Grid Tie Inverter Schematic 2.0

[plodhi2002]

Thanks Bill
Fendel when you try the schematic do let us know here. I wonder if anyone of the technical experts here can compare this with the transformer type inverters (GTI) given and talked about in some of the discussions here in these forums. I tried to analyse the schematic by comparing the switching devices like transistors/fets being substituted with the Triacs. These are getting switched on during the +ive and -ive half cycles when AC is present and that is through the capacitor and the circuit being completed through the battery or solar panel. I think that the capacitor substitutes for the transformer, the big difference is, it is on the DC side of the bridge and 2nd big difference is the +ive of Battery/solarpanel is going to -ive of bridge circuit formed by the four diode connected Triacs. Now in conventional transformer type of inverter the voltage is stepped up by using the windings inductively coupled and having a suitable turns ratio and then the windings with the switching devices are switched on/off alternately. In the case of GTI the switching devices would be triggered by a small transformer in parallel with the AC mains. Here I believe the capacitor is doing that?? Any ideas on this.

 

[Fendel]

Great work Nikos.

I make the Neo Aerodynamic schematic with 4 triacs BT139, 4 diodes IN4007 to gate, 4 diodes IN5408 in place of the fuses and one 6 uf 450V capacitor, and begining tests with 6 VAC and +/- 3 to 12 VDC...
The tests results was pessimistic, but still the components are not burned...
So I began study the BT139 and BTA25 dataschets, and I understand that the gate current must be betwen 1 mA and 35 mA.
Also I have to put some resistors on the gate.
Can you help me to aprimorate that Neo Aerodynamic schematic, to 220 VAC, because I first will make a cheap, robust and simple square GTI.
Thanks.

Why square? Because all grids are realy infinite comparated to the litle generators.
For 3 decades I am tring to implant the ENERNET:
**broken link removed** a resume in english and
**broken link removed** a complete revision in portuguease.

 

[Fendel]

View attachment 67926
that´s the "square GTI"!

 

[plodhi2002]

Yes Nikos has also done some other great GTIs using transformers with 2N3055s and they worked ok but we used 30V AC for just to experiment. Like the other more technically qualified guys say on this and other threads here its risky to do it with the higher voltage AC. This new one he gives seems a bit more difficult to understand,seems a bit complicated! So I see you have tried the Neodynamic GTI and are going to try further well that motivates me to try here too and lets see how soon I can do it. read your website and very impressive ideas, salute to you. here is a redrawn schematic of the neodynamic GTI in continuation with my discussion yesterdayView attachment 67934

 

[Fendel]

Hi Plodhi
On figs 13, 15, 20, 28 of:
https://www.electro-tech-online.com/custompdfs/2012/10/pv_appnotes.pdf
Are some schematic of direct high voltage outputs, and they seems very simple.

In your schematic one triac is inverted?
I am just trying to understand these directions and how to make the minimum gate controls.

As say Tcmtech, the grids are like very great and very strong AC "acumulators" and, also its easy to put litle LC filters to round the stronger spikes.
And I say that some injected spikes from our GTIs cancel some spikes made from other devices and other consumers, because they are in real oposite directions.
Also, a cable feed from 2 extremes, can transports doble the usual energy, and if plus feed in the midle, suports 4 times more, and so on, with plug and play controlers; but at first, with smal and simple GTs, the increase of only 20% from grids capacities, is almost free and benefits all involved. That is the ENERNET, very different from the "stupid grid" they dreams.

 

[tcmtech]

Hmm... 10 uf in series with a 220 volt AC 50 - Hz system.

Per this site, , that works out to a series impedance of about 318 Ohms.
So given a 220 volt AC input that works out to around 220/318 = .69 amps or roughly 152 VA capacitive load.

Now given the 5 volt DC input as mentioned earlier that would give a theoretical feedback of, ((5/318 = .016) x 5)/2 = .04 watts with a power factor approaching zero.

However with a 220 VDC input that works out to around 76 watts with a power factor around .5

**broken link removed**

 

[Fendel]

Hi Tcmtech

Did you have some sugestions to me, to make a litle square GTI?
What do you think about 4 random SSR?

 

[tcmtech]

Not sure what a square GTI is. H bridge?

I suppose you could make one any shape you want. I prefer to use whatever used box or chassis I can find myself.

As far as simple, scalable, and actually works I have a number of different very basic units already posted in other GTI threads in this section.

Two transformers, two switching devices, two resistors, and a diode is as simple as you can get and have a solid working self synchronizing design with true line isolation and the capacity to use mismatched DC input/AC output voltages.

 

[plodhi2002]

View attachment 68024
Hi Fendel I have corrected the schematic as you pointed out the error. Now it is as per the original given in the Neodynamic site except for the fuses. I see that now other technical friends are coming in and we will hopefully have some output on this schematic. By the way the originator of this Phi Tran did not answer the queries. But I guess since we have so many experience people like tcmtech and nikos here to learn from no need to worry.

 

[Fendel]

Tcmtech, your boxes are not square, they are all parallelepipet.
If your cd is not yet finnischt, why do you not sell the archives incomplete?
Wath are the price espected for it?

