Kai Deng GTI300W Grid tie inverter reverse engineering

Ladeelektronik, Messgeräte, Steuerungen etc.

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon microsolar » Mo 11. Feb 2013, 21:29

Hi vasonline,

there are tow off this ICs (IC2 and U2) on my board (GTI200W). One on the front and one on the rear side.

R58 and R59 goto the LED.

When the converter was working (last year) and the input from the solarpanel was about 10-15 W, the converter also always reseting.

microsolar
Dateianhänge
U2.JPG
U2
U2.JPG (291.19 KiB) 11728-mal betrachtet
IC2.jpg
IC2
IC2.jpg (224.67 KiB) 11728-mal betrachtet
microsolar
 
Beiträge: 36
Registriert: Mo 10. Sep 2012, 10:33

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon microsolar » Di 12. Feb 2013, 08:22

Hallo Zusammen,

jetzt habe ich mir die zwei Chips auf meinem GTI200W noch mal genauer angeschaut. Ich denke IC2 ist der Mircocontroler mit 18 Pins und U2 mit 16 Pins entspricht evtl. dem UCC3806 auf dem board von dem GTI300W. Werde mal prüfen ob bei meinen GTI200W der IC2 überhaut eine Spannung bekommt.

Viele Grüße,
microsolar
Zuletzt geändert von microsolar am Di 12. Feb 2013, 22:12, insgesamt 1-mal geändert.
microsolar
 
Beiträge: 36
Registriert: Mo 10. Sep 2012, 10:33

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon vasonline » Di 12. Feb 2013, 12:37

Hi Microsolar,

Vielen dank fur ihre photos :)
Many thanks for your pictures.
It appears your GTI200W uses the same design as mine. Just with a different layout.

Cool that you found "the same" bug: it cycles endlessly when the input power is below ~15W. (green led on/off)
I measured the dc input: it drags the solar voltage down to 7-9V at which point it falls below it's minimum input range (10,5V) And shuts off.
A few times later DC rises again to ~18V and the whole thing repeats until there is enough Solar input...)

This led me to think that the firmware is identical or nearly identical.

The two LEDS are wired the same on my IC2. I now believe it's indeed a PIC. I have read some of the partially sanded of name.
And with my pic programmer I can connect to it. Unfortunatly I believe mine is code protected.

I see a programming connector that is wired the same. GND is the middle pin.

*edit* All parts number appear to be the same :)
vasonline
 
Beiträge: 60
Registriert: Sa 2. Feb 2013, 20:46

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon vasonline » Di 12. Feb 2013, 13:02

Updated drawing:
Bild

For your GTI200W:
- We need to know if the fault is on the DC side or the AC side.
- Does the red led light up when connected to DC input? If no: check if there is 5V present on pin 14 of IC2 (see my drawing)
- If the red led lights up: measure the AC fuse: if blown: measure all FET's rf7, rf17, rf18 rf8 (multimeter auf diode-stand)
- I had no problems with the DC, but you can alse measure those for shorts. rf10, rf9, rf12, rf11, rf6, rf5

Do these part numbers match?

Greetings & good luck!!
vasonline
 
Beiträge: 60
Registriert: Sa 2. Feb 2013, 20:46

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon microsolar » Di 12. Feb 2013, 22:10

Hallo vasonline,

- 5A 250V (AC) Fuse is OK
- 20A Fuse (DC) is OK
- Power Switch is OK
- Fan is OK
- RED and GREEN LED is OK

Ich habe den GTI200W an eine Labornetzteil angeschlossen.

input :

0 - 5V --> ca. 0.05 A
5,5 V --> ca. 0.8 A
6,0 V --> ca. 1,2 A

RF9,10,11,12 is an "IRF3205 MT 1H3H" HEXFET® Power MOSFET

after about 60 s, the RF11 is hot and RF12 warm. RF10 and RF 9 normal.

- Does the red led light up when connected to DC input?

No.

- If no: check if there is 5V present on pin 14 of IC2 (see my drawing)

Input is ca. 6.0 V --> Pin 14 is ca. 2.2 V

So I think, the DC part is defekt.

rf7, rf17, rf18 rf8 part numbers matching. its black but type not readable
rf6, rf5 are not on my board
D24 is a C1125 (ON) 1660CRT
D?? is a (ON)1208 U1660G

How to measure FET's with (multimeter auf diode-stand) ?
Can I do this if the FET's are on the borad ?

Thanks for your help !

microsolar
microsolar
 
Beiträge: 36
Registriert: Mo 10. Sep 2012, 10:33

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon microsolar » Mi 13. Feb 2013, 09:13

Hallo vasonline,

Auf den 2 Transformatoren meines GTI200W ist die Aufschrift: GTI300W 15V/230V
L3 schaut optisch identisch aus.

