Common pack design mistakes, how to avoid?

[Pasha]

Just wanna know your thoughts on the design which would be from each parallel group two 10awg silicon tinned copper wire would connected to the next parallel group in series.

[goatman]

Thanks litespeed for your reply. My question is that would it be better to use two 10AWG(ampacity ~120amps) copper for serial connections or pure nickel from each cell to to next cell in the parallel group.

ES member Agniusm aka NESE cell holders comment on copper and nickel
viewtopic.php?f=6&t=101548#p1545141

if youre fusing each cell to a bus "bar" and then doing serial connections on each side of the pack, so 2-10awg. 1 on each side of pack. im trying to picture how pure nickel can be used. the only thing I can picture is no more cell fusing, no more bus bar just a strip of nickel across the cells like a NESE cell holder, that's why I included the link, is that what youre meaning?

[Pasha]

Thanks goatman for your input. My final design would be using 4p pure nickel strip on the cells and then do the series connection by soldering 10awg wire (two on each side of the pack). I guess this would result in better current flow.

[goatman]

have you looked at the very first post of this thread, I think its docbass's chart on current flow through a pack and where to do the serial connections?
viewtopic.php?f=14&t=84412#p1235095

4p? do you mean 40p instead of 60p?

[Pasha]

have you looked at the very first post of this thread, I think its docbass's chart on current flow through a pack and where to do the serial connections?
viewtopic.php?f=14&t=84412#p1235095

4p? do you mean 40p instead of 60p?

Exactly this is what I am trying to follow. Since my one parallel group of cells will be consisting of 60 cells so running two cables for serial connections will be enough?

4p nickel strip means 4 nickel strips joined together

[goatman]

I was actually hoping to learn from what youre doing, and am trying to understand, maybe someone else will jump in here.

[Pasha]

I was actually hoping to learn from what youre doing, and am trying to understand, maybe someone else will jump in here.

What I meant by 4p nickel strip was that four strips joined together. If you google 4p nickel strip you will get to see the images. As my each parallel group will in 4X15 cell spacer

[amberwolf]

Since my one parallel group of cells will be consisting of 60 cells so running two cables for serial connections will be enough?

The size (gauge, square millimeter area, etc) of serial connection you need doesn't depend on the number of parallel cells.

It depends on the specific amount of current you want flowing continuously, and peak, thru them.

So what is the peak and continuous expectation you have for the completed pack itself?


If they're hidden away where there is no airflow, and bundled together or very close to each other where heat from one is also affecting the other(s), then depending on the chart you look at, typically 10g wire might be good for 20-50A continously, depending on the insulation type, and the length of the run. Shorter the run, the more current it could handle. More airflow, more current.

So a pair of them inside the battery where all the heat is contained and held in, but where the wires can pass their heat into the battery cells (assuming the cells themselves wont' heat up due to using them too close to their ratings), and at most a few feet long, you could probably put 100A thru them continuously. Maybe more, depends on your conditions and the actual wire in question.

Also depends on how you crimp those J-hooks to the series cables themselves. If it's a really good crimp, they won't add significant (or even noticeable) resistance. If it's a bad crimp, then it depends on how bad the resistance is, and how it grows worse over time. If it's soldered, it depends on how good a mechanical connection you started with (a good crimp will be better), and the solder itself if it's a poor mechanical connection.

[Pasha]

Since my one parallel group of cells will be consisting of 60 cells so running two cables for serial connections will be enough?

The size (gauge, square millimeter area, etc) of serial connection you need doesn't depend on the number of parallel cells.

It depends on the specific amount of current you want flowing continuously, and peak, thru them.

So what is the peak and continuous expectation you have for the completed pack itself?


If they're hidden away where there is no airflow, and bundled together or very close to each other where heat from one is also affecting the other(s), then depending on the chart you look at, typically 10g wire might be good for 20-50A continously, depending on the insulation type, and the length of the run. Shorter the run, the more current it could handle. More airflow, more current.

