18650 DIY powerwall 48v

[kje]

Yup, that'd be fine. Since all 4 strings will be on the same side, no real worries about paralleling them together. Just combine 2 strings and go into one Input, and parallel the other 2 and go into the other port.
Or, take 2 strings and parallel and input into the Solaredge SE 5000, and then a single string into each of the ports of the Voltronic. That'll give you 3 controllers to work with

Yes, but the SE 5000 is not an offgrid inverter...

 

[Korishan]

Ohhhhhh, I see what this thing is. I didn't realize it was a Grid-Tie Inverter that injects into the existing line. Okay. Well, in that case, you could *probably* configure it to be in parallel with your other inverter. Just put the installation where the other one is the "Main Inverter" and the SE taps into the output. Basically you are creating your own island.
It's just that it can't be used "by itself", it must have an upstream AC source.

 

[kje]

Ohhhhhh, I see what this thing is. I didn't realize it was a Grid-Tie Inverter that injects into the existing line. Okay. Well, in that case, you could *probably* configure it to be in parallel with your other inverter. Just put the installation where the other one is the "Main Inverter" and the SE taps into the output. Basically you are creating your own island.
It's just that it can't be used "by itself", it must have an upstream AC source.

Do you mean ac out from SE5000 - to ac in on Voltronic Axpert Max?

 

[Korishan]

Do you mean ac out from SE5000 - to ac in on Voltronic Axpert Max?

No
The AC Out of the SE5000 will be in parallel with the AC Out of the Voltronic. But all loads will be *after* the SE5000 connection

Voltronic
       \ AC-Out
        \
         \   SE5000
          \ /
           |
      Loads Panel
           |
         Loads

It's a Grid-Tie, it ties into the Grid on the Loads side of the Grid. It's not for upstream connections. You could also essentially plug the SE5000 into a wall outlet in a room and effectively do the same thing. Altho, that would limit your output to 15A more than likely

 

[kje]

No
The AC Out of the SE5000 will be in parallel with the AC Out of the Voltronic. But all loads will be *after* the SE5000 connection

Voltronic
       \ AC-Out
        \
         \   SE5000
          \ /
           |
      Loads Panel
           |
         Loads

It's a Grid-Tie, it ties into the Grid on the Loads side of the Grid. It's not for upstream connections. You could also essentially plug the SE5000 into a wall outlet in a room and effectively do the same thing. Altho, that would limit your output to 15A more than likely

Ok, so Voltronic and SE5000 in parallel on ac out? Is there any risk for the Voltronic inverter involved, can it be damaged configure it this way?

 

[Korishan]

Yeah, the are parallel on the output. As far as being damaged, I'm not sure. I doubt it, considering that the devices that are designed to be grid-tied that way are also safe to be put on the national grid without risk of backfeeding the grid, especially if there's a downed line.
The way they work is the monitoring the line for a valid signal and only output enough to not zero it out, aka they won't output so much that they overpower the mains (in this case the Voltronic) output, so there'll always be a bit of power coming in through the line, up to the max output of the grid-tie unit.
I would definitely contact both Voltronic and Solaredge to verify this, though. I personally haven't messed with this type of configuration so I'm a little fuzzy on the details, but that's how I understand them to work together.

 

[kje]

Yeah, the are parallel on the output. As far as being damaged, I'm not sure. I doubt it, considering that the devices that are designed to be grid-tied that way are also safe to be put on the national grid without risk of backfeeding the grid, especially if there's a downed line.
The way they work is the monitoring the line for a valid signal and only output enough to not zero it out, aka they won't output so much that they overpower the mains (in this case the Voltronic) output, so there'll always be a bit of power coming in through the line, up to the max output of the grid-tie unit.
I would definitely contact both Voltronic and Solaredge to verify this, though. I personally haven't messed with this type of configuration so I'm a little fuzzy on the details, but that's how I understand them to work together.

I think I'll do it simple and just using the Voltronic Axpert Max inverter in an offgrid setup at my cottage with the Stion solarpanels (Voc: 77.4v Isc: 2.6a). And maybe selling the Solaredge SE 5000.


Back to my powerwall and solarsetup at home:

I'm now working on battery #2. I'm using the same BMS as battery #1; JK-B2A24S-20P. It can handle 200A continuous discharging and charging current. I will connect the two battery banks in parallel, and then into a circuit breaker before my Voltronic Axpert Max inverter.

Is it so simple as combining the two positive cables from battery #1 and #2, and the two negative cables from battery #1 and #2, and then into a circuit breaker before the inverter?

Is there any settings I should be aware of or change when connecting the two battery banks in parallel?

 

[Korishan]

Is it so simple as combining the two positive cables from battery #1 and #2, and the two negative cables from battery #1 and #2, and then into a circuit breaker before the inverter?

Is there any settings I should be aware of or change when connecting the two battery banks in parallel?

