Thinking about battery backup

[MASolarGuy]

"Yes, each sell will have its own fuse using the battery hookup fused nickel sheets. I believe they pop at anything over 10amps directly to the cell."

Ahh, so there is method to their madness! And here I thought that was just mechanical flexibility, which also happens to provide thermal relief if you do solder.

Is that supposed to be nickle throughout, not just nickle plated something else?
If you can identify it you can maybe find its ampacity.
Perform 4-wire resistance measurement to be able to calculate voltage drop across it. The geometry is tough, but could could measure from battery terminal to body of strip, also end to end. The end to end resistance would be real good to calculate voltage sag along the strip. Just build a ladder in Excel of cell current and strip resistance. Add the copper wire in parallel.
You may find that at 120 amps there is a significant voltage difference between cells due to not having perfectly balanced resistance.

Yes, the nickle strip has its own ampacity restrictions and resistance. Eyeballing those wasp waists vs. the spiral fuse, they can't handle more than 100A. If someone tapped off the end for 125A without soldering a copper wire along the way, it would pop.

I still say the best assembly would be to use one row of that strip for a row of 34 battery positive terminals, and the other row for a row of 34 battery negative terminals. Then for the terminals at last group of batteries each end, unravel a 2/0 battery cable, insulate each twisted bundle, and solder each one to the strip between cells 1&2, 3&4, etc. That will provide absolutely perfect symmetry.

It appears 2/0 has 19 twisted groups. For 34p arrangement of cells you would connect 17 of those.

View attachment 20447

They are nickel plated steel. Here is a link with more info that you may find interesting. https://batteryhookup.com/products/...8650-cell-level-fusing?variant=29596921823311.

This was interesting:
" For example, if you had a 20 cell parallel pack 2 cells wide by 10 cells tall you can carry 6 amps from each row of 2 cells times the 10 rows tall for a total of 60 amps considering you connected your main positive or negative up and down connecting to the 10 rows to evenly distribute the 60 amps, 6 amps per row."

With mine being 17 rows, i should get 102amps if i run a bar down the middle? Ill also be running it the full length of both side, not just one.

 

[Hedges]

They are nickel plated steel. Here is a link with more info that you may find interesting. https://batteryhookup.com/products/...8650-cell-level-fusing?variant=29596921823311.

This was interesting:
" For example, if you had a 20 cell parallel pack 2 cells wide by 10 cells tall you can carry 6 amps from each row of 2 cells times the 10 rows tall for a total of 60 amps considering you connected your main positive or negative up and down connecting to the 10 rows to evenly distribute the 60 amps, 6 amps per row."

With mine being 17 rows, i should get 102amps if i run a bar down the middle? Ill also be running it the full length of both side, not just one.

Yes, if your bar itself is sized for 102 amps. That's about 6 awg if you use ampacity in free air, 90 degree C. Which is what it could subject the battery to.

Not sure what you mean by "both sides". If the sheet metal is welded to 2 rows of batteries, only one side is accessible.

With the 4 cell wide roll, I think you can have 2x17 battery positive terminals and 2x17 battery negative terminals joined, and maintain perfect balancing of resistance across all cells. No wires to solder except the for where inverter cables connect. (plus, of course, small gauge wires for balancing.)

NICKEL FUSE 2P-6P WIDE CONTINUOUS ROLL BY THE FOOT! 18650 CELL LEVEL FUSING

Each foot equals 15 cell connections long by 2, 3, 4, 5, or 6 wide. For example 2p wide would be 30 cell connections per foot (2 wide x 15 long) and 6p would be 90 cell connections per foot (6 wide x 15 long) Introducing custom cell level fused nickel design that can be spot welded with ease...

batteryhookup.com

 

[MASolarGuy]

Not sure what you mean by "both sides". If the sheet metal is welded to 2 rows of batteries, only one side is accessible.

By both sides i mean the top Positive side and the bottom Negative side. Both will have 2wide x 17long Nickel fuse, which will put them all in parallel.