So a pair of them inside the battery where all the heat is contained and held in, but where the wires can pass their heat into the battery cells (assuming the cells themselves wont' heat up due to using them too close to their ratings), and at most a few feet long, you could probably put 100A thru them continuously. Maybe more, depends on your conditions and the actual wire in question.

Also depends on how you crimp those J-hooks to the series cables themselves. If it's a really good crimp, they won't add significant (or even noticeable) resistance. If it's a bad crimp, then it depends on how bad the resistance is, and how it grows worse over time. If it's soldered, it depends on how good a mechanical connection you started with (a good crimp will be better), and the solder itself if it's a poor mechanical connection.

Thanks amberwolf for jumping in. Peak current draw would be max 100amps or 3500watts from this 10S60P (two 5S60P connected in series through XT150 connectors). And I would keep the airflow open keeping two sides of the packs open. 10awg silicon tinned copper wire which I am going use is good for 120amps.

[Pasha]

Since my one parallel group of cells will be consisting of 60 cells so running two cables for serial connections will be enough?

The size (gauge, square millimeter area, etc) of serial connection you need doesn't depend on the number of parallel cells.

It depends on the specific amount of current you want flowing continuously, and peak, thru them.

So what is the peak and continuous expectation you have for the completed pack itself?


If they're hidden away where there is no airflow, and bundled together or very close to each other where heat from one is also affecting the other(s), then depending on the chart you look at, typically 10g wire might be good for 20-50A continously, depending on the insulation type, and the length of the run. Shorter the run, the more current it could handle. More airflow, more current.

So a pair of them inside the battery where all the heat is contained and held in, but where the wires can pass their heat into the battery cells (assuming the cells themselves wont' heat up due to using them too close to their ratings), and at most a few feet long, you could probably put 100A thru them continuously. Maybe more, depends on your conditions and the actual wire in question.

Also depends on how you crimp those J-hooks to the series cables themselves. If it's a really good crimp, they won't add significant (or even noticeable) resistance. If it's a bad crimp, then it depends on how bad the resistance is, and how it grows worse over time. If it's soldered, it depends on how good a mechanical connection you started with (a good crimp will be better), and the solder itself if it's a poor mechanical connection.

Thanks amberwolf for jumping in. Peak current draw would be max 100amps or 3500watts from this 10S60P (two 5S60P connected in series through XT150 connectors). And I would keep the airflow open keeping two sides of the packs open. 10awg silicon tinned copper wire which I am going use is good for 120amps.

Continuous current draw would be 40 to 50amps and max peak of 100amps would be may for 5 to 10 seconds

[hemo]

10s 60p we are talking 600 cells here yes or is this a typo ?
A theoretical 600a current draw.

[Pasha]

10s 60p we are talking 600 cells here yes or is this a typo ?
A theoretical 600a current draw.

Yes it’s 10S60P. 600 cells

[goatman]

now that I understand what you mean by 4p, if it was me and 15 cells per row I would run the serials between cells 4 and 5 and the other between cell 11 and 12. that basically gives you 4 cells on each side of the 10 awg wire. if pure nickel maxs at 10 amps. youre running 2 amps so 4 cells is 8 amps. maybe double the 4p nickel strip?

[hussman]

Please correct me if im wrong, but in the case of a building a high capacity high discharge battery pack. Would this be the best way to build the pack to be able to handle the high current, deliver the high current, and keep the battery pack 'cool'.

25mm.jpg (127.57 KiB) Viewed 942 times

[pwd]

Please correct me if im wrong, but in the case of a building a high capacity high discharge battery pack. Would this be the best way to build the pack to be able to handle the high current, deliver the high current, and keep the battery pack 'cool'.

25mm.jpg

I just completed an odd shaped pack using a similar spot welding layout. I've only got a few cycles on it so far but it seems to perform well and the cells are running cool and staying in balance (less than 0.02V difference).

My only criticism with the photo you've shown is having a single main discharge cable one side of the pack. I would think having the main discharge termination coming from both sides would be a better design and allow for theoretically better current share.

[fechter]

Please correct me if im wrong, but in the case of a building a high capacity high discharge battery pack. Would this be the best way to build the pack to be able to handle the high current, deliver the high current, and keep the battery pack 'cool'.