Yup, pretty easy. I would recommend each battery having its own breaker, tho. For ease of maintenance and safety. And then you can either combine into a single breaker before the inverter or not, that's a bit optional.

The only main concern to be aware of is that both batteries should be pretty close to the same voltage before connected them together. Try to be less than 0.1V if possible. That way you don't have too much power flowing from one to the other when connected. No sparkies.

 

[OffGridInTheCity]

I'm not clear if you have LifePo4 or 18650/Lithium-ion. If you have LifePo4, not sure if it's cells or bricks. So forgive me for that as I write extra to mention each case.

IF you have LifePo4, you may want to charge each battery - e.g. powerwall and then 'new battery' to 100% - before connecting in parallel because the charge/discharge voltage curve is soooo flat thru most of the middle range the voltage is not a good metric to determine if OK to parallel in terms of them equalizing to each other to have the same SoC.

IF you have Lithium-ion (my case) then the voltage curve is slanted enough this isn't a problem. I routinely add in new batteries at the middle voltage range (at close voltage as @Korishan suggests) and they balance out quickly while in operation using Batrium Auto-Level - I don't need to do the 100% charge technique used for LifePo4. If you don't have effective balancing... then the very close voltage is good or 100% charge technique is good either way.

Breakers? To me there's 2 issues. 1) Protect against short/catastrophic current flow and 2) convivence of enable/disable.
And this goes back to LifePo4 bricks vs individual cells as to my thinking. For LifePo4 300ah bricks in series I would consider breakers or fuses on each battery (to the common buss) to limit catastrophic current flow as a level of protection. For small cells then fuse wire can provide the 'limit current flow' protection.

In my case I have ~14,000 cell level fuse wires - so each cell has protection and don't need breakers for catastrophic short purposes on each individual battery. And since it's physically daunting to 'remove' a battery from parallel with the others I didn't bother with breakers for convenience. *I need to work on 2 of my batteries this winter and the lack of battery level breakers will definitely be an annoyance and even a bit of 'short management' risk as I have to bolt/unbolt wiring and loose wiring can be risky.

Each of my 10 batteries are paralleled with 4/0 awg to a common buss and then routed to a single 400a SACE S3 as the master breaker - which is also a shunt-trip that Batrium can throw if cells get out of spec. The master breaker is part of my BMS protection system.

 

[kje]

I'm not clear if you have LifePo4 or 18650/Lithium-ion. If you have LifePo4, not sure if it's cells or bricks. So forgive me for that as I write extra to mention each case.

IF you have LifePo4, you may want to charge each battery - e.g. powerwall and then 'new battery' to 100% - before connecting in parallel because the charge/discharge voltage curve is soooo flat thru most of the middle range the voltage is not a good metric to determine if OK to parallel in terms of them equalizing to each other to have the same SoC.

IF you have Lithium-ion (my case) then the voltage curve is slanted enough this isn't a problem. I routinely add in new batteries at the middle voltage range (at close voltage as @Korishan suggests) and they balance out quickly while in operation using Batrium Auto-Level - I don't need to do the 100% charge technique used for LifePo4. If you don't have effective balancing... then the very close voltage is good or 100% charge technique is good either way.

Breakers? To me there's 2 issues. 1) Protect against short/catastrophic current flow and 2) convivence of enable/disable.
And this goes back to LifePo4 bricks vs individual cells as to my thinking. For LifePo4 300ah bricks in series I would consider breakers or fuses on each battery (to the common buss) to limit catastrophic current flow as a level of protection. For small cells then fuse wire can provide the 'limit current flow' protection.

In my case I have ~14,000 cell level fuse wires - so each cell has protection and don't need breakers for catastrophic short purposes on each individual battery. And since it's physically daunting to 'remove' a battery from parallel with the others I didn't bother with breakers for convenience. *I need to work on 2 of my batteries this winter and the lack of battery level breakers will definitely be an annoyance and even a bit of 'short management' risk as I have to bolt/unbolt wiring and loose wiring can be risky.

Each of my 10 batteries are paralleled with 4/0 awg to a common buss and then routed to a single 400a SACE S3 as the master breaker - which is also a shunt-trip that Batrium can throw if cells get out of spec. The master breaker is part of my BMS protection system.

I have 18650 DIY powerwall 48v.

My battery #1 has a 200a dc breaker before the inverter. I've now ordered a second 200a dc breaker for battery #2.

 

[Fukuokian]

Blocking Diodes are generally only installed at the ends of a string of panels if they are to be connected with other strings of panels. The overall voltage of a string needs to remain independent in case one drops in voltage, this is where the Blocking comes in. It keeps the higher voltage string from pushing into the lower voltage string. This causes hot spots on the lower voltage string and can cause damage over time if there's enough power flowing. This could also lead to panel fires.

