External limiting an GTIL2 SUN-1000G2 with raspberry

Announcement - Help us fight the BOTS! Please report all spam including stuff in your inbox!

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,569
Cheap4-life said:
Series. It is also recommended to use a resistor to make the resistance of the dual cable similar to the resistance of a cable that uses one ct. a trim pot helps fine tune

ajw22 said:
I've done mine in parallel, to add the currents produced by the CTs. No additional resistor.
OK now I am really confused.............. :huh: :huh: :huh:
One says Series one says Parallel.
I'm glad I have mine individual,1 for each leg............
Wolf
 

Cheap 4-life

Member
Joined
Aug 3, 2020
Messages
372
This is a pic of the cable opened up from the largest seller of the dual sensing cables to supply 240v to a split phase panel. He now adds a trim pot for fine tuning.
I have shown the graphs from iotawatt to confirm that this works wonderfully. One month I only payed $4 for electric.
https://www.facebook.com/groups/1455824754596738/

image_vskcrv.jpg

I couldnt have ANY backfeeding so I even went to 60ma clamps instead of 50ma. The 60ma clamps allow the grid to supply 15-35w instead of the 0-10w typically seen. 0 was to close to backfeeding imo. The amount the inverter says the grid is supply isnt correct when using the 60ma clamps but thats ok with me because now there definitely isnt any backfeeding with room to spare. I monitor the amount the grid is still supplying with the iotawatt which is more accurate anyways.


Parallel might also work. The inverter simply produces till the clamps are near zeroed.


This is what the main seller of the dual sensing cables said many moons ago on this thread.

image_cnbpmh.jpg

Thats why the dual sensing cables are made. Most splitphase utility meters only have one ct. The dual sensing cables combine 2 CTs into one so it reads like the utility meter at 240v.


There is a little sight glass on most meters that allows you to see inside to see how many CTs your meter has. Its small but if you have good vision you can see the ct. If theres one ct inside the meter then use the dual cables and supply 240v from the inverters. If it has 2 CTs inside the meter then use a separate ct for each phase-leg and supply 120v instead.
 

Cheap 4-life

Member
Joined
Aug 3, 2020
Messages
372
ajw22 said:
I've done mine in parallel, to add the currents produced by the CTs. No additional resistor.


The goal of the dual sensing cables is not to add the current from both mains. Instead they are supposed to read the 240v sinewave power the home is using. This is why one ct has to be flipped opposite of the other ct when using separate 120v-120v inverters just like when using the 240v inverters.
My understanding is that the CTs measure amps but send a low voltage reading to the inverter. This low voltage reading is added together when the CTs are wired together in series.
 

copec

New member
Joined
Apr 22, 2020
Messages
16
Is the DB9 port a standard serial port? I have a USB<->Serial adapter plugged into it, and am attempting to read it (first anything I've done with modbus ever), but can't seem to get it to work. I know the serial adapter works because it is working on my MPPSolar inverters.

root@raspberrypi:~# mbpoll -c 125 -o 10 /dev/ttyUSB0 -b 9600 -d 8 -s 1 -P none -R -F -v
debug enabled
Set mode to RTU for serial port
Set device=/dev/ttyUSB0
mbpoll 1.0-0 - FieldTalk(tm) Modbus(R) Master Simulator
Copyright 2015-2019 Pascal JEAN, https://github.com/epsilonrt/mbpoll
This program comes with ABSOLUTELY NO WARRANTY.
This is free software, and you are welcome to redistribute it
under certain conditions; type 'mbpoll -w' for details.

