External limiting an GTIL2 SUN-1000G2 with raspberry

It sure looks like you have RS232 then.
To communicate, you need a Modbus master. I used a demo applicationfrom here with parameters set as I posted here.
I never found an English version of the protocol document, but was able to get the gist of it using Google Tranlate.
 
watts-on said:
It sure looks like you have RS232 then.
To communicate, you need a Modbus master. I used a demo applicationfrom here with parameters set as I posted here.
I never found an English version of the protocol document, but was able to get the gist of it using Google Tranlate.
You may right, it is probably RS232,
I saw this diagram on internet and i put the picture i taken yesterday, do you think this USB-RS232 Serial cable might work?


image_ztvqvn.jpg
 
Any USB to serial should work fine.
As I said before, I made a lead to go between the serial adapter and the GTIL because I was concerned about pin 9. It may be fine, or it may not, I just didn't want the chance it and fry something.
From memory, I think it was pins 2 to 2, 3 to 3 and 5 to 5. Although it was a while ago and I don't have it set up any more.
 
watts-on said:
Any USB to serial should work fine.
As I said before, I made a lead to go between the serial adapter and the GTIL because I was concerned about pin 9. It may be fine, or it may not, I just didn't want the chance it and fry something.
From memory, I think it was pins 2 to 2, 3 to 3 and 5 to 5. Although it was a while ago and I don't have it set up any more.

Mine doesn't work using USB-RS232 Serial, it might be defective or we need to use reverser null cable (2->3, 3->2). But on my Wifi Serial Device, ,i can now parse data
 
kmsj13 said:
jkammler said:
Hallo,

I also use this unit - and was looking into controlling the output. From what I learned so far :

- My box has a RS232 output, same board as shown in the last post from kmsj13 (the chip is a MAX3232).

- Could confirm the 4 pins being used (pins 2,3,5,9).

- On pin 9 some 12.x volts against pin 5 (ground).

- I can control the power output (Watts) by using a 5kohm poti when connected to the normally not used, open 2 pin socket, in the row of sockets where the externel sensor cable also goes to (see the 2nd picture from 12-31-2017, 12:10 AM, from rolands post, the black and red cable). Did not need to connect anything else to ground so.

- This control is possible only when the "Limit Mode" checkbox is NOT checked.


That is what I could confirm on my box.


Next step:

- Use a MCP4132 as Digi_Pot controlled by a Raspberry PI (I think I get this to work).

- For current meassuring I will use ACS758LCB-050B modules, where the analog (out2) signal is converted by a ADS1115 to I2C on the Raspberry.


Question: Did I overlook a translated version of the "Eybond Modbus RTU Protocol V100" document ? Since I did not yet get the RS232 interface to talk to me yet, so I assume I do something wrong. Does anyone know of a quick test (some confirmed string to send) for this interface?


Your help is greatly appreciated.

Hi,
does the pcb looks the same as yours v1.2?
I tried using usb-rs232 serial but cant get data, i tried usb-rs485 serial but i was getting some data corruption and cant decode strings...

Yes, pcb looks the same.
 
watts-on said:
OW1000 said:
Is there any news on this topic? Would be very useful to be Able to control the GTIL2 externally.

I have put together my diy limiter design and am waiting for the boards to come in from China:


image_hcdole.jpg


It has two comms ports that can be configured for TTL UART, RS232 or RS422/485 in either half or full duplex.

The idea is that one port will be for control (I'll probably use modbus protocol) and the other for monitoring only / software updating.
It also has a relay allowing the GTIL it to be disconnected from the mains when not in use to save power.

There is also a spare output that could theoretically be used to control an external contactor to disconnect the DC for the same reason.

It's a fully external unit, i.e. no modifications required to the GTIL inverter (in theory, if it works :) )
hi watts-on
this working like you want ?
 
Info:
- Can now talk from my Raspberry via RS232 to the box.
- My CRC calculation also works.
- For power control (as described before) a 680 ohms resistor gives me about 120 W output.
- I used pins 1 and 6 on the internal RS232 board to put the current control signal to the outside world. They were not connected to anything.
- I am going to make a custom cable to the raspberry for the serial interface and the power control.
- Will also make a little test program to see what registers (see Modbus dokumentation) give me some useful values back. Probably tomorrow ;)
 
watts-on said:
Patalhetas said:
watts-on said:
hi watts-on
this working like you want ?

It's definitely doing something :)


image_hekuhu.jpg

image_bosrfc.jpg


I didn't get the voltage monitoring circuit right first time, so had to bodge on a mod. Current monitoring seems fine so far though.
...and controling the inverter ? or is only for monitoring!
 