I've been following this thread, and want to say thanks for all the good information! In my case, I'm planning on having 3 strings of 3 panels (in series), all on the same side of the roof, with no shading issues. I'll use a combiner box, 4-1 so I could always add another string later. I was wondering...

Do blocking diodes also give protection if there was ever a short circuit somewhere in a string, so as to prevent current (especially the higher current from paralled strings) from back-feeding through the panels and damaging them? (I'll have circuit breakers on each string too, but I know they are not instantaneous in cutting off a string) It seems to me that blocking diodes might add this layer of protection. ??

 

[Korishan]

Do blocking diodes also give protection if there was ever a short circuit somewhere in a string, so as to prevent current

Not really. They have an upper limit. That's why I saying to do the combining "after" the breakers for each string. That's what will protect from the shorted system. Breakers will pop within a few seconds, or less, depending on how high current load is.
I sized my breakers just barely higher than what the string can do. My strings can do about 8A wide open, so the breakers are 10A.Then they merge after and the combined breaker is a 20A. Even if they surge to max current, they can't overload the breaker on proper use. But if there was a short, they'd overload.

 

[kje]

Update:
My battery #1 in my 18650 powerwall I discovered to weaker cellpacks. They where sinking in voltage.

Solution: I disconnected the two cellpacks and desolder all the cells on the positive side. Let them sit for two days and then checked the voltage on each cell. I found one or two cells sinking in voltage. I changed them and solder in the new ones.

 

[kje]

I made a fume extractor with a bathroom fan, some cardboard and gaffa, and even some 12v led strip for soldering the 18650 cellpacks.

 

[OffGridInTheCity]

A few years ago I found that rare 'handyman' that was both really good but also didn't mind doing weird stuff - like put an exhaust fan in my office disguised as a furnace vent. It works surprisingly well as I build packs....

He had to crawl a significant distance in ceiling to get this in place + run wire down the wall to on/off switch - I'm still amazed he was willing.

 

[kje]

I promised myself not beginning with battery #3 before I'm finished with 18650 battery #2. But I couldn't resist opening the Nissan Leaf battery wich is for battery #3. I began disassembly Nissan Leaf modules and store in the metal cabinet.

 

[Fukuokian]

I like the repurposed lockers! Do the threaded rods go into added vertical bracing, or just the side panels? Will the bottom/sidewalls support the full weight once it's fully loaded, or are you planning to put support/blocking in the space under the flooring?

When you assemble the batteries in their final stacks/groups, check the batteries metal cover surface for bulging. Usually some are concave and some are convex a little bit. Try to put convex into concave... Imagine these laying flat... battery module --> [ ) ) ] <--next battery module.

I kept my batteries under compression as much as possible... especially when I was charging them.

Good Job!

 

[kje]

I like the repurposed lockers! Do the threaded rods go into added vertical bracing, or just the side panels? Will the bottom/sidewalls support the full weight once it's fully loaded, or are you planning to put support/blocking in the space under the flooring?

When you assemble the batteries in their final stacks/groups, check the batteries metal cover surface for bulging. Usually some are concave and some are convex a little bit. Try to put convex into concave... Imagine these laying flat... battery module --> [ ) ) ] <--next battery module.

I kept my batteries under compression as much as possible... especially when I was charging them.

Good Job!

Good point, I have to do something about the weight here. The left side of the cabinet with the Nissan Leaf modules I plan to cut away the threaded rods and support the floor.

Yes I discovered that the metal cover surface is convex / concave. I'll try to remember that when I assemble the batteries in their final stacks.

I didn't know they needed compression when charging or storing. Thank you for pointing this out.

I've now taken them out of the metal cabinet and used the end pieces I got out of the battery bank to store them compressed. I don't know how much compression though... Do you know?

 

[Fukuokian]

Good point, I have to do something about the weight here. The left side of the cabinet with the Nissan Leaf modules I plan to cut away the threaded rods and support the floor.

Yes I discovered that the metal cover surface is convex / concave. I'll try to remember that when I assemble the batteries in their final stacks.

I didn't know they needed compression when charging or storing. Thank you for pointing this out.

I've now taken them out of the metal cabinet and used the end pieces I got out of the battery bank to store them compressed. I don't know how much compression though... Do you know?

View attachment 33637

I think the torque settings are in a link to a Nissan battery assembly/disassembly document I posted before. I think the corners of the modules have a plastic tube (edit: nope... there are some metal collars, and the ends of the pouches have plastic pieces... I should add that I'm guessing for the rest of the sentence. ) that prevents over compression as long as they aren't overtorqued where those tubes would get damaged. I think it was around 10 Nm, but definitely check first.

 

[Fukuokian]

metal cover surface is convex / concave

They are originally all concave, and I think are engineered to allow for some swelling. The consensus I found when looking into it is that slight bulging isn't a big deal. Re: https://www.diyelectriccar.com/threads/damaged-nissan-leaf-battery-bulge-module-im-screwed.189642/.