Getting there. Panels on the right are just loose at the moment.
Ronski's Solar & battery DIY build with whole house backup
- Thread starter Ron-ski
- Start date
All panels are on and we are charging the battery.
AC wiring at end of day 1
0.9kWh generated on the new panels.
0.9kWh generated on the new panels.
AC side should be done tomorrow, but they'll still need to change the house consumer unit next week.
Smart meters are being replaced next week as well, finally, but just in time, then need to see if I can get on Flux for the summer at least.
Smart meters are being replaced next week as well, finally, but just in time, then need to see if I can get on Flux for the summer at least.
We've been running off the battery/solar since early afternoon 
A few teething issues, which was a couple of settings in the Victron system, but other than that at the moment its doing what its supposed to.
Electricians done a very neat job, they will be back soon to update the consumer unit in the house. I've still got numerous little bits to sort out.
A few teething issues, which was a couple of settings in the Victron system, but other than that at the moment its doing what its supposed to.
Electricians done a very neat job, they will be back soon to update the consumer unit in the house. I've still got numerous little bits to sort out.
As the weather going to be bad for a few days, I decided to force charge the battery - it needs a full charge to calibrate the battery SOC, just as the Mrs put the kettle on.
5.6kw into the battery, 4.5kw to AC loads.
5.6kw into the battery, 4.5kw to AC loads.
Last edited:
It wasn't displaying battery % because it needed a full charge to calibrate it, it is now, it will also display time left.
Yes I need to setup home assistant, then get that to talk to the Victron system, then send it to PV Output, it will be done, but not just yet.
Yes I need to setup home assistant, then get that to talk to the Victron system, then send it to PV Output, it will be done, but not just yet.
See what?
Open advanced, go to widgets(top right of the screen) chose what widgets you want displayed for each individual component.
(it will give you the total widgets for each device to chose from)
That is where the real magic of the Victron VRM is
No, the real magic is in Node Red
If the above was directed to me I already have loads of widgets setup - you can take a look via the link below in my signature.Here's the VRM with % remaining, and time to go, time to go is based on the last three minutes use, but you can alter that 3 minutes to what you want.
Here's something I've playing with in Node Red, Node Red uses Java script, which unfortunately I've never programmed with, so I need to learn it.
The titles at the top for the panels are not correct, its very much a work in progress.
Node Red runs on the Cerbo GX, which is the heart of the Victron system, and what I'll need to use to setup scheduled discharging.
I've not added up the complete costs yet, and I'm hoping the spreadsheet has some errors in it as it's rather high, and I haven't had the invoice from the electrician either.
Yes , it can see the entire PV, if you look on the advance tab in the VRM there are two graph's, one show each individual string, the other the total generation.
The new consumer unit is just for the garage, the existing fuse box dated from cira 1974
The earth cable running left is to a spike, this is solely for when running off grid, the inverter will switch to this if the grid fails. This is the only bit I'm not keen on, it could have been installed nearer to the other stuff, then the spike/cabling outside would be less visible.
What would you like to see pictures of?
We might require the addition of a 100a MCB, in theory the inverter can pass through 100a and add an additional 34a/8kw which would exceed the cabling and rating of the CU.
The earth cable running left is to a spike, this is solely for when running off grid, the inverter will switch to this if the grid fails. This is the only bit I'm not keen on, it could have been installed nearer to the other stuff, then the spike/cabling outside would be less visible.
What would you like to see pictures of?
We might require the addition of a 100a MCB, in theory the inverter can pass through 100a and add an additional 34a/8kw which would exceed the cabling and rating of the CU.
Last edited:
I'll get a picture tomorrow for you, although that side is not yet finished, electrician needs to come back and change the CU, they said it needs to be up to date as they are altering the supply to it, so I'm having an up to date one with RCBO's fitted.
The electrician did say IIRC that the earth spike can be connected to the DNO earth point, I'm not actually totally sure which way they did - I'll have a look tomorrow.