Patalhetas said:
...and controling the inverter ? or is only for monitoring!

Controlling the inverter is the end goal. But before that, I have to test and calibrate the different sections of the design to ensure they work OK. Then I have code to write to bring it all together.
 
kmsj13 said:
I opened my GTIL2 and saw this Wifi Plug Ver 1.2 which is different from PCB of Sir Watt.
The IC doesn't have any label so i wonder if its MAX3232 IC.

image_weokpd.jpg



I tried to use lights to see the connection:
DB9:
PIN 5 -GND
PIN 2 - Might be TX
PIN 3 - Might be RX
PIN 12 - 12v


image_cibaoz.jpg

image_pwefrm.jpg

image_nyyurn.jpg


clean well with isopropyl and gently remove the coating that appears to be silicone and you will see the code of the chip
 
So my RS232 connection works well. BUT I am a little disappointed from what information the inverter corresponds, just had hoped for more.
Besides registers addresses 0 to 7 (basically the serial # of the box) there are only 6 other values:
Registers 43 (? no idea what I can do with this), 59 (a status), 60 (Day active Power), 63 (Total active power), 70 (Grid Voltage), 86 (Power out) and 109 (Battery/Collector voltage).
So I will log the last 4 values, and do the controlling of the box via ext. HW connected to this 0 to 1.6 V internal connector.
 
Hello
and good morning everyone and congratulations for the work / ideas / explanations you are giving.
I have a 1000 inverter, like the dakoal inverter, which has a
problem: it worked perfectly for more than 1 year up to 10 days ago now the PV input short-circuits the photovoltaic panels.
The capacitors are OK, and I'm afraid they are mosfets.
I read that someone from the group replaced them, could you give me the initials? So replace them.
Many thanks and congratulations again
 
Last Update:
Reading the Inverter data and setting the output power works (via RS232). Thanks for the fine information in this thread that helped me.


image_vyokgg.jpg


image_ywvzbo.jpg


image_ivycru.jpg
 
Hi folks,

I have invertor SUN-2000G2. Can you help me with decoding of output data from modbus?

Currently, the inverter is not connected to panels. I expected zero values :(.



Code:
> client.read_holding_registers address=0 count=8 unit=1                                                                                       
{
  "registers": [
    2, 
    256, 
    1, 
    12601, 
    12594, 
    12594, 
    12336, 
    14641
  ]
}
 
Doin it said:
I am not grounding the CT or shielding at all and it always reads accurately...
also did u install the required resistor in your sensor cables wiring?
U are reading negative on both L1 and L2.. if u only read negative on one of the legs (once u correct the probelem) then this is normal.. the reason negative on one leg is normal (not backfeed) is due to the way your electric company meter reads your home usage..,, it reads both phases as a whole so negative on one leg gets subtracted from other leg to net zero..

Doin it, it turns out the GTI with limiter was defective. I swapped it out and things seemed to be working. Limiter circuitry would lose its mind and GTI would just start over producing.

I want to make one note to anyone out here who is using 240v GTI with limiter . Everything posted in the past with polarity of CT on L1 and L2 is correct with the exception of one thing I noted.

I believe at some point in this thread it was said that the 2 CT should be connected in series. Well that is wrong, they should be connected in parallel. They are after all Current Transformers and CT connected in series does not sum the currents, currents are summed in parallel. So CT on L2 and L1 connected in parallel will sum currents.

I was looking over my install one day and noted the problem.

I am of course referring to 240v GTI installed in NA where we have 240v split voltage .
 
MartinCZ said:
Hi folks,

I have invertor SUN-2000G2. Can you help me with decoding of output data from modbus?

Currently, the inverter is not connected to panels. I expected zero values :(.



Code:
> client.read_holding_registers address=0 count=8 unit=1                                                                                       
{
  "registers": [
    2, 
    256, 
    1, 
    12601, 
    12594, 
    12594, 
    12336, 
    14641
  ]
}

Try this, should read grid voltage:

client.read_holding_registers address=46 count=1 unit=1
 
@lepdiypowerwalls, thank you.

Does any documentation exist for this? Something like @jkammler's app screenshot (comment 63741).
I want to create a module for sending information from the inverter to MQTT broker (Home Automatization) by ESP32.
 
Some Info of - and a test programfor - the RS232 communication attached. ;)
 

Attachments

  • spi_SerIO_20200719_SUN_1000W_Grit_Tie_Inverter.c.txt
    11.2 KB · Views: 934
  • Log_20200719_SUN_1000W_Grit_Tie_Inverter__relevant.pdf
    22.9 KB · Views: 859
Back
Top