Aufbau der Platiene (Rückseite):

+--------------------+--------------------+
| .DC.12V........|...Logic...........|
+--------------------+--------------------+
| .AC.230V.............................|
+--------------------+--------------------+

DC 12V Teil ist über den Transformator mit dem AC 230V Teil verbunden.
Logic Teil ist über 2 Optokoppler (?) ( 2 chips mit 4 pins ) mit AC 230V Teil (neben dem Transformator) und einen blauen Kondensator (?) (zwischen rot FUSE 230V und Connector J6 ) verbunden.

Logic beinhaltet:

- LED Anzeige
- FAN
- Power Switch
+ ??? PWM + ???

Was macht pin 18 Voltage sensor an IC2 ? Ist der für den AC-Teil oder DC-Teil ?

Wir sollten bestimmen welche I/O, ADC, DAC Ports der IC2 hat und welche Funktion sie haben.

Eventuell läst sich ja der "Logic" Teil Nachbilden, dann könnte man auch einen anderen Microcontroler verwerden. Eine Software für die PowerJack gibt es ja wohl hier im Forum, veileicht ist das ähnlich aufgebaut.

Ist der connector J6 auf den Bild zum Programmieren des IC2 ?

Viele Grüße,

Microsolar
Dateianhänge
Conector_J_6.jpg
Connector J6
Conector_J_6.jpg (165.37 KiB) 11676-mal betrachtet
microsolar
 
Beiträge: 36
Registriert: Mo 10. Sep 2012, 10:33

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon vasonline » Mi 13. Feb 2013, 09:51

Hi Microsolar,

I believe you are on the right track. :)

Indeed I would think you have a DC problem.
Check out the FET's on the DC side for shorts (multimeter set to ohms)

On my working unit I get: infinite resistance in one direction, and around 6Megaohms in the other direction).
Measured from center pin to the left pin and from center to the right pin (and from right to center, left to center)
It's no problem to measure them in-circuit.

Power consumption should be very low, when powered from a 6V power supply: on my unit it is around 10mA.

I believe pin14 of IC2 does not get much voltage because everything is going trough one(?) shorted DC FET.

Pin 18 of IC2 is used to sense the startup and shut off voltage (10,5V and ...28V) into the microcontroller. It is via 2 resistors connected to the + (after the power switch)

The programming connector on my unit is also labeled J6.
Cool: the 5 pin connector is already installed on your unit!
I had to solder in the pin's myself.
If you can get the unit working I would be very curious to see if you can extract the code from the PIC. Mine appears code protected.
vasonline
 
Beiträge: 60
Registriert: Sa 2. Feb 2013, 20:46

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon vasonline » Do 14. Feb 2013, 11:29

Updated drawing:

Bild
vasonline
 
Beiträge: 60
Registriert: Sa 2. Feb 2013, 20:46

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon microsolar » Do 14. Feb 2013, 12:38

Hallo vasonline,

Ich habe gemessen:

Jeweils 1-ter, 2-ter, 3-ter Pin

PINS___RF11____RF12____RF9____RF10
+ 1-ter Pin
1 -> 2 = 358 K, 333 K, 329 K, 327 K
1 -> 3 = 328 K, 327 K, 328 K, 321 K
2 -> 3 = 0 K, 0 K, 0 K, 0 K (Schwankt zwischen -10K und +10K, wird wohl der Kondensator geladen oder entladen )

- 1-ter Pin
1 -> 2 = 210 K, 204 K, 208 K, 211 K
1 -> 3 = 201 K, 204 K, 200 K, 204 K
2 -> 3 = 0 K, 0 K, 0 K, 0 K (Schwankt zwischen -10K und +10K, wird wohl der Kondensator geladen oder entladen )

Viele Grüße,

microsolar
microsolar
 
Beiträge: 36
Registriert: Mo 10. Sep 2012, 10:33

Re: Kai Deng GTI300W Grid tie inverter reverse engineering

Beitragvon vasonline » Do 14. Feb 2013, 20:12

Hi microsolar,

Your values seem good/fine.
However, rf11 -as you said- should not get warm or hot at all. (below 10,5v everything should be still 'sleeping')

Maybe you can measure the voltage across those fet's when you supply around 6v DC input?
Do they also get warm when the on/off switch is set to "off"? That would exclude the logic from beiing at fault.

Because, apart from the capacitor, fuse, there are not much other powers components on the dc side
vasonline
 
Beiträge: 60
Registriert: Sa 2. Feb 2013, 20:46

VorherigeNächste

Zurück zu Elektronik für Windkraftanlagen



Wer ist online?

Mitglieder in diesem Forum: 0 Mitglieder und 8 Gäste

cron