont hook up the power transformer yet!
Ready? Start by applying a limited voltage and current to the DC input. This will keep you from burning anything out. If your phasing is correct you will hear the power transformer buzz and be able to get voltage readings on the output side. IF you did the H bridge phasing right!
If its working the low limit of the IGBT gate triggering level will automaticaly be giving you a safe Zero cross dead band and compensating for that slight phase delay in the control transformer. plus being a sine wave the Gates are being ramped up and let down smoothly. you wont be getting that super sharp turn on and turn off edges of a true square wave. They are rounding the waveform just enough to keep the harmonics and switching noise down to the level that the filter cap can easily clean up.

Step three:
Assumming you have not smoked anything try running at the voltage range you plan to work with. Take out the current limiter and crank up the input voltage.
You should have a noticeable hum with that slightly harsh undertone. thats because the power transformer is being ran with a slightly rounded square wave input.
Check your output voltage and connect one lead of the power transformer to the power line common. Check your voltage between the other lead and the power line itself.
If you get a big number you are probibly 180 degrees out of phase on the output. reverse your output leads and see what you get.
You will still see a voltage difference, but much smaller. It will never be exactly the same. You are reading a sine vs square wave with a slight phase lag.

Step four:
turn off the control side power and power up the power transformer.
Scarry part here. You may want to use a variable voltage source to raise the power transformers voltage slowly. If you get fully powered up without any problems turn on the control power.
WOW! nothing happened. The IGBT's are not switching anything. No DC input yet, the only thing they have to switch is that small filter caps charge. If you have the right size capacitor the circuit wont even have enough charge to make a noticeable sound durring each part of the cycles charge and dump points.

Step Five:
Start bringing up the input DC voltage, the power transformer should start to get that raspy buzz to it agian as the input voltage climbs and your DC source should show a steady climb in the amps draw. Much higher than the ealier test you did without the mains connected.
Congradulations! if you chose the right ratio of power transformer to DC source voltage you are now feeding power back to the grid!
Variable input of volts and amps! how sweet is this!
Unless you were way off on your transformer choice or you burned up all your stuff because of incorrect phase testing, insuficient current capacity or to low of working voltage limit on your IGBT's!

And yes you still need line filters and some power factor tweeking to fully clean up the output.

I will cover that later along with some basic system controls to automate the on/off process.
This should give you something to play with for a while. have fun!

 

[tcmtech]

Dumpster Diving Tip!

Ever want to pull IC's off old circuit boards quick and cheap without destroying the pins or overheating them?
Canned Air!
Stand the circuit board on edge and heat the solder side joint with your soldering gun or pencel. Shoot the hot solder out from the IC side with the canned air! That little straw on the can focuses the blast right down the IC socket hole.
Plus canned air is usualy made up of R134a refrigerant. Keeps the IC cooler while its removing the hot solder!
Make sure your hand is not inline with the solder when it blasts out the little hole. And you may want to make a back catcher out of aluminum foil so your not blasting hot solder fragments all over your stuff!

 

[fernando_g]

Thanks, tcmtech, for your valuable method to remove multi-pin devices from a board.
Tried that a few minutes ago. Works GREAT!
Thanks again!

 

[mneary]

Most of the "canned air" I've bought recently is highly flammable, so don't be tempted to use a torch on the solder side.

 

[Nigel Goodwin]

Most of the "canned air" I've bought recently is highly flammable, so don't be tempted to use a torch on the solder side.

Almost all aerosols are now - since they took out CFC's they went back to propane and such

 

[tcmtech]

Absolutly right on the potential flammability!
Never looked at my cans as of late. One says flammable, and one says potentialy flammible in the fine print. The others dont say anything about flamibility. But they are the R134a based type.
One Difluoroethane based can says "This product can be ignited under certain circumstances" but is not directly labeled as flammable.
The other Diflouroethane says flammable on the back side about half way down in the bigger fine print.

I still could not get it to light up off my 230 watt buzz gun even with the end bright red!
But the butane powered pencel can make a flame thrower with a little practice!

So I wonder how many times I just missed blowing myself up over the years and making another addition to my F ups file!

 

[mneary]

As you noticed, the Diflouroethane requires a flame to start it and it goes out as soon as the ignition source is taken away - at least at room temperature.

So, don't use a flame to desolder!

Safety glasses are a must for desoldering parts from boards. Mask over mouth also vital. Everything you want to keep functional and free of permanent scars should be at a safe distance; hot solder sticks!

Don't wear artificial fibers when soldering; it melts and bonds to your skin. Natural fibers can be separated from you as soon as you feel the heat and possibly before the damage is permanent.

 

[Nigel Goodwin]

Don't wear artificial fibers when soldering; it melts and bonds to your skin.

Oh yes - blob of solder falling on nylon sock - OUCH!!!!!.

 

[bablu]

hiii i m new will u help me in my traffic controller plz

 

[tcmtech]

Dumpster Diving Tip!

Need a good low ohms resitor with some wattage capabilities?
Try the metal spiral out of a notebook! They are chrome plated steel on most or some may even be a low grade chromiumsteel alloy.
Once shaped into place it can run red hot all day without burning up. but flex it to much and the outer covering can flake off!
Resistances are a bit varied though. typical 8.5" x 11" 100 pager is around .5 to 1 ohm.
Stop by your arts and crafts section and pick up some large glass or ceramic beads and you have a homemade low volt heater/ dummy load to dump your power too!