Opening /dev/ttyUSB0 at 9600 bauds (N, 8, 1)
Set response timeout to 10 sec, 0 us
Protocol configuration: Modbus RTU
Slave configuration...: address = [1]
start reference = 1, count = 125
Communication.........: /dev/ttyUSB0, 9600-8N1
t/o 10.00 s, poll rate 1000 ms
Data type.............: 16-bit register, output (holding) register table

-- Polling slave 1... Ctrl-C to stop)
[01][03][00][00][00][7D][85][EB]
Sending request using RTS signal
Waiting for a confirmation...
ERROR Connection timed out: select
Read output (holding) register failed: Connection timed out

I have a feeling it isn't a standard DB9 because I never see any bits back.
 

copec

New member
Joined
Apr 22, 2020
Messages
16
I got it working! While writing a script to try polling with every combination of possible parameter values it occurred to me that in the GUI on the device that there was a setting for device number, so I tried changing it from the default '1' to '2', and now I can communicate with device number 2. (Trying to set it back to 1 confirmed that it wont work when set as device number 1 for some reason, perhaps the mbpoll is an implied device number 1?)
 

lambda

New member
Joined
Dec 12, 2020
Messages
6
Here I have a test program for you, with which you can test the communication to the inverter.
 

Attachments

  • gtil-test.zip
    959 bytes · Views: 215
  • gtil-test.JPG
    gtil-test.JPG
    51.9 KB · Views: 346

lambda

New member
Joined
Dec 12, 2020
Messages
6
Hi,
does anyone have an external limiter in use and could record the signal with an oscilloscope? I have the feeling that this is a one-sided RS485 interface.
If so, we can try to build our own controller with it.
 

watts-on

Member
Joined
May 27, 2017
Messages
181
Hi,
does anyone have an external limiter in use and could record the signal with an oscilloscope? I have the feeling that this is a one-sided RS485 interface.
If so, we can try to build our own controller with it.

I investigated that a while back and wrote about it in this post.
It shows that it isn't any kind of serial data, but more of a Demand/Satisfied signal.

With that info, I then went on to design my own external limiter which I can control myself through RS485/Modbus. I wrote about that Here.
 

lambda

New member
Joined
Dec 12, 2020
Messages
6
I investigated that a while back and wrote about it in this post.
It shows that it isn't any kind of serial data, but more of a Demand/Satisfied signal.

With that info, I then went on to design my own external limiter which I can control myself through RS485/Modbus. I wrote about that Here.
I would like to thank you very much! Your solution works better than expected!
Have a nice Christmas and a happy new year
 

lambda

New member
Joined
Dec 12, 2020
Messages
6
Hi all,

here is a test program for an external limiter.
made with an arduino nano, a 220Ohm resistor and a TA12-100 Sensor (i will remove in a later version)
 

Attachments

  • limiter-test-V1.01.zip
    39 KB · Views: 194

BartDP

New member
Joined
Jan 22, 2021
Messages
12
Hi all,

here is a test program for an external limiter.
made with an arduino nano, a 220Ohm resistor and a TA12-100 Sensor (i will remove in a later version)
Hello Lambda,
Looking to buy 2 pieces 2000W to parallel for singel phase 230V (Europe).
But before buying I like to know if the detected current readable on the display is also available on the communications port.
Additional : is that port a RS232 ( as seems from above ) or a RS485 ( as see on the device itself )
Or did you find the serial port internaly ?
Did you see the current ( from the grid and/or to the grid ) in your readouts ?
Thanks and greeting !
 

lambda

New member
Joined
Dec 12, 2020
Messages
6
Hello Lambda,
Looking to buy 2 pieces 2000W to parallel for singel phase 230V (Europe).
But before buying I like to know if the detected current readable on the display is also available on the communications port.
Additional : is that port a RS232 ( as seems from above ) or a RS485 ( as see on the device itself )
Or did you find the serial port internaly ?
Did you see the current ( from the grid and/or to the grid ) in your readouts ?
Thanks and greeting !
Hi BartDP,

But before buying I like to know if the detected current readable on the display is also available on the communications port.
Unfortunately no,

But before buying I like to know if the detected current readable on the display is also available on the communications port.
Unfortunately no,

The output power is available. but its an average of 5 minutes. It cannot be used for any regulatory system


Additional : is that port a RS232 ( as seems from above ) or a RS485 ( as see on the device itself )
Its a RS232 Port

Or did you find the serial port internaly ?
I use the "Remote" Port outside of the device


Did you see the current ( from the grid and/or to the grid ) in your readouts ?
For this I use my SmartMeter from the grid company

I have now solved the whole thing with an Arduino and a Raspberry.
The Raspberry reads the current from the smartmeter and sends a value to the adruinio

The Arduino now simulate an external signal for the GTIL.