As per above, I'll need to check exactly how they connected it.
Some pictures for you, note its not totally completed, new consumer unit to be fitted, labelling and some tidying up to be done. Covers removed for taking the picture.
House wiring, by consumer unit. You can probably work it out without me explaining.
Inverter wiring. Grid comes in on AC in 1, which is on the left, supply to consumer units goes out on AC out 1.
Switch gear and earthing. If you follow the earths, you can see that they are all linked together with the ground spike. Spare 25mm SWA is just a bit I had left over, so I put it in the ground and ran it back to the house wall, where in the future it could be used to supply a heat pump.
Emergency by pass switch gear/isolators (lockable)
If the inverter is turned off, the house loses power, so if there was ever a problem with the inverter, or it needed to be worked on, it can easily be bypassed, its just two change over switches linked together.
Some of you may have noticed the fused switch in the first picture was turned off, and this is the grid supply via the meter and DNO fuse, you have to turn it off to take the cover off, so whilst I removed the covers, took the pictures and put them back on we were running completely off grid.
So what happened you may ask? Well, pretty much nothing at first, the power stayed on, the server stayed running.
Then I noticed the SolarEdge PV had stopped generating, but I still had power, so in theory it should have been generating. Checking the frequency there was a small increase to 51.6Hz, not sure how long this lasted as the graphs are only in minute intervals, but the graph shows it was no longer than two minutes, then the frequency settled back to 50Hz and the SolarEdge PV started generating again. This proves that frequency shifting does stop the SolarEdge system from generating, this is very important.
In the extremely rare situation of the grid being off, the batteries being full, and the solar generating, the excess power needs to go somewhere, and that somewhere is the batteries, but eventually the BMS will shut off charging to protect the cells, also at some point the inverter will shut down. To prevent this the system is set to shift the frequency (this can only be done off grid) and shifting the frequency will shut down the SolarEdge PV. I do need to test this actually works, I believe it should happen at 95% state of charge, but now I do at least know the SolarEdge System does shut down. The Victron MPPT's will throttle back when the DC battery voltage reaches a set point, so these arn't an issue, just AC connected PV.
The other thing I noticed is that my server UPS went onto battery, probably due to the frequency increase, so need to check the settings there.
Best I get on with some work.
House wiring, by consumer unit. You can probably work it out without me explaining.
Inverter wiring. Grid comes in on AC in 1, which is on the left, supply to consumer units goes out on AC out 1.
Switch gear and earthing. If you follow the earths, you can see that they are all linked together with the ground spike. Spare 25mm SWA is just a bit I had left over, so I put it in the ground and ran it back to the house wall, where in the future it could be used to supply a heat pump.
Emergency by pass switch gear/isolators (lockable)
If the inverter is turned off, the house loses power, so if there was ever a problem with the inverter, or it needed to be worked on, it can easily be bypassed, its just two change over switches linked together.
Some of you may have noticed the fused switch in the first picture was turned off, and this is the grid supply via the meter and DNO fuse, you have to turn it off to take the cover off, so whilst I removed the covers, took the pictures and put them back on we were running completely off grid.
So what happened you may ask? Well, pretty much nothing at first, the power stayed on, the server stayed running.
Then I noticed the SolarEdge PV had stopped generating, but I still had power, so in theory it should have been generating. Checking the frequency there was a small increase to 51.6Hz, not sure how long this lasted as the graphs are only in minute intervals, but the graph shows it was no longer than two minutes, then the frequency settled back to 50Hz and the SolarEdge PV started generating again. This proves that frequency shifting does stop the SolarEdge system from generating, this is very important.
In the extremely rare situation of the grid being off, the batteries being full, and the solar generating, the excess power needs to go somewhere, and that somewhere is the batteries, but eventually the BMS will shut off charging to protect the cells, also at some point the inverter will shut down. To prevent this the system is set to shift the frequency (this can only be done off grid) and shifting the frequency will shut down the SolarEdge PV. I do need to test this actually works, I believe it should happen at 95% state of charge, but now I do at least know the SolarEdge System does shut down. The Victron MPPT's will throttle back when the DC battery voltage reaches a set point, so these arn't an issue, just AC connected PV.
