Thread Rating:
  • 1 Vote(s) - 5 Average
  • 1
  • 2
  • 3
  • 4
  • 5
External limiting an GTIL2 SUN-1000G2 with raspberry
(08-05-2019, 12:24 PM)watts-on Wrote:
(08-04-2019, 06:57 PM)ZARK Wrote: Hello all. Having read the entire tread I am still lost concerning remote monitoring of  this GTIL inverter. I have a GTIL2 1000 watt inverter ver 5.1 with the DB9 connector for remote monitoring.
however i do not have the wifi plug unit. I want to remote monitor my unit localy and not go to the chinese cloud. It seems some have done this here in the thread. I would like some help to do the seem please :-). First of, which standard/protocal does it use rs485 or rs232 with modbus. The manuel seems to imply 485. Or can it use both. The pin out for the DB9 connector on the unit would be helpful. I have a wifi RTU unit (model USR W630) that i want to use to connect it to my network and then monitor using openhab2 with the modbus binding. I cant get it work :-( any help Please. Thanks

It's a while since I dabbled with this, but here is some info and images I created at the time if it helps:

The interface is RS232:

The essential functions of the interface board:

The 8 pins on the left represent the ribbon cable from the main board.  The 9 pins on the right represent the RS232 D connector.  The centre pins are the connections to the RS232 level shifter IC (charge pump etc components not shown for clarity).
I'm not sure what the connection to pin 9 on the D connector is, but I suspect it is to provide power to a wifi type dongle, so watch out for that if what you connect to it is not expecting that as that is not standard on an RS232 connector.

Some screen shots of settings I used in a modbus application:

Thanks very much for your responce Smile . You have been very helpful. The pin out diagram was very useful. I have now manage to connect and communicate with the inverter Cool , although I cant make much sense of the data. I was successful with a modbus tcp client. I want to monitor the inverter using openhab2.  The modbus binding for openhab support tcp client anyone knows how to set this up. had no luck thus far Confused
I have 2 2000w gtil2s..they are both connected to one set of ct clamp sensors and one of them is great, no backfeed.
Do u guys think I can add a second set of clamps and have original set of clamps just go to one inverter then second set to the second inverter..
Reason I want to do this is so that I have permanent master and slave inverters. I’m thinking that I’ll put both sets of sensor clamps far apart from one another so that one set of clamps will see-send the signal first to the master inverter, then the second set of clamps will only send to the slave the signal for what the master can’t provide? Will this work...thoughts plz
Hi, guys-
Has anybody (like tocan or cherry67) found out in conversations with the factory people in China who actually makes the daughterboard (called the dc controller board) that has the connectors on it for external inputs to control the inverter output power? The inverter main board is made by Deye, but reading thru this thread makes me wonder about the dc controller board. Seems things would be a lot easier if we could order a bunch of these with hacked firmware to suit our purposes. Ideally, getting hold of the guy who actually wrote the code initially and paying him would be quicker that massaging the code with the inherent risks of overlooking a line of code here or there with possible disasterous consequences for the fets, igbts, and scrs in the power handling sections of the inverter.

The pic on the daughterboard is a Microchip dsPIC33F four channel controller which manages the inverter input boost smps (4 fets for a GTIL1 and 8 fets for a GTIL2). Seems the original developer could whip out some new code to do what we want if somebody could find the guy. A boo boo in the code could fry all of them.

Just wondering. Shout out to Wolf, we have met before
Using 2 240v inverters and one set of cts in series then paralleled to each inverter let’s the grid supply more power at minimum than only having one 240v inverter. Just fyi
(08-18-2019, 03:35 AM)Doin it Wrote: Wolf.....Seems like u are right,,,the 240v inverters do backfeed if loads are unbalanced. .....................the 120v inverters connected pv can only supply loads on one leg.. so that’s a lot of wasted pv energy.......................
Ouch that's gotta hurt. Tongue

In my case the "wasted" PV energy charges the batteries so not much loss there as the batteries will supply both the 120v GTIL after the sun goes down till the voltage cutoff. Also in the AM the batteries get charged before the GTILs reboot voltage is reached.

