Mixing and matching panels (with Solis AC Inverter)

Charge Controllers (MPPT, PWM)
111

This is an interesting approach, but one I won’t take. You would have to remember exactly what you did 5 or 10 years down the line if for any reason you need to disconnect or remove any one of the multiple units.

The CT should be open circuit the moment you disconnect the faulty unit and 10 to 1 the smoke will exit the CT rendering all the other units useless till you return from the electrical wholesaler with a new ct.

Forgetting to remove the CT from the live wire with a CT per inverter setup will still send you off to the wholesaler, but without affecting the other units.

Me and you might remember to go remove the CT before disconnecting the wires, but a DIY enthusiast might not…

112

This is precisely why I have advised against using standard CTs and rather modified “open-circuit able” ones.
It seems though that manufacturers are engineering solutions into their designs without the customers even being aware. Take for example the ZAPPI EVSE equipment that can work wirelessly or with multiple CTs wired back with CAT5 cable. Those wired CTs are internally protected and the customer is blissfully ignorant of it.
It leads to a sense of complacency around CTs.
Many a DIY enthusiast will adopt the approach that if the wire is disconnected it is safe, but with a CT the opposite is true.
In my line of work, CTs are treated with big respect.

I would install CT terminals at each Solis, if I put them in series on a standard CT because I don’t have a perfect memory either:

image

Edit: I better explain these terminals because they might confuse some readers as it probably isn’t self-explanatory to everyone
The CT wires come in as pairs at the bottom and leave at the top.
There is a shorting strip for every terminal pair and a sliding terminal that opens the connection above it. They are mechanically interlocked so that the straight-through connection cannot be opened unless the cable pair is shorted. This allows work to happen downstream of the shorted pair and keeps the upstream CT happy at the same time.
The terminals are shown in the un-shorted position at the moment and should always be mounted in this orientation so that gravity will make the shorting strip fall into the shorted position ( if it is loose).

113

Thanks, now everything is clear…

114

Right, so all office computers were off this weekend, the 2 panels I have will cater for that.

image
image1017×418 23.8 KB

Max the panels can do on good days, is about 22kWh.
image

But, it is still nowhere near enough for recharging the batteries

image
image1012×415 13.6 KB

BUT, I can set them to Optimized with Battery Life and “Let it goooo …”

I may have an option for a 150/35 … will see where that leads as that could mean up to 2kw extra, starting with 2 x panels, expanded wisely over time, and it is a lower price than a ±2kw Solis, and integrates flawlessly with the existing system.

Solis added will take “strain” off the Victron system. So there is that too.

My core goal is to bring that figure down, wisely, and not “throw money at the problem”.

image
image772×601 31.7 KB

115

Right, now that I know Solis works with Victron, and how to connect it all … the end result of this entire thread?

Sold the 250/100 and ordered a 450/100. The universe was kind to me.

Now I can, as suggested, have all the same model panels on the one MPPT (8 x 350), and the mixed range on the 2nd MPPT (7 x 355-420w).

Or sell the mixed batch of >350w panels and start “new”. Will see where this leads.

2 Likes
116

Just remember an 18s battery is now off the cards. The 450 severely derates above 57 V.
You can only just manage 16s.
See:
https://community.victronenergy.com/questions/121186/warning-rs-450-controllers-derate-current-severley.html

1 Like
117

Thank you for pointing it out! And the link Phil.

Yes, I saw the 62v, checked on the Demo on VConnect, and it limits to 62v.

Sold half the bank, have 3 cells on order, so I’m limited to 17 in any event. Using 1 previous version cell currently.

My thinking:
17S at 3.45 = 58.65v
Or Gelach’s tried and tested 3.375v per cel = 60.75v for 18 cells
18S at 3.45v = 62.1v so that is a no chance.

60v / 18 = 3.33v is a bit too low.

BUT I presume dropping the charge at that +60v to 5a is also going to limit the grid-tie side, which is a HUGE consideration IF I understand it correctly?

So max 17 cells at my normally accepted 3.45v per cell it seems to me.

1 Like
118

60V/17 = 3.52V/cell which I suppose will be OK, but 18s won’t work.

