Grid Tie Inverter Schematic 2.0

[tcmtech]

If you are thinking of my 15 - 20 KW experiment yes that was direct line connection. It still had sizable inductors on the AC line side of things to clean up any switching noises.

As far as the IGBT's were concerned I used 600 amp 1200 volt rated ones so switching at anything under 400 VDC is not a problem.

Never finished the prototype though. No present need for it or a large enough AE based power source to drive it. I gave it a few test runs to see if the design concept worked then set it under the bench in the shop and that is where it has been ever since.

 

[Fendel]

yesterday I put my firts watts on the wall plug, from your here described GTI:
I used a fan as suport to the dissipators, and plug it on to reduce the igbt temperature

View attachment 68085

Today I will make more tests, until it beguin to smoke
In some igbt datashets is writen:
"Excellent Current Sharing in Parallel Operation."
To put some igbts in paralel (because price and offer) wat I have to do? Change the series resistors?

 

[tcmtech]

Just use the same size resistor for each IGBT that is being driven.

The IR2113's driver IC's outputs can source and sink about 2 amps so given the rather low line frequencies they are being run at that works out to them being able to easily handle at least a dozen IGBT's on each of their outputs if each IGBT gate is connected through a 100 ohm resistor.

Given that the biggest IGBT's I have ran with them have gate capacitance rating of around 42,000 pF they can take a load of smaller IGBTS without issue.

View attachment IR2110 - IR2113 HALF BRIDGE DRIVER.pdf

Here are a couple of good data sheet websites as well.

http://www.datasheetcatalog.com/
http://www.datasheetarchive.com/

 

[Fendel]

View attachment 68128

That is the thirth time I make these post, I hope now send...

My current and power meters are yet not fiables.
At 500 RPM it got 2,5 A, 53 V and 147 seconds to make a round on the k = 2,4 W/R analog power meter, that gives:
133 W DC site and 59 W AC site.
At 600 RPM - 5 A, 58V, 50 s = 290 WDV and 173 WAC
At 760 RPM - 10 A, 67 V and not time to red the time... because the 12A IGBT dissipator want hot.
AT 15 A no time to read anithynk... only observe the nice fire work at the IGBTs...

 

[tcmtech]

At 15 amps you would need at least 60 amps of IGBT capacity on each of the four switching points of the H bridge assuming your zero crossing dead band is not set too narrow.

Turning the switching devices on and off too low on the sine wave really drives up their peak current carrying load and can easily hide the possibility that they may actually be carrying far more than 4x the average amp readings.

Believe me I toasted more than just few switching devices learning that little detail and its also why I highly recommend overbuilding the H bridge relating to its current handling capacity.

On a properly set up and tuned GTI 4x current handing capacity has never given me much problems however getting to the point of being set up can get to be very expensive test and tune experience!

I am assuming your IGBT's are the type that have the built in reverse conduction diodes?

I would also say that if you are aiming for a 1 KW or slightly higher working range that having a 290 watt input for a 173 watt output is already running about 60% efficiency which is not bad at all for being at roughly 25 - 35% peak working capacity! There is a very good chance once you have the system properly set up you may see peak efficiency well above 90+%!

 

[Fendel]

Today arrived to me a new digital dc current meter, and then I compared mesured Amps in the old are about 30% higer than in the new one, at the compared rates. For higher Amps, above 20 A (old) the diferences rise to 60%!!!
So, the efficience of the burned GTI was also not so bad.
I also dont put any capacitors at the GTI output, I yet not find the values on the texts.
Also arrived a Arduino CLP and I understand only a litle Ladder language, but the program must be writen in C. Thats not for now.
I waiting the arrive from a simple VDC controler, to put my mashine on the top.
Only for curiosity, Tcmtech, how do you make to improve by 10 the mesure of your kW meter? You rewind it? they are 2 ore 3 windings...
And are yours new digital kW meters in usa also only unidirecional like here?

 

[tcmtech]

The output side capacitors are for a combination of power factor correction and waveform shaping and there values are not all that critical.

