Nickel Strip ratings

[Ham]

Hi All,

I am just wondering if someone can point me to a chart or resource that will show me the current ratings for this:

https://eu.nkon.nl/accessories/battery-solder-strip/nikkel-batterijpack-verbinding-soldeerstrip-27mm.html

nickel strip 27mm x 0.15m.

I have asked Nkon and they don't know!

Thanks

 

[spinningmagnets]

If it is truly 0.15mm pure nickel, it should be fine for 10A per cell.

 

[Ham]

If it is truly 0.15mm pure nickel, it should be fine for 10A per cell.

Thank you, 10a per cell, continuous or peak? ...any tips on how I work that out on a pack basis? I have 196 cells in my pack...I assume I use the parallel count as it is the paralleling of cells that ups our amp hours?!

 

[kdog]

Spot welding repository,.. first page

 

[spinningmagnets]

196 cells is a lot. You haven't mentioned the voltage or how many cells are in parallel.

When using the ladder-shaped nickel bus material, the series and parallel current paths are both the same width and thickness. This adds physical reinforcement to the pack, since nickel is very stiff. That being said, the parallel paths can be much thinner and narrower, because the current they carry is very low under all conditions. The issue of concern is all of the series current paths.

If you are using 14S then the P-count is 14P. At 10A per cell, that's 140A. That ladder-bus strip can handle that, but the two ends of the pack cannot. You will need to overlay the positive and negative ends with thick copper as a collector.

 

[Ham]

Thanks guys.

We built the pack as two 6s16p so I can join them together in series to make a 12s16p pack on the bike and have a modular pack for future builds I have planned.

We ran the main pos and