I presently have 14x 435W Canadian Solar panels on my roof, wired in 2 series strings of 7. Panels have a Voc (NMOT) of 45.7V and Isc (NMOT) of 9.15A, meaning that each string has a maximum voltage of 7x45.7 = 319.9V. 

My inverter (Axpert 7.2kW MAX) has two MPPT chargers (4000W each). At present, I have connected one string to each controller, so I am well within the maximum and minimum voltages of the unit.

Now, I would like to add an additional 2kW of solar panels onto a different roof of my house, but I cannot get the identical panels. I was planning to add 4x panels, and wire them in a 2S2P configuration, and then add one string to each of the existing strings. However, this will then cause a voltage issue as the new strings will only generate around 100v per string. 

I don't really want to run additional cables to the other panels in order to split them into strings of similar voltages, so I was thinking of purchasing a charge controller such as a Victron 100/20 in order to accommodate the four additional panels. Could I then just wire the charge controller directly to the batteries (48V), so that these 4 panels just charge the batteries and the inverter continues to supply the loads etc? Any issue with both the inverter and the Victron controller charging the batteries?

*Edit:

I'm presently running ICC on a Raspberry Pi, and I noticed that ICC can include the information directly from the Victron MPPT, via the VE USB cable. Would it be as simple as connecting the Victron, as per the above setup, and then connect it to the Pi so that ICC can update values, such as PV production etc?

Edited December 6, 20223 yr by Schnavel
Additional Info

23 minutes ago, Schnavel said:

purchasing a charge controller such as a Victron 100/20 in order to accommodate the four additional panels. Could I then just wire the charge controller directly to the batteries

I would think that should be fine, you just would probably have to ensure that both this charge controller and the inverter use the same voltages for bulk/absorb and float...

46 minutes ago, Schnavel said:

I don't really want to run additional cables to the other panels in order to split them into strings of similar voltages, so I was thinking of purchasing a charge controller such as a Victron 100/20 in order to accommodate the four additional panels. Could I then just wire the charge controller directly to the batteries (48V), so that these 4 panels just charge the batteries and the inverter continues to supply the loads etc? Any issue with both the inverter and the Victron controller charging the batteries?

*Edit:

I'm presently running ICC on a Raspberry Pi, and I noticed that ICC can include the information directly from the Victron MPPT, via the VE USB cable. Would it be as simple as connecting the Victron, as per the above setup, and then connect it to the Pi so that ICC can update values, such as PV production etc?

Yup - did that in my old system setup. Was running 2x Axperts (older models) with no panels connected and a Victron 250/100 MPPT. The batteries, the Victron, and the battery input of the Axpert all just need to be connected to a busbar. Then the Victron will charge the batteries and feed the inverter during the day.

With the VE USB cable you can then get ICC to display the PV production from the Victron (a setting somewhere)

15 hours ago, Kalahari Meerkat said:

I would think that should be fine, you just would probably have to ensure that both this charge controller and the inverter use the same voltages for bulk/absorb and float...

Definitely, I would need to set the maximum bulk charging voltage to the same for both in order to not over charge the Pylontech batteries.

My concern was more around the safety aspect of having both the inverter and Victron charging the batteries - but I guess in this case, it would be like having 3x MPPT chargers all charging the batteries (the inverter has 2).

Edited December 7, 20223 yr by Schnavel

17 minutes ago, Schnavel said:

My concern was more around the safety aspect of having both the inverter and Victron charging the batteries - but I guess in this case, it would be like having 3x MPPT chargers all charging the batteries (the inverter has 2).

Yes,this is why you add the Busbar as a central "delivery" location for all charging devices and to-be-charged devices rather than direct connecting 20 different pairs of wires direct to the battery

Edited December 7, 20223 yr by PsyWulf

Just keep in mind: the Victron 100/20 will only be able to charge at 20A*48V(battery voltage) so almost 1kW.

1 hour ago, Schnavel said:

Definitely, I would need to set the maximum bulk charging voltage to the same for both in order to not over charge the Pylontech batteries.

I would recommend setting the voltage on one of your devices a bit lower - probably on the Victron as this has the least PV connected to it. That way they don't 'interfere' with each other when they change over to float

1 hour ago, Schnavel said:

Definitely, I would need to set the maximum bulk charging voltage to the same for both in order to not over charge the Pylontech batteries.

I notice you are not working with Voc from the STC for your panels. If as has been suggested to connect 2S to a Victron 100/20 you don't have much space for the Voc that can be expected at 0-10 deg on cold days. Thus you will be running at risk but may be it is just me that would not want to get close to Voc at low temps. 

1 hour ago, P1000 said:

Just keep in mind: the Victron 100/20 will only be able to charge at 20A*48V(battery voltage) so almost 1kW.

Ah, that's a very valid point that I hadn't considered - I was planning under 100v at under 20A on the PV side, but then as you mention the 48V "step down" is a problem.

Thanks everyone for the input. I will need to relook at costing and the exact panels I want to buy and then see how best to do the installation. It may make more sense to split up the existing panels and run additional cables in order to use the existing inverter MPPT controllers, as opposed to adding a separate charge controller.

6 minutes ago, Schnavel said:

It may make more sense to split up the existing panels and run additional cables in order to use the existing inverter MPPT controllers, as opposed to adding a separate charge controller.

I would definitely expect this to me more cost-efficient.

Not sure what voltage your MPPT can handle, but maybe see if you can squeeze more of your existing panels onto 1 string. And then put the rest of the panels together with the new ones onto the 2nd string

18 hours ago, Schnavel said:

I would like to add an additional 2kW of solar panels onto a different roof of my house

What distance between your current panels and your planned installation location are we talking about?

18 hours ago, Schnavel said:

I cannot get the identical panels.

I know that challenge 😶 Have you tried to see if you can get panels with similar specs? Mixing panels is not easy but also not impossible

17 minutes ago, wolfandy said:

Mixing panels is not easy but also not impossible

when trying to mix on a serial basis, all panels should produce the same current, the Voltage is not relevant other than calculating the total Voc and Vmp...

when try to mix on a parallel basis, all strings should provide close to the same Voc and Vmp, then the current is not relevant other than calculating what power you can expect from the panels in total...

same current +/- 5%, I'd think would make any losses close to irrelevant and the same goes for Voltage differences..

9 minutes ago, Kalahari Meerkat said:

same current +/- 5%, I'd think would make any losses close to irrelevant and the same goes for Voltage differences..

Indeed, it would be necessary to source similar spec panels - will have to have a look what is out there and see if I can make it work.

I would need to either add additional panels to the existing strings in series, which will be fun from a wiring perspective (wire routing will be around 50m), or add an additional string of panels to bring the voltage up to similar values as the existing string, but would need to mindful of the MPPT operating criteria (90V - 450V)

2 hours ago, Schnavel said:

Indeed, it would be necessary to source similar spec panels - will have to have a look what is out there and see if I can make it work.

I would need to either add additional panels to the existing strings in series, which will be fun from a wiring perspective (wire routing will be around 50m), or add an additional string of panels to bring the voltage up to similar values as the existing string, but would need to mindful of the MPPT operating criteria (90V - 450V)

Your next bet is Togo for a 150/20 Victron. Not sure if you get one otherwise the 35A version. This is costly if you want to use Victron. Other good brands are a lot cheaper and can meet the 150V Voc spec.