The simplest way to do it would be to hook your transformer up to the AC power and do a reading of its no load current draw and keep adding capacitance to it until it reaches its lowest current reading. At that point your no load power factor will be very close to 1 and that should be enough for the system to run without passing any serious levels of harmonics or switching noise.

Yep I just rewound the current sensing coils to a 10:1 ratio over the old ones. I also have a 100:1 I rewound for very fine readings. As far as our digital units here it all depends on how they are programmed.

One trick you may like to know is that if you need a way to further isolate your new windings from the iron core that they go around coat the iron core in epoxy first and let it cure. You can get your windings very tight that way and not loose any space which when trying to fit 10x or especially 100x number of windings in those little spaces is critical.

 

[Fendel]

https://www.electro-tech-online.com/custompdfs/2012/10/irg4pc60fpbf.pdf

Make it sense ad external diodes to above IGBTs? and if, wath diodes?

At the DC imput, behind the diode, I put a 330 uf 400V electrolitic capacitor, and after the measures and fire, it was a litle warm. Is that ok?

Do you know how your digital normal usual wattmeters are programed?

With a SCR or Triac, how can I make a simple normaly closed swith? I intend to automatic close the PMG output if the grid fail, to protect and brake the wind turbine.

 

[tcmtech]

Ideally you want one with similar ratings for speed as the IGBT has but given the low frequency that these circuits run at most any good power diodes would work. Anything with a amp rating over 1/3 of what your IGBT is conducting should be close enough.

Depending on the ESR of the capacitor and overall current flow that's fairly normal for them to warm up some. By behind do you mean the DC input side C1 or the IGBT side of the input diode C2?
The IGBT side capacitor C2 should be fairly small like around 10 uf. The DC input side C1 should be fairly large and can be anything from around 1000 to 10,000+ uf. I recommended a 2500 uf (2.5mf) but it's also a component that has a large amount of leeway regarding its size.

As far as the digital watt meters go I don't have a clue. I only have two of them and I have yet to ever bother trying to find out how they are programmed. Odds are If I took them to my local power company the guys in the shop would tell me and probably program them for me as well for next to nothing.

 

[Fendel]

Can you help me to define the folowing circuits, like values and pieces for:
- a normaly closed SCR or Triac for the output PMG or imput GTI, between + and -, until 250 VDC and 40A... (= safety brake, in case of grid fail) (like these switch open wen it has a 12 or 24 VDC input, and close if it is zero)
- a normaly open SCR or Triac to ad a dum load at the DC site, to protect over charge of the GTI, also switched by a 12 or 24 VDC imput.
In the atempt to make these circuits, I burned a SCR, and dont get turn off the Triac...

 

[tcmtech]

What I have done with mine is to just use a large contactor like a riding lawn mower or small automotive type starter contactor with a continuous duty rated coil to short circuit the power leads for my generators.

By using a simple relay driven from the AC line side to trigger the contactor or a voltage sensing circuit that latches the contactor on if the input voltage goes to high from high winds with a manual reset to bring it back on line after a high wind shutdown you will have both sides to of the system capable of triggering an automatic shutdown.

Of course the contactor and control circuits will need a battery based power source once the generator or line power goes off.

 

[Fendel]

Today I tried to link a Triac BTA41-600 for 40A and 600V... and:
I put some resistors betwen gate and MT2 and as result I hope it serves as emergengy brake for the wind generator.
The measures of V are betwen the positive and negative pole of the generator rectifier, also conected between terminal MT1 and MT2 :

Ohm - V - RPM
3k3 30 270
6k6 60 550
8k2 75 660
9k9 90 800

At the mesuared voltage, the Triac closes and brake the generator until a manual reset, or until the wind stops and make free the Triac, or... the triac burn.
Now a will put the Triac on a bigger dissipator, to see if it not burn...

 

[Fendel]

The great heat sink dissipator I put, lets the triac cold, but his legs want very hot, at 20 and 30 A...
So I thinking put some in paralel... but, dont know if they dispare at the same time, or voltage, and if not... only one will connect and his legs can burn...
Perhaps solder the bigger vires direct on the case, at the beguining from the legs?