Solardad likes this post
For Info Google Drive

Great ABB shunt breaker seller River City Industrial
Great prices and superb service.
 [Image: em2566%20(1)__19172.1521541365.jpg?c=2]

Not your average Wolf       
sorry i an new here.
Wolf and anyone supplying a split phase main panel with a single phase 240v limiting inverter... the pic (the pic using total power) is actually showing what the 240v inverters supply and how electric company reads our usage using 240v single phase inverters..the pic using main1 and main2 is incorrect and not how ecompany meter reads our usage.....settings in iotawatt I have allow negative values selected as I should but I am also combining main1 and main2 (each leg) cuz that’s how the ecompany meter nets our power usage...I got my iotawatt running and I got some results after a fairly lengthy discussion with the engineer of graphs when I measure each leg separately at 120v and graph the phases separately, they look the same as wolfs did with overfeed-backfeed and underfeed..But this isn’t what the ecompany meter sees..The ecompany meter will only see backfeed if there Isn’t an underfeed on the opposite leg, with a 240v limiting inverter it will have an underfeed aswell as an overfeed so they net to zero..The ecompany meter actually equals out the backfeed and underfeed on the graphs (at 120v on each leg to equal 240v overall like ecompany meter is reading) to a net zero..The standard home meter (same meter I have) most ppl in the USA have reads this way..Most ppl when they go grid tie actually get what’s called a “net” meter from the electric company and that meter doesn’t charge the consumer for backfeed (consumer gets paid for backfeed) but again it still only senses backfeed if it’s happening on both legs. But if u oversupply one leg by 80% more (using a 240v single phase limiting inverter) than the other leg then ecompany meter will read error and most likely backfeed or most likely would get a visit from electric company..

My test confirms this, although I havnt accumulated many kwh, I still would see the ecompany meter saying I use more than what the iotawatt says I’m using if the ecompany meter actually reads the backfeed as power we pay for.. but it’s not, if anything iotawatt is saying I’m using slightly more or I could say the same (only .8kwh difference) as ecompany meter .
Test started with ecompany meter: 27252
Test ended with ecompany meter: 27310
Equals: 58 kwh used
Test started with iotawatt reading: 59.6
Test ended with iotawatt reading: 118.4
Equals: 58.8 kwh used
When using 120v limiting inverters without the special grid tie ecompany meter installed, if there is ANY backfeed on its leg it is wired to (which there shouldn’t be using limiters) it will be seen by electric company as backfeed therefor getting a visit from them..the reason for this is that if one leg would be backfeeding the other leg isn’t being underfed to even out the net usage like the ecompany meter does...using 240v inverters (that supply the sum of both legs) it will appear to backfeed and underfeed when reading each leg separately, but it’s not read that way by the 240v ecompany meter cuz the other leg is being underfed by the same amount it’s being overfed so it reads net zero.. The common ecompany meter on most houses reads at 240v and in order for that meter to read at 240v (no neutral in the meter) it reads the balance of current between each phase..this is why the ecompany meter will still see backfeed if it’s happening on both legs cuz the push and pull balance isn’t equaling out if power is flowing the same backwards backfeed direction on both legs..
Good morning everyone. 

I can't say how much I appreciate all the leg work and research you all have put into controlling the GTIL2 via serial / modbus. 

I have (5) GTIL2 (version 2) that I am looking to control via modbus. Building myself a peak shaver with about 22Kwh (today) and easily scale-able to 33Kwh next year by adding a few modules. 

I am going to try and connect to one of the inverts via RS-232 using the pinout watts-on provided only thing is that it looks like my inverter has a slightly newer version of the DB9 control board. 


Doin it and Patalhetas like this post
Like the rest of you, I am trying to hack the DC Controller Board for the GTILs. I think somebody posted early in this thread about speaking to the developer of the board. I have attached two photos of the board showing the logo of the actual manufacturers. If somebody can identify the logos, then contacting the real manufacturer/programmer would be a lot easier and maybe would give everybody access to the schematic and code, so we could hack it at will. Also, I have a list of the components (ICs) used on the board if anybody wants me to post them.

The IC U1 is the PIC that runs the whole board, a Microchip PIC33FJ. The datasheet for the chip says that it runs C and Assembly in this application. I bought a Microchip MPLAB programmer/debugger and am going to learn how to use MPLAB IDE so that I can figure this thing out (maybe). Big Grin
Patalhetas likes this post

Forum Jump:

Users browsing this thread: 2 Guest(s)