But, in reality, anything above 60V and you are not getting the full functionality, so really 60V is the limit at the MPPT terminals. (And remember, you’ll lose “comma something of a volt” in the cabling at 100A before it gets to the batteries).
But I think you should be OK with a 17s bank.

119

Interestingly, watching the system whilst working at my desk, iPad for BMS in front of me, and a webpage for the BMV, the BMS SOC gets “stuck” at 80% SOC for a very “long” time before it moves up.

BMV moves up much more “consistently” i.e. pure coulomb counting from BMV shunt vs BMS coulomb count + volts with a shunt.

(parenthesis because I’m talking concepts of what I see, not science)

So, my thinking is, and I stand to be corrected, once it moves above 80% SOC, having been stuck there for a while, the batts have pretty much “absorbed” the bulk of the charge, being far FAR away from 60v at THAT point in time.

The graphs of Lifepo4 seem to say similar to what I unscientifically have observed.

FWIW, I KNOW I need more watts with a grandchild going on 6m in the house.
Couple that with this video of Andy, which made me sell half the bank at cost, 560ah bank dying of “old age”. Dang, even Will (in the comments) agrees with Andy:

120

So, giving it all some more thought re. the 60v and the 17/18 cel thing, also listening to the advice on that Victron can run up to say 150v, it seems that a 250/85 (4S3P) and a 150/35 (3S/1P) is a better move, slightly cheaper too.

But that is not all …

Having a 150/35 = the off-grid camping MPPT, it being able to do 12/24/48v.
… can also go back to 18 cells (62.1v) if I can find some.
… and no issue ever on high batt volts and lower grid-tied production, the 60v limit.
… don’t need to change the existing array, just add the 3 panels onto the 150/35

What am I missing?

121

If I understand you already have the 150/35, then a no brainer actually, in fact I’d say even keep the 250/100 if you still have that option, you never know where you want to expand to in future.

If I remember the 150/35 can take something like 1.6 kW so that won’t leave you much for future expansion, but you can always cross that bridge when you get to it.

You can buy a 100/30 for the camping setup at a later stage.

122

The cost of a 700w Solis came close to R5k with all the parts and whatnot I wanted to add. So unless one goes bigger, it did not make sense for 700w more.

The 250/100 is sold, like in “passing it on” sold as I was not prepared to push the volt limits if I went 5s/3p and I did not have space for 4s/4p without more costs on frames and holes in Kliplock sheets.

So a 250/85 sorts the existing array to a T, basically a direct swap ito cost. I win, the buyer wins.

The 150/35 is there to cater for the 3 extra panels (1050w) I have laying around now AND it is able to handle 500w on 12v, more than enough “space” for some serious off-grid camping if I want to.

Also to mention. A pal has a brand new 150/35. He bought himself into a corner having installed 2 x 3kva’s with a 17 cell 280ah bank. So ideally I take over his 150/35 and he buys a bigger MPPT. Again, we both win some.

End result: The cost of 3 more panels + 150/35 + mounting costs as I have the wires on the roof already.

Is this a financial sensible transaction? Don’t really care as it falls in the “HOBBY” category. :wink:

I do know the additional 1050w of paneling, me then on 4200 + 1050 = 5250w, is more than fine, data to substantiate that.

And WHEN CoCT makes a move to wanting some of my power, then I go Solis on the spot.

123

This is the Solis I now have:

e8f88ae552850ab0cb605ae422fedce9c44cb849
e8f88ae552850ab0cb605ae422fedce9c44cb849986×1855 158 KB

And I know how to connect it, no, have it connected by a specialist - yeah, not my cup of tea:
7c92a664d2d09e6a4d2dd445df11507158cea120_2_463x500

However, it is still “maak 'n plan”, not “plug and forget”. Not like the Fronius, integrated, so, I’m not going there yet, at least, not unless Eskom gets worse. And I don’t think that is going to happen, seeing as Cpt is making serious moves. The private sector is making moves, and many in SA have made moves. Will see.