Then cames me the idea to put the triacs in the AC 3 fase legs of the generator, before the diodes...
I make some simulations, and I think its a great deal.
So, I want make 3 triacs to schort circuit at 70 VAC and more 3 triacs to conect the dump load at 55 VAC.
If the rotation, or tension, or wind speed is under these numbers, the triacs open automatic, and the MPPT GTI can make wath it wish.
Wath do you think about?

 

[tcmtech]

Wath do you think about? .

I think English is probably your second language?

As far as short circuiting things with solid state devices thats fine provided you do a bit of calculating first. You need to know your overall system resistance and the peak voltage you plan to short out.

So say your system has a series resistance of the generator windings plus line sets of around 1 ohm and you are putting it to a dead short at 70 VAC the peak of the AC voltage will be nearly 100 volts and 100 volts dumped through a 1 ohm resistance will produce a short circuit peak current of 100 amps.

So from that obviously you need semiconductors capable of safely handling that peak short circuit current. In realistic application I would design the short circuit switching devices to handle at least double that.

 

[Fendel]

No, no, english is my fourth language... before comes portuguese, german and spanish.

To make a dump load at VAC site of the PMG with Triacs, I have to make 3 resistors, in the near of 6 ohm each.
Not so easy make wind cooled resistors for 6 ohm and 500 W each, near the PMG.
Any sugestion?

 

[tcmtech]

English is my only language barely.

Home built high capacity wire wound power resistors are easy. Get yourself a few cheap notebooks with the spiral wire wound backs.

The wire is usually a chrome plated steel and if given reasonable air flow so that the chrome does not melt or flake off work very well as large power resistors.

Figure that depending on the wire gauge and length a typical notebook wire is between 1 and 3 ohms and good for 200 - 300 watts with any reasonable air flow.

Pick up a bunch of ceramic or glass beads at a hobby shop and you have dirt cheap high temp insulators too!

 

[Fendel]

Spiral wire in old notebooks? Who are these spiral? What it makes at a notebook?

Short circuit at PMG is not so good as a dump load, to make it as a emergency brake for higher rotation.
A load makes the PMG turn slower than a short.

 

[tcmtech]

Spiral wound notebooks are those paper books that children use at school to write things in. Around here you can buy them for about $1 each.
The cheap ones have their spines held together with a single chrome colored wire that resembles a heater coil.

The other option is to buy the cheapest replacement heating element you can find for an electric cloths dryer and cut that up to make your load bank out of. Something like this for example. https://www.ebay.com/itm/Replacemen...654?pt=LH_DefaultDomain_0&hash=item5aea69eb1e

All of my PM motor based wind turbines always turned slowest at a dead short. It's maximum load at minimum output voltage and it always kept them well under their stall speeds so that they could never get enough RPM's up to get into their efficient speed range or higher and start running way.

Maybe with a homemade PMG its different but I have always used off the shelf commercial Permanent magnet AC or DC servo motors as my generators.

 

[Mr RB]

Here in Australia you can buy different types of "jug element" as used in jugs you boil water for coffee etc. They are commonly available everywhere, hardware stores to supermarkets etc. One type has an exposed stainless steel wire over a ceramic former, with two long brass power rods.

Submerged in water they are about 1500W at 240v AC, but they work fine in air as dummy loads at lower wattages of course.

 

[Fendel]

Here we have big schower resistences - 8 Kw - 220V... naked wire, in direct contact with water...
I go cut some piece, woun around the holes of a brick, conect an put it in a 200 liter water barrel.
Thats god for the DC dump loat at ground level.

Now I am triyng to make a 50VDC controler, to limit the PMG voltage, because the chinese GTI not accept higher voltages.
Firt I tried control the DC site, but the triacs dont desconect the schort or the dump, until the voltage backs to zero volt.
So I got to the 3 fase AC site, with 2 Triacs, but to control that with the proper AC voltage is not so easy for me, as I think first...