Mount it in an old metal can and use the fan from a old computer power supply and you have it! (12-24v 150 - 600w load capable with fan!)

**** Apply common sense as needed.****

Stupidity is self eliminating! Either you learn from it or it builds up to a fatal level and kills you! either way, some stupid just eliminated itself!

 

[tcmtech]

Mathematical Monkey Poo

Math is everywhere, we all use it everyday it’s a part of our lives even if we don’t realize it or even think about it.
So is feces, it’s a part of life! Everyone and everything makes it. Some is good for fertilizer and can be used raw to makes things grow and develop better.
So are some forms of mathematics. Practical, usable and great for making ideas grow and develop.
Then there are the other forms of feces that have limited use or no real use unless processed or modified to work in a particular way, same with mathematics too.
Monkeys are known for throwing feces at each other and at anyone or anything that upsets or gets them defensive!
Sometimes types of mathematical formulas get thrown around more defensively than constructively! Hence the term “Mathematical Monkey Poo”
Often thrown around with no regard to weather it’s being used appropriately or constructively and most often it has been triggered by someone getting defensive or upset!
And just like the angry monkey the mathematician is trying to squeeze out the biggest pile he can make, even if the situation does not truly warrant it!
So are you spreading fertilizer to help make things grow better or are you just throwing Mathematical Monkey Poo?

 

[blueroomelectronics]

hiii i m new will u help me in my traffic controller plz

Not even close to a grid tie inverter.

 

[be80be]

Inverters work by taking the 12 or 24 volt DC voltage from the source, such as solar panels or micro hydroelectric generators and 'chopping' by turning it on and off at grid supply frequency (e.g. 60 Hz) using a local oscillator and a power transistor. This chopped DC signal is then filtered to make it into a sine wave (removing the upper 3,5,7 harmonics that make up the square wave and then applying it to a transformer to up the voltage to 120 or 240 to supply the needs of load.

A grid tie inverter does the same but has two key differences. Firstly the frequency has to be matched in phase to the grid. This means the local oscillator has to be in sync with the grid. Secondly the voltage of the inverter output needs to be variable to allow it to be slightly higher than the grid voltage to enabling current to flow out to the grid. This is done by sensing current flow and raising the voltage on the output (or duty cycle of the transformer input) until the current flow results in the resulting output power matching the input power from the DC supply.

This don't look like it would be to efficient for low power output. Low being any thing less then 10Kw

Transformer designers cannot change I, or the current portion of the I2R losses

Most of what I have seen has a transformer you would have design one that is solid state to have a good grid tie inverter. They make them not cheep. But after reading all of this post I got me some solar cels and some thing to play with till my head hurts.

 

[tcmtech]

I have seen similar statements made elseware about GTI setups just like what be80be quoted.
What people often dont look at is you are not trying to make money off the utillity grid selling back to them. You are only trying to save yourself from spending more money buying energy that you could otherwise be harvesting yourself.
Every KW I make myself is a bout 9 cents the power company does not get! I made $0 but did not spend any either! Think avoided cost, not actual profit!
The big 15Kw I am working on should produce all the power I need year round. Total profit per year? $0.00, operatational cost probibly $400 annualy (line connection fee and misc wind generator maint.) Avioded power purchase cost about $1400 a year and rising.
In theory that is.
Payback time to break even on the system install, around 3 to 4 years give or take.
Knowlege learned, priceless! Feeling of acomplishment, priceless! Thumbing my nose at the corporate utilities, Priceless!

That is what low scale alternative energy is about, is it not?
Its like having a garden. Sure there is some personal effort involved but you dont have to go to the store and spend $300 on veggies if your garden is good. Plus you did it yourself! That makes everything you get from your own efforts taste that much better!

Those big corporate level output only numbers are what bugged me to the point that I built my first one years ago.
The smallest one I have made is a 30 watt I built about 3 years ago to work on a solar panel. Once it was set up it turned on at about 3 watts feedback while drawing about 9 watts. It topped out at 30 watts cont with a 45 w peak feedback before it would overheat. Top efficiency was about 80% at 30 watts returning give or take.
The whole thing fit inside an old computer power supply case!
I actualy had to rewind the current windings on an old spinner wattmetter (I took the one turn out and replaced it with 50 of much finner wire. 200 amp is now a 4 amp!) to get it to read fine enough to see it feedback at 3 watts!
Sure I will die of old age before it makes enough to cut my electric bill down even once, but its still cutting even if at the micro level!
Actualy its not any more. I gave it to a friend. hail distroyed the panel. He hooked the GTI up to a truck battery and killed it too! Such is life.

I will try not to hurt you head! I did that enough myself figuring out how to do this years ago.

Anyone out there got a real world GTI they want to build? Post your specs!
Power source volts and amps capabilities, line voltage and frequency, price range.

I would love the challange of helping build a real life application one.
After all is that not what this GTI group is about?
Right now I feel like I am talking to myself about the basics. I have real world knowledge I want to share but I need an actual system to put numbers to.

 

[be80be]

Here a little some thing this is not the circuit I'm going to use but you can get the idea
Use pwm to switch a hi output h-bridge step that up to your mains voltage.
with the pwm you can bring it In phase now need to just boost the voltage a little so it can feed.