I used a similar approach and it has worked very well. If the copper wire soldered to the end cell nickel is heavy enough, the voltage drop across it will be so small the current sharing will be good enough. I used a separate buss wire along the nickel but attached the main lead closer to the center to help even the current.

Img_1098A.jpg (99.6 KiB) Viewed 885 times

[Smoke]

Please correct me if im wrong, but in the case of a building a high capacity high discharge battery pack. Would this be the best way to build the pack to be able to handle the high current, deliver the high current, and keep the battery pack 'cool'.

25mm.jpg

That stacking without spacers could cause problems.

I'm going to stack like that with spacers and I'm a bit worried.

Also two levels of cells is going to make the interior side of each cell toasty.

I think it would be great for a high capacity battery with conservative amps per cell. If you are going for really high current draw like 4C+, I would expect heat issues.

If you did a square layout like the one above but with spacers, I think that will create enough space for air flow between cells that high discharge on a pack made up of cells that are good at high discharge will maintain battery life much better than a layout that really concentrates heat at the center of the pack.

Right now I'm doing a battery test on some LG M50T 21700s in an Opus battery charger. It is set at 1A charge and discharge. About half of the cell is exposed to air, the other half is blocked by the charger and I have it set inside a .30 caliber ammo can with the lid open. Ambient temperature is in the upper 70s and with that 0.2C charge/discharge rate, the cells get warm to the touch. My cruise current is going to be more like 0.5C and my finished pack will be much more thermally insulated so I'm worried about heat.

I'm not worried about thermal runaway or failed cells, I'm worried about accelerated wear and losing capacity.

[fechter]

Good point about the spacers. I used Nomex paper and fiberglass tape which won't melt or abrade too easily. My cells get pretty warm during discharge but one face of each brick is against the container wall so heat has a place to go.

[Smoke]

My spacers create a 22mm cell spacing in a cubic pack (orange crate style), with a 21mm cell, that's a 1mm air gap, pretty small but not zero.

The spacers are plastic and I am going to hot glue my cells to them with a small amount of glue but hopefully enough to really lock the pack together. Since the plastic is an insulator, I'm not worried about abrasion. If I can make it work, I want to wrap the perimeter of my pack with aluminum sheet metal and a layer of silicone TIM pad to prevent the heat shrink from abrading and shorting the cans. An extra layer of physical protection, thermal mass and insulation. Of course I'll have heat shrink over that so convection cooling is essentially nonexistent. I'm just trying to do as much as I can.

[nardcox]

Img_1098A.jpg

On how many spots did you solder the busbar on the nickel? I'm working on a similar setup.

[jimmyhackers]

i wish id seen this thread a lot earlier.

id always thought that as long as the wires between cells are thick enough to take the max amps all would be fine.
i never though it could effect individual cell performance, leading to voltage sag, overstressing individual cells and premature cell death.

i havnt had cell death but im going to rewire my battery tommorow and see if it helps reduce voltage sag at all.

from the intial picture i will be going from the worst to the best possible configuration, so fingers crossed.

[fechter]

On how many spots did you solder the busbar on the nickel? I'm working on a similar setup.

It is soldered continuously along the length. It took a lot of solder. I used hemostat clamps to hold the copper against the nickel during soldering.

[nardcox]

On how many spots did you solder the busbar on the nickel? I'm working on a similar setup.

It is soldered continuously along the length. It took a lot of solder. I used hemostat clamps to hold the copper against the nickel during soldering.

Yeah, also figured that out after i cleaned the tip of the soldering iron, it's pretty easy then...

[jdvmi00]

I always have a hard time wrapping my head around why it's important to connect the positive and negative wires to every cell in the parallel group when terminating the ends of the battery pack. I think it's so that the distance is shared for the current traveling to the terminal wires. Is this picture showing the terminating wire OK because the 3P cells are so close together and not in a straight line?

[fechter]

The picture didn’t come out.
Yes you want the current in all the cells to be nearly equal. With only 3p this won’t be hard.