I hope this information can help you a little.

Have a nice weekend

Lambda
 

BartDP

New member
Joined
Jan 22, 2021
Messages
12
Hello Lambda.
Thanks for you fast reply and answer !
I have already a ESP8266 on the P1 of my digital meter, so will need to use the data from there.
Or even redo the current ESP8266 ( used for data logging digital meter ) and integrate the P1 port and its data in a new system integrating the GTIL2's
Kind regards to all.
 

lambda

New member
Joined
Dec 12, 2020
Messages
6
I also have the program and circuit ready that the TA-12 is no longer needed. I upload that as soon as I am at home here.
 

copec

New member
Joined
Apr 22, 2020
Messages
16
Here I have a test program for you, with which you can test the communication to the inverter.
Sorry, I hadn't been paying attention to this forum. I'll try it!

So I've only had these two inverters operating for like nine months and we had one of our 30 second power outages the other day - I live in "the middle of the desert" so I think distribution lines get hit by lightning relatively often - and when the power went out I heard a loud pop from the garage, and not only does the inverter not work anymore, the two-pole polarized breaker was fried.

I ordered a couple of non-polar breakers (they are tied to a battery), but connecting the inverter directly shows it just takes about ~1.5A and shows zero voltage. Even connecting it directly to a panel shows the voltage drop to zero and only ~1.5A, which doesn't make sense because when I shorted the panel to test it, it made 8A, so I think the inverter is bridging an AC waveform that the meter isn't picking up:

The vendor (eco-worthy.com) says there isn't a warranty or repair service, so I'm not sure I want to roll the dice again for a new one. We have fairly new meters so I think I'll try to wire my one remaining with both CTs and see if that works - using the distribution transformer through the meter. I assume the meter will measure a real power usage, otherwise power factor would throw a wrench in metering?
 
Last edited:

copec

New member
Joined
Apr 22, 2020
Messages
16
I figured out how to make one work with both legs with the information above in this thread. I wonder if it is also easier on the 1000W models, since they only have to deal with half the current as at 120V?
 

floydR

Member
Joined
Aug 23, 2017
Messages
1,279
With 120v you are dealing with 1000/120 8.33A same amount of curent 2000/240 8.33A. When the 1000w models are hooked on on both legs of a 120/240 system then the current is reduced 1000/240 4.16A
later floyd
 
Joined
Mar 7, 2018
Messages
1,051
Having had a week of neck deep decoding modbus and RS485 baud for my own MPPT, Juntek meter and inverters (long standing overdue tidying up and fixing from a few years of just hacking to get something working) I had another read through this thread..... and the solution as to how to control the unit is in the protocol document.

I have a SolarRiver inverter which I can also programatically change the maximum output, although it's slow to respond to command changes (seconds) and that involved changing the type of same setting for the GTIL...

Within the registers, take a look at register 40 (Active power adjustment) which is a read write register that takes 0 to 1200 as a divide by 10 representation of a percentage of "target output".

The trick is to make the unit try and output 2000W (attach the current clamp to a load like your computer) with the clamp sensor and the PWM will push the unit to output maximum wattage. Then write to register 40 to control that output between 0 and 100% with an accuracy of 0.1% or in the case of the 2000W the resolution will appear to be 20W steps.

Also, all 3 registers may need setting the same so as not to create a wild unit condition
040 Active power adjustment
041 Reactive power regulation
042 Apparent power adjustment

If someone is able to give this a try it might save a looooooooot of work.....
 
Last edited:
Joined
Mar 7, 2018
Messages
1,051
Oh, forgot to add.... the very principal reason as to why they don't (openly) want to make the unit controlable via a user sending commands is that it actually can invalidate thier grid certification of the product for connection to the grid (it does in the UK as it is a specification as to what parameters are locked from users changing them - read home owners OR the actual people who own the kit, i.e. you)....
 
Joined
Mar 7, 2018
Messages
1,051
Also it would appear that the units may be capable of being programatically switched between wind mode,solar mode or even locked in a narrow output MPPT voltage window for hydro mode...
 
Top