Plodhi, the matter seems me like to assemble a car, with various diferent pieces, from diferent marks.
So, we have to put a 6 bore weehl in a 5 bolt axe.
To the experienced people, that is light, they put the wells on the cnc maschine and the work is done.
But for us... we have to make the bores by hand.

 

[tcmtech]

Tcmtech, your boxes are not square, they are all parallelepipet.

So what does the format of a schematic have to do with the functionality of the design?

 

[Fendel]

My dear Tcmtech

Whit square, we mind square wave, curve... not sinusoidal...
Today I travel 250 km specialy to buy igbts, triacs, scrs, random optocoplers, two new transaformators, one 48V 1200 VA and the other 48V 960 VA, and so on.
What do think about the use of 2 or 4 random ssr in place of the igbt? because igbt are very expensive here.

 

[tcmtech]

I see now. Square wave based switching system.

Well for the most part once the initial square wave from the primary side of the GTI goes through the transformer and through the AC line side filters its not really much of an issue. Besides these designs are intended to be the simplest and most basic versions of a DIY GTI as possible.

Given that these units are likely going to be built to work at power levels under a few KW's at best any line noise is not going to get very far.

I have ran these same circuits with a number of different transformers in capacities up to around 3+ KW peak feeding back into a normal 120 volt 30 amp line and never had any issues with them causing any levels of interference with other devices on the same line or with any radio device as well.

If the zero crossing dead band between each half of the cycle are set right the actual AC line side output blends right in with the normal utility systems sine wave shape.

As far as using SSR's for the main switching devices I cant see where you would be able to buy any of them with DC switching ratings in large enough sizes that would be cheaper than any IGBT or power Mosfet device.

 

[Fendel]

Great job, Tcmtech.

As far I understand it, the zero crossing conection happens in the next cicle of the wave, also in the oposite direction...
So, a random device seems me easier to use.

Here (and in all places of the world) we have very stupid laws, and so, for me it is cheaper and easier to obtain strong ssrs than igbts... and ssrs are finisched and easy to assemble.
The question is, if an random ssr connects and desconects as fast as an igbt in every cicle. I dont find these informations in datashets.

 

[tcmtech]

Is it a DC rated SSR? If not it wont work on the DC side of the system.

As I mentioned earlier Mosfets are a direct drop in for IGBT's in these type of switching circuits.

Also transistors will do just fine as well however the base drive current is higher so the drive circuit needs to be able to handle the added current flow.

As far as I know the IR2113 IC's I have in the schematics are also capable of driving the bases of transistors just fine. The only change is that the resistors on their two outputs need to be sized different to limit the peak base current to what ever the IC can safely handle and the power circuit and voltage regulator (U1) need to be able to handle the added current draw as well.

 

[Fendel]

I buied 4,5 V Gate threshold voltage Igbts. So the series resistors may be 500 ohms?

 

[tcmtech]

I would stick with the 100 ohms or anything close. Given the low frequency switching that they are doing this part is not all that critical and I suspect that anything from a few tens of ohms up into a few hundred ohms would not have all that much effect on their switching characteristics.

What are their current ratings and what input power are you going with on your GTI? Volts amps power source stuff like that?

 

[Fendel]

My dear Tcmtech

In place of the CD, why do you not sell entire GTIs?
I bought one China GTI 1200 W in Germany for us$ 330, and a friend bring me it.
In the banch I tested my pmg made from a old 1 kW servomotor copled to a 2 hp motor and a 1 hp inverter.
I found that the MPPT function is to slow and not good for a wind generator, and I am studing a adicional control to put the generator on my 45 m height swing tower.

View attachment 68062

I begun to read these great GTI treds and I am very happy to found them, and read it for many days.
Until it, I made some litle conected hidropower and vegoil and biogas with assyncronous motors as generators, during the last 30 years. All in clandestine. Now in next december we will have a law to permit the conections.
So I want make litle 0,5 to 1,5 kW conected wind machines, to sell.

 

[tcmtech]

Well largely in the USA I cant sell GTI units legally without having UL certification and possible patent rights along with an number of other highly expensive and time consuming legal actions having been done first.

Mostly I am just content with giving out the basic knowledge of how to build very simple but highly adaptable and reliable units for the DIY people who wish to build their own.
Believe me I know what its like to be poor and just getting by. I lived that way most of my life so I understand the potential value of how these simpler but still fully functional designs can be appreciated by those who have more time and knowledge resources in hand than money.

As the CD goes since I had started working on it a few years ago I now have a high paying job, a great wife, a pile of stocks that are doing very well, and a new house I am building mostly by myself which means I don't have the time to work on it or the need for the money it could have brought.

That and to make room for the new much much larger house I had to clear out my old electronics work shed and burn it down. (Presently the wife won't let me put up a new one on the temerary house we are living in.)

 

[Fendel]

If I understand right, you said you once want built a direct GTI for 2 kW. Is that right?
Direct is without transformator?
Here we dont have litle PMG nor PMM. So I think to make some with pieces of normal assyncronous motors, like 0,5 kW and 8 poles, making a new rotor with niodimium magnets.
So we have generators in the range of 900 RPM at 380 V -3 fases - 60 Hz... that give us 220V at 500 RPM... and perhaps 200 W?
Wath hapens if I put your H bridge direct on the grid? with so a hight voltage PMG?
With high voltage, all are less... minor cables, minor igbts, and so on...