Niel

Yes. It can limit both export feed and supply (essentials+non essentials) feed.
Meaning if you buy 16kw you can limit it, to make it do whatever you want it to supply (2kw/3kw,10kw,etc)
Hope it makes sense

@Neilthank you so much for the valuable and prompt response. I stand by my reference of you, but the correct English should have been  'Valuable Human Resource.'👍
@Ronald Ware you on track, or do need additional info?
 
Kind regards.

Ha ha ha , @87 Dreamyou beat me in this race by 1 minute 😁 was just about to mention @Niel as usefull resource regarding this matter.

Solcast is what you want if you want to ‘predict’ potential PV for the day. You seed their algorithm every 5 minutes with your PV production over that period and based on that, plus your location and panel angle/direction, it starts generating forecast after a week or two. Another user put me onto it and it seems reasonably accurate so far. I’m using Home Assistant though, not sure if this is available for OpenHab

That really depends on your coding language you use, but have a look at combing two bytes into a word. In C it’s easy, you just bit shift the high address with 4 bits and add the low address (hbyte <<4 + lbyte). That’s the most efficient way of doing bitwise operations, but it depends, like I said, on the high level language you using.

Bit shift by 8, but it is a big endian encoding so the most significant byte is on the left.
Plus be aware some values need some (sign) conversion magic "if value > 32768 then value = value - 65535".
Look at the python sources, they are pretty self descriptive.
 

Some good news.  It's taking a bit of time but I have my first data from the Goodwe coming into OH!
 

Some good news.  It's taking a bit of time but I have my first data from the Goodwe coming into OH!
 

It seems various inverter models and/or firmware versions slightly differ in the response data (resp. their lengths).
So I've replaced the primitive response length check with proper checksum and response type validation.
Now it should be more robust in respect to different inverters.

Yebo, you can’t use wire shark that way. While UDP packets are broadcast, your switch prevents those packets from flooding the network. The switch is routing the relevant packets directly between your phone and Goodwe dongle. It’s not available on the rest of the network. You’ll have to run a packet sniffer on the phone with the PV master app to get the relevant packets. That’s how I got the data initially.
Head over to the GitHub page listed above. The integration has come a long way over the last week. Unfortunately it’s directly integrated into HA at the moment. Ideally it should be done through MQTT, but I don’t think it’s on the roadmap for the moment.
‘You might want to consider trying HA. I prefer it over openhab anyway.

So I am ready to take the next step in my off-grid/on-grid solar adventure.
You can view my 1st post here
Now to start measuring all this eskom power solar battery stuff.
I started by looking at all the inverters and the monitoring software they come with.
Mixed results on that, some work ok, some are total crap.

Solution - D.I.Y. seems to be the way to go on this too.
I have looked at some of the forum posts here and also again a load of youtube videos and power forums.
My conclusion is that it is 100% better to D.I.Y. and I have settled on the software, hardware etc on doing this.
My first setup is now completed and I can start measuring power consumption from various sensors and get a visual representation of the cost of power from eskom.
For now I have started off just measuring eskom main power usage, will be expanding this out from here to start measuring individual circuits and then start bringing my solar production and battery measurements into the mix.

Currently the stats above come from the following hardware.
1- Raspberry Pi4 4gig as the main influxdb and grafana server.
2 - Pi Zero W at the main eskom db to do usage measurements
3 - RPIZ_CT3V1 - voltage and current sensor board (attached to pi zero W)

This initial setup will be further expanded to add sensors to all the main home circuits individually (Lights, plugs, geyser, stove etc)
Extra hardware will be an expanded ac current sensor that can accommodate all the circuits and the extra sensor hardware to measure the solar and battery side.
To date all this has cost less that the retail equivalents out there (sonoff effergy eyedro vue etc)and I have more precise control over exactly what and how I can measure everything.
And of course I can visualize everything exactly how I like using grafana.
If you have any tips or suggestions then do let me know and share your sensoring/monitoring experiences with the community so we can all learn from each other

Ok, I'm kinda following what you guys are trying to do and will love it if it works.  But I'm still wondering whether it's possible to hack the PV Master app.  It updates a lot more often than once a minute and somehow gets all the data from the inverter that I'm looking for so the solution must be in there somewhere.
 

see attached report. @Rautenk has given me the permission to publish it
 
 
Irradiance report001.pdf

I guys. I would like to share my schematics and source code for a small project that allows to control/monitor and intergate Pylontech batteries with home automation via WiFi.
Here is a video of this in action:
https://youtu.be/7VyQjKU3MsU

Schemetics and source code:
https://github.com/irekzielinski/Pylontech-Battery-Monitoring/

Hope you find it useful
Cheers,
Irek

I've done firmware upgrades a few times in the past, some successful, others basically turned the inverter into a blue doorstop.  But Goodwe's support via email has been very good and they really keep trying to help until the problem is solved.
The Ezconverter solved our battery communication problem and everything is working fine.   I still have the battery as "self-defined" as the US3000 is not an option but ultimately I don't think it really matters.  The Pylontech communicates it's SoC and charge/discharge capacity to the Goodwe and that is really all that is really important.
As far as I know, the SN 9 models are the newer "ARM" models which should be able to communicate directly with the Pylontech.  If not, I'd suggest a firmware update.  Failing that an EzConverter.  I don't think a US2000 will communicate with the Goodwe any easier than the US3000.
 