The Battery hookup site states that if you run a copper wire the full length of One side you distribute the Amps evenly across all 17 rows. Does putting a full length copper wire on both the pos and neg increase the amps it can pull?

For EX: With 1 copper wire the full length of the Positive side of my 2x17 pack i can pull 102 amps.
If i add a second wire, the full length of the Negative side, can i pull more than 102 amps?

 

[Hedges]

By both sides i mean the top Positive side and the bottom Negative side. Both will have 2wide x 17long Nickel fuse, which will put them all in parallel.

The Battery hookup site states that if you run a copper wire the full length of One side you distribute the Amps evenly across all 17 rows. Does putting a full length copper wire on both the pos and neg increase the amps it can pull?

For EX: With 1 copper wire the full length of the Positive side of my 2x17 pack i can pull 102 amps.
If i add a second wire, the full length of the Negative side, can i pull more than 102 amps?

No. The copper wire is needed for both, because 102 amps of electrons come out the copper wire on the negative side, go through your load (inverter) and 102 amps of electrons go back in the positive side through the copper wire.

A switch at either positive or negative side stops current flow. A fuse at either blows if current is excessive and tops the flow.

You only get more current with things in parallel, e.g. if you make two of these battery banks and connect them side by side. The each produces 102 amps for a total of 204 amps.

But what you can do is run the copper wire all the way along the sheet metal strip, off both ends. You can pull 102 amps from one end, and another 102 amps from the other end, for a total of 204 amps. Assuming you have enough batteries and the sheet metal fuses can handle it. 34 batteries in parallel, 204/34 = 6, so that would be drawing 6 amps from each battery and through each fuse. The sheet metal fuse vendor's page mentions both 3A and 5A limits, and says a design with higher current is coming later. So if you use this one, staying at 3A per cell for 102A makes it less likely you'll blow a fuse.

With just two cells (e.g. two series strings of big batteries like lead-acid), tapping opposite corners of positive and negative terminals for inverter cables perfectly matches current draw and voltage drop for every cell/battery. With 3 or more strings, current is drawn unevenly between cells due to IR drop across wires. Other wiring configurations could restore perfect balance.

Your set of 34 cells in parallel, in an array 2 cells wide and 17 cells long with a single copper wire joining them and connected at one end, would cause 9.1 mV difference in cell voltages if each cell sourced 3A and the wire was 6 awg. (Actually, at different voltages the current they sourced would vary, but I've ignored that in creating this curve.)

 

[MASolarGuy]

Ok, it sounds like 6awg wire will work for running the length of the pack with a ring terminal on the end to connect to the next pack in series, correct?

I also picked up some watt meters for the appliances i plan run off the batteries.

Here is my Fridge. Its normally around the 169w/1428mA but jumps to 361w/2921mA ocasionaly.



Here is my chest freezer. I havent seen it move much.

 

[Hedges]

Ok, it sounds like 6awg wire will work for running the length of the pack with a ring terminal on the end to connect to the next pack in series, correct?

It looks like 6 awg would just barely carry the current, with significant temperature rise (which the cells will feel.)

Why "with a ring terminal on the end"? Why not put ring terminals at both ends, so current splits and only half the current goes through only half the wire length?

Why "to connect to the next pack in series" when you could use 4x17 strips of the sheet metal to connect batteries in series/parallel? Then there would be no wires except at the ends.

 

[Ampster]

Why "to connect to the next pack in series" when you could use 4x17 strips of the sheet metal to connect batteries in series/parallel? Then there would be no wires except at the ends.

I am not following you. We seem to have a different picture. Again I think @MASolarGuy is following a best practice of HBPOWERWALL who has shown many videos of how his packs work. I think he is following that modular concept.

I respect your technical knowledge but some of us are visual learners and pictures are much better than a thousand words. For example what is a 4X7 strip? How is that better to connect than ring terminals as shown in the pictures. The concept was modularity. With ring terminals it takes two bolts to take out one of the modular parallel groups and replace it with another.

 

[Hedges]

Thank you, I'm also very visual but being lazy have tried to convey with words. I do have a deep background hands-on on-the-job, as a shade-tree mechanic, and from undergrad and grad school.

