For example, one route to take:
Another Route to take seeing it is not an official installation:
Has a Venus built in … and a MPPT.
Options options …
If you have a Venus/Cerbo:
Use a standard network cable to connect inverter to the Venus/Cerbo.
Need a VE.Direct cable for MPPT’s.
Has the ports already for these cables. No USB required.
If you have a Rpi:
Need a MK3 to USB cable for the inverter connection.
Need a VE.Direct to USB cable for MPPT’s - you don’t need it yet.
Need that cause Rpi has only 1 UTP (for network) and USB ports. So you “convert” the cables with the USB addition to them.
Future:
If you connect to the main DB, you need a Carlo Gavazzi ET112 with its USB cable to connect to the Venus/Cerbo, or the Rpi. No need for one now.
1 Like
The Hubble AM-5 is a 16 cell LFP battery. You should be able to wire it in parallel with any other 16 cell (51.2V nominal) LFP battery, with the caveat that there won’t be any communication with the additional battery. Some batteries have an LCD display interface that you can use to check up on the health of the battery, so lack of communication with the inverter is not always a deal-breaker.
There are many cheap options available. A quick search revealed these:
(Not endorsing the company, just pointing out that you have other options if you don’t want to buy another Hubble.)
Nothing wrong with tinkering - if not for tinkering @TheTerribleTriplet wouldn’t have found out you can use a no12 spanner for arc welding… 
4 Likes
The battery would be wired in parallel with the existing battery. I can’t advise on cable gauge, either between batteries or between battery bank and inverter, you should ask about this.
Then the usual thing is to have one battery set up as the “master”. This will be the existing battery and it will continue to communicate with the inverter. The “slave” has a comms cable to the master.
This makes sense when taken with the rest of what you say here. The system produces a lot more than it can store.
How much grid do you use during the day? In kWh, not rands.
I ask because I’m tellimg you how nicely I get through the night, but I always draw some grid power in a day. Unless one goes completely off grid, as in has no physical connection, it seems to be very difficult to get to zero.
In my opinion, when starting from scratch, with the price of the Cerbo GX or S GX these days, it is hardly worth going the PI route.
PI 4 plus proper power supply plus SD card plus RS485 HAT to talk to the battery BMS plus the Victron cabling, it quickly adds up.
Total different story of course if you already have the PI plus accessories lying around and also some of the Victron cables maybe.
2 Likes
If you draw 10W for 24 hours… that is 240Wh, or 0.24kWh. A quarter of a unit almost, and noticeable by anyone who watches their system with a magnifying glass 
1 Like
Found this again yesterday …
Damn dit was seer …
Now I know so much better … the right way of doing it. 
Have not blown “any shiite up” the last 12 months … at least! 
2 Likes
Yup, parallel as you explained. I wrote in far too short 
During the day grid use is about 6 kWh - the remaining variation of 1-4 kWh comes from the night use. In the 6 kWh I’m also including the base sync load with the grid which is a total of 0.5 kWh for the day.
Of course. I don’t want/need to get to zero all the time. It is for the 95% of the loads. The odd peaks in usage, etc. I am not bothered about. I don’t even have the option to go full off grid anyway.
This is what I was reminded about. It is even stricter. It is Hubble Am-5 with another AM-5. Not even another 51.2V battery they might have. It is not that it CAN not be put in parallel. It just should not be to keep the warranty.
I can only wish. I do about 2 a day on average.
1 Like
Well that’s not very lekker. That’s 7 to 10 kWh per day from the grid. What’s your total use?
Sorry I wasn’t clear. What I’ve mentioned is consumption. The from grid is also about 1-3 kWh. Basically looks like the evening load.
This is why I am not enthusiastic about spending money towards this (but yes, you should be able to put something together that can put some unused PV into a battery).
Even though I have various restraining orders against me for mathing without comprehension my sums suggest this.
Using 5kWh daily grid consumption over 18 months, including a 30% tariff increase for 12 of the 18 months, and the additional 10% increase in consumption (falling in the increased tariff period) you will pay ~R11 500 for the consumed electricity.
If using the battery you already have (rewired as 48V or not), it will only allow you to avoid direct grid consumption of about 2kWh per day. Assuming it works perfectly every day, and using the same tariff increase scenario above, over 18 months you will avoid ~R4150 in direct grid consumption.
So, unless you are very keen to build a UPS/trolley that you can take along when you move you need to cap your expense at ~R4000 for the trolley option and that seems unlikely. Alternatively, you need to come in under R11k to end up with just a battery to take along when you leave.
OK. So now I’m thinking you’re going to spend good money for a not very great return.
Yes, this looks increasingly likely. And certain if I go blue Only makes sense to go ahead with this, if the grid avoidance is more than 4 years I think. Last I checked my math.
Yes, the sums are a tight one.
Depends on your long-term plan.
Me, I look at it as “school fees”. Goes for anything solar. Any brand.
Mistakes will be made. Misunderstandings. Perfectly normal.
If you shop right, and get it in as affordable as possible, the learning curve just from that, will save you a ton into the future if you want to mitigate Eskom costs going forward.
You can pay someone to do it for you, ROI takes longer.
Or you can do the effort yourself, the ROI is shorter.
But that it is going to cost, that is a fact.
Here, on this forum, we have a HUGE repository of “WTF Not to Do” … and what works best long-term.