At a product training @ Rubicon, Goodwe said their SA import partners is Sinetech and Rubicon, but yes Segen also offers them. As far as i understand Segen imports them via their international connections (with very good prices i should add  )

It's almost a week and everything seems to be working fine.  It is however interesting to see how the BMS manages the batteries.  When the SoC is under 50% (approx), the charge current is limited to 29A (that is for both US3000's).  As the SoC increases, this increases to the max of 76A.  I'm sure this is to prolong the battery life, probably charging too quickly when depleted causes damage/ wear.  It does however mean that they don't charge at full capacity in the morning and significantly quicker in the afternoon.  So if you have a NE and NW facing roof and plan to use Pylontech's, place more panels on the NW and less on the NE.  And/ or move some more load to the morning and less to the afternoon.
 

It's almost a week and everything seems to be working fine.  It is however interesting to see how the BMS manages the batteries.  When the SoC is under 50% (approx), the charge current is limited to 29A (that is for both US3000's).  As the SoC increases, this increases to the max of 76A.  I'm sure this is to prolong the battery life, probably charging too quickly when depleted causes damage/ wear.  It does however mean that they don't charge at full capacity in the morning and significantly quicker in the afternoon.  So if you have a NE and NW facing roof and plan to use Pylontech's, place more panels on the NW and less on the NE.  And/ or move some more load to the morning and less to the afternoon.
 

a while ago we discussed if the CoCT is following the inverter maximum output limits as published in NRS097-2-1 2014. Unfortunately they are not (see correspondence with the CoCT below). The limit in CT for a 60A single phase supply is indeed only 3.5kVA and not 3.68kVA as in NRS. This is restricting suitable inverters even further. A lot of suppliers (like Goodwe) only have 3.6kW models ☹️.
 
 
Good day,



Our requirements is modified to that of NRS, thus 3.5 kVA will be the limit for SSEG installed for a 60A single-phase installation. 
Regards

Shiraaz Swartland (Pr Tech Eng)
Principal Professional Officer: Service Connection Planning (SCP) – North

Tel: 021 444 2177 | Fax: 086 696 6068 | Cell: 083 324 2908 |  Email:      [email protected]  | Web: www.capetown.gov.za


 



 
 

 
Dear Mr. Swartland,

thank you for the reply. We will follow up with Goodwe in China to find a suitable solution for their Inverters > 3.5kVA connected to the grid in Cape Town. Another issue we would like to clarify to prevent any future registration problem is the exact maximum total generation capacity. According to NRS 097-2-3:2014 it is 3.68 kVA on a 60A single phase supply but your table refers to 3.5kVA. Is the CoCT prepared to accept inverters which a total maximum output capacity of 3.6kW?

Regards

 
 
 

And I finally got it to work.  It turns out the little switches on the Ezconverter (0 to 5) was incorrect.  The communication is working and I see the charge/ discharge current limit was updated.
The battery is still set up as "self-defined" Li ion.  I'm not sure what to change this to as the US3000 is given as an option and I might just leave it as is for the time being.
A thank you to Segen, they were very helpful.  If anyone wants more info, I now have some experience on the Ezconverter!
 

And I finally got it to work.  It turns out the little switches on the Ezconverter (0 to 5) was incorrect.  The communication is working and I see the charge/ discharge current limit was updated.
The battery is still set up as "self-defined" Li ion.  I'm not sure what to change this to as the US3000 is given as an option and I might just leave it as is for the time being.
A thank you to Segen, they were very helpful.  If anyone wants more info, I now have some experience on the Ezconverter!
 

Thou shalt not use an inverter not created in the name of NRS-097 unless you be smitten by the fist of COCT.
Clear ?

Thats pretty much what I did, I swopped out my 4kw element with a 1kw element and put my geyser on a time switch to only turn on during daylight hours.
The water takes longer to warm up but is warm by the time I need it, works fine for me

I had independent elements make me up a 1kw element: https://independentelements.co.za
I supplied them with my existing element which was dead and they used the flange and base and fitted a 1kw element to it.
I guess they could make the element any size you like
I think they charged about R500

Hi all,
I was looking at this chart from yesterday and I thought I'd share it since it shows how nicely my hybrid grid-tied setup is working.  Shows the advantage of the grid-tie approach even without council feed-in.
 

 
The blue is the power coming from the panels.  The green is shortfall that is being drawn from the grid.  The purple is power going into (-ve) or out of the batteries.
The brown shows how much power the inverter is pushing back to the grid side, and the red line is how much is going to the "essentials" circuits.
Since the inverter was grid-tied it essentially feeds the grid and the essentials together.
You can see that power from the panels goes first to the essentials and the grid side, trying to reduce draw from the municipality down to zero.
If local demand is met and there is still excess from the panels then it goes to charge the batteries.
If we don't have enough local power from the panels then what I do make is still 100% used and the nett difference drawn from the municipality.
When the sun is gone then power is drawn from the batteries so the excess from the day is used in the evening.
This was a high usage day - hot water, electric oven and etc - but in the middle of Cape Town winter I managed to use up 13.7kWh from my 3150Wp setup - looking at this chart none of the sunshine went to waste.
I'm going guesstimate that that is twice the yield I could have got from a non-grid-tied setup.