What I'm trying to do is make voltage drops through all conductors balanced to all cells, as well as minimized.

I'll see what I can put together using paper dolls.

In the picture below I've made an assembly of batteries (AA alkaline) in 2s6p configuration.
The paper "sheet metal" strips are 4x8, copied from vendor's website and printed out. It could support 16p, but I didn't have enough cells.
The sheet on the bottom has been "welded" to the negative terminals of 6 batteries in one bundle (left), and the positive terminals of 6 batteries in another bundle (right).
The sheet on left has been "welded" to positive terminals of 6 batteries, and the sheet on right to negative terminals of 6 batteries.
cable-tie "wires" have been "soldered" to the sheets on left and right, providing symmetric connection to cells.

Untwisting a 2/0 cable would provide 19 stranded wires, which could then be insulated with heatshrink and used for those balanced lengths of wire.
His entire assembly of 14S34P could be assembled this way, no wires between cells. Just battery cables coming off the ends.



Materials used (plus a bunch of cells, of course.)

NICKEL FUSE 2P-6P WIDE CONTINUOUS ROLL BY THE FOOT! 18650 CELL LEVEL FUSING

Each foot equals 15 cell connections long by 2, 3, 4, 5, or 6 wide. For example 2p wide would be 30 cell connections per foot (2 wide x 15 long) and 6p would be 90 cell connections per foot (6 wide x 15 long) Introducing custom cell level fused nickel design that can be spot welded with ease...

batteryhookup.com

 

[MASolarGuy]

It looks like 6 awg would just barely carry the current, with significant temperature rise (which the cells will feel.)

Why "with a ring terminal on the end"? Why not put ring terminals at both ends, so current splits and only half the current goes through only half the wire length?

Why "to connect to the next pack in series" when you could use 4x17 strips of the sheet metal to connect batteries in series/parallel? Then there would be no wires except at the ends.

Are you suggesting the first pic and or the second?

 

[Hedges]

Are you suggesting the first pic and or the second?

The first pic. That's 2x better in terms of current vs. ampacity and 4x better in terms of IR drop. Just by connecting at both ends to make an "O" instead of 1 end to make a "U".

Then at least do the same for inverter cables at the end - either use a longer bus to form a complete "O" or split the 2/0 cable into two halves, connecting both ends of the bus.

 

[MASolarGuy]

The first pic. That's 2x better in terms of current vs. ampacity and 4x better in terms of IR drop. Just by connecting at both ends to make an "O" instead of 1 end to make a "U".

Then at least do the same for inverter cables at the end - either use a longer bus to form a complete "O" or split the 2/0 cable into two halves, connecting both ends of the bus.

Should be easy enough to connect the POS bottom and top of one module to the neg bottom and top of the next module in series.

 

[Hedges]

But why do all that with copper wire and soldering, if you can built the whole pack just by spot welding all the batteries to a bunch of 4x17 (or 2x34) strips of metal bus/fuse? Also, it'll have less resistance and better balancing of current between cells.

 

[MASolarGuy]

But why do all that with copper wire and soldering, if you can built the whole pack just by spot welding all the batteries to a bunch of 4x17 (or 2x34) strips of metal bus/fuse? Also, it'll have less resistance and better balancing of current between cells.

Those would be 2s34p packs instead of just 1 34p pack. The idea is that if one pack acts up, needs maintenance, etc.. i just unplug that row and diagnose the pack/34 batteries. The one draw back of these fused cell sheets is if you have to replace a cell, you cant jut pop it out without pulling the nickel strip off the batteries around it.

I also like that, in the unlikely event that one catches fire for what ever reason, its not in the middle of a pack surrounded by 67 other cells.

 

[Hedges]

Those would be 2s34p packs instead of just 1 34p pack.

I meant 14s34p, the whole thing.

"i just unplug that row and diagnose the pack/34 batteries"

Your 34 cells are all in parallel, so not easy to diagnose individually. Anything is series already is, regardless of how connected (copper wire or sheet metal).