Also, when CoCT is ready for home user feedback - 2025 (?) - I can sell back in the summer using the Solis. Also realized, one can only sell back if there is no LS … so that call must be made wisely.

Anycase, will mount the Solis right next to the main DB, a small rewire and the Solis can be on the Critical Loads DB with the Victron … if I’m “forced”.

For the foreseeable future:
Stock standard grid-tied Solis with the main DB.
Eskom is on, Solis heats geysers primarily, as all are on the Non-Critical DB in any event.
The rest is on the Critical Load DB, and there I have a 5.2kw array.

Thinking is:
Can use more batts at night in winter and recharge them “faster” during the day if the Solis takes care of the geysers, 2 of them.
Spare from the main system, there will be, can help with geysers for the 2 hours per day … seeing as I can heat 2 at a time, or sequentially on not-too-bad weather days.

How do I envisage it?
Solis does Solis, Victron does Victron ito to not feed anything back initially.

The first question now:
Does the Solis CT come before the Carlo, or after the Carlo?
My “feeling”, and it is “biased”, is that the CT of the Solis comes first, then the Carlo, then the main DB.

Any thoughts will be appreciated.

EDIT: O, the Solis is all about ROI. 7 adults and 2 geysers, is a challenge.
FWIW, the 3rd Geysers is the “happy marriage” geyser. 50l kitchen one and it is permanently on Eskom,. I don’t care. :slight_smile:

124

If no loads are tapping off between the two, it won’t matter in terms of function.
(Same current seen by both).
I prefer to keep my equipment (CT & Meter) downstream of the respective MCB used.

125

Perfect. Exactly as in the picture.

image
image527×523 39.1 KB

(Just one CT, not 3xCT)

1 Like
126

Next question: Panels - what would “you” do?

I have:
Currently on the 250/85:
12 x Canadian Kumax CS3U 350P: 4.2kw
Pmax: 350W
VMPP: 39.2V
IMPP: 8.94A
Voc: 46.6V
Isc: 9.51A

Currently on the 150/35 - yeah, the mix and match array - 1.05kw
1 x Canadian Kumax CS3U 355P
Pmax: 355W
VMPP: 39.4V
IMPP: 9.02A
Voc: 46.8V
Isc: 9.59A

2 x Canadian CS3W-420P
Pmax: 420w
VMPP: 39.5 V
IMPP: 10.64 A
Voc: 48 V
Isc: 11.26 A

Question:
Would it be a good idea, to take 8 x 350w panels, and put them on the Solis - 2.8kw?
Then get like say 10 x 600w panels on the 250/85 - 6kw? It is then 122% over specced.

Or I’m going to HOBBY Level … just get 5 x 420w panels and make a 7 x 420w array - 2.94kw?

Cause 4.2kw (existing) + 2.9kw = 7.1kw = enough now!!!
EDIT: Forgot …

  • Camping system … use the “leftover” 355w on the 150/35 and if I want, add more panels if one can pick up some “cheap”. So one 355w added = 7.45kW (… and I use everything I have already … )

versus

6kw (new) + 2.8kw (existing) = 8.8kw = lekker man!
… with spare panels left over after the big expense of 10 x 600w new

127

The Solis could go downstream of the Multi as well.
Hint: That’s where I’d put it.

128

Remember, I don’t use the AC_Out2 … so that will mean the CT of the Solis is “on the” Critical Loads DB? Ja?

Did you see this?

129

You lose AC out 2 during load-shedding, so you wouldn’t put it there anyway.

Yes, I saw that write-up, but I don’t know why it behaves like that.
I operate mine without a CT and pure frequency control so I can’t offer insight into that behaviour.

Even so, bugs aside, putting it upstream of the inverter guarantees 0W output every time.

130

Thanks. I think, for a start, I prefer the 0W output “every time”.

But where did you think of connecting it?

As I’m thinking, 2 inverters on a PAYG meter can create some interesting results. Just a gutfeel.

Also, gutfeel again, Victron has added workload like charging the batts and keeping a 20w draw. Yes, it can “dump” to the batts, but still, if batts are full quickly, it does take seconds. I “think” the Solis can “act faster”, hence MB then Solis CT then Carlo.