The other thing I noticed is that my server UPS went onto battery, probably due to the frequency increase, so need to check the settings there.
Best I get on with some work.
Last edited:
Yep, got an alarm on my phone pretty instantaneously, which I would get any where in the world so long as both ends have a data connection.
I will add up the costs, but it is going to come in very expensive, but still less than some quotes I've seen on here for far less capable systems. One benefit of it taking so long is I've actually been able to pay for it with only using disposable income, and some from the Mrs.
I now have 8.74kw of solar, an addition of 4.74kw, 14.5kWh battery, but that is if I used the full capacity from 3.65v down to 2.5v, I'm going to be using the 2.9 to 3.45v range, but capacity drops off very quickly either side of those voltages. I will add a second battery, as the system really needs it, and the extra capacity will be useful.
I will add up the costs, but it is going to come in very expensive, but still less than some quotes I've seen on here for far less capable systems. One benefit of it taking so long is I've actually been able to pay for it with only using disposable income, and some from the Mrs.
I now have 8.74kw of solar, an addition of 4.74kw, 14.5kWh battery, but that is if I used the full capacity from 3.65v down to 2.5v, I'm going to be using the 2.9 to 3.45v range, but capacity drops off very quickly either side of those voltages. I will add a second battery, as the system really needs it, and the extra capacity will be useful.
@HungryHippos think I misread your previous reply, there is no real lag when going off grid, Victron state the following.
The Quattro can charge the battery at up to 140A (48v DC side) from the grid, great for those short periods of cheap rate electric. At 09:50 this morning I was seeing 86A going in to the battery just from PV. 140 amps is 0.5C which stresses the cells more, having two batteries would half the current to each, same goes with discharge, the Quattro can pull 160A at full capacity. Spreading this load across two batteries will hopefully mean they last much longer.
The Quattro can charge the battery at up to 140A (48v DC side) from the grid, great for those short periods of cheap rate electric. At 09:50 this morning I was seeing 86A going in to the battery just from PV. 140 amps is 0.5C which stresses the cells more, having two batteries would half the current to each, same goes with discharge, the Quattro can pull 160A at full capacity. Spreading this load across two batteries will hopefully mean they last much longer.
Was just looking at the VRM (its rather addictive whilst its all so new), and AC loads went to over 6000w, grid went up a little to about 240w then immediately settled down to around zero. Mrs had just put the oven and kettle on at the same time, apart from the slight fluctuation it was all supplied via the inverter
@katie279 I noticed today that the PV Power graph had 3 Y axis scales, which is why the three arrays seemed to be very close together on the graph, I've now set it to a fixed scale for all three arrays.
SolarEdge system produced 13.07kWh
East Garage Array produced 5.65kWh
West garage array produced 4.42kWh
So on a sunny day but still some clouding the 4kw SolarEdge system beat the 4.74Kw Victron system by 3kWh, on a cloudy overcast day the Victron system does better.
SolarEdge system produced 13.07kWh
East Garage Array produced 5.65kWh
West garage array produced 4.42kWh
So on a sunny day but still some clouding the 4kw SolarEdge system beat the 4.74Kw Victron system by 3kWh, on a cloudy overcast day the Victron system does better.
@Journey message sent.
Smart meters finally changed
It's all coming together.
It's all coming together.
When I add it up, but the electricians bill came to £2.5k, the bypass switches they used were £300 each 
It's gone well over what I thought it would be, but it's only money.
I've ordered 16 more cells, cell busbars and BMS for the second battery. At least the cells were £500 cheaper this time round, they are expected late April.
It's gone well over what I thought it would be, but it's only money.I've ordered 16 more cells, cell busbars and BMS for the second battery. At least the cells were £500 cheaper this time round, they are expected late April.