Gross failure will already be diagnosed for you by the fuses.

To identify one cell out of 34 in parallel, you won't be able to see state of charge differences by voltage, but while under charge/discharge conditions you can measure voltage drop across the fuses, using them as current-measurement shunts. Or measure temperature, e.g. with IR camera.

If you want to be able to replace cells, get sheet metal with room for a few extra cells, like 36 or 38. The copper wire soldered to it will be useful so to balance current between cells in this case. Just snip the fuse (or blow) to isolate. It can be physically removed only if in no greater than 2-cell wide rows.

 

[MASolarGuy]

I plan to connect them in series by putting a ring terminal on each end and connecting it to the next pack by bolting them together.

The other option is to simply pull a single battery from the 34p to make it 33p and a can do 3 rows of 11 cells. This would allow me to for a complete loop using the copper wire on a single pack.

 

[Hedges]

I plan to connect them in series by putting a ring terminal on each end and connecting it to the next pack by bolting them together.

The other option is to simply pull a single battery from the 34p to make it 33p and a can do 3 rows of 11 cells. This would allow me to for a complete loop using the copper wire on a single pack.

So why not do 3x12 with 34 batteries, leaving two sites empty where you could later install replacement cells if needed.
Put a ring terminal at each end, so length from any battery to a ring terminal is shorter and half as much current goes through the wire, split 4 ways now.

By the way, copper wire soldered to the strip is likely to put mechanical stress on the fuse elements. Even stranded wire will have a tendency to not want to be flat.

 

[Ampster]

By the way, copper wire soldered to the strip is likely to put mechanical stress on the fuse elements. Even stranded wire will have a tendency to not want to be flat.

Can you explain? It looks like there is a lot of room for that spiral to move both vertically or horizontally. Is this stress from experience doing that? There was no mention of this in any of the HBPOWERWALL videos. Did you see the part in the video where he pre tinned the wire and the nickel?

 

[Hedges]

Can you explain? It looks like there is a lot of room for that spiral to move both vertically or horizontally. Is this stress from experience doing that? There was no mention of this in any of the HBPOWERWALL videos. Did you see the part in the video where he pre tinned the wire and the nickel?

Just my judgement from looking at it. The spiral fuse is very limber, won't provide much resistance to motion without extending all the way.
See how flat you can make a piece of copper wire. I don't think it will break he fuses, but will probably deform them. You may want to make some sort of mechanical support.

If those plastic frames seen in some photos are used, the bundle of batteries will be held together and can be handled that way.
Securing the sheet metal to the frame might be good.
In the photo of an array, do I see sheet metal used for one group of batteries, but individual wires for the rest?

https://diysolarforum.com/attachments/1597940256226-png.20419/

 

[MASolarGuy]

So why not do 3x12 with 34 batteries, leaving two sites empty where you could later install replacement cells if needed.
Put a ring terminal at each end, so length from any battery to a ring terminal is shorter and half as much current goes through the wire, split 4 ways now.

By the way, copper wire soldered to the strip is likely to put mechanical stress on the fuse elements. Even stranded wire will have a tendency to not want to be flat.

Is this what you were referring to? The grey spots are empty. For the soldering, ive heard of guys zip tying the copper bar to these nickel plates with great success as well. No solder needed.

 

[Hedges]

Is this what you were referring to? The grey spots are empty. For the soldering, ive heard of guys zip tying the copper bar to these nickel plates with great success as well. No solder needed.

View attachment 20660

Exactly.

Yes, maybe zip tie, fiberglass tape, etc. around batteries, resist busbar pulling sheet away.
I see many packs wrapped in heatshrink or similar, but you want airflow to cool cells used at higher power.

But I wouldn't trust pressure contact from a zip tie for electrical connection. You want uniform contact resistance near every cell, or else the current will take a different path and voltage at cells vary.
Maybe thread a thin copper through every hole in the strip, wrapping around the heavier wire and cinch it up like a shoelace. Then tack solder at every hole and to the heavy wire. Or just press the wire against the strip and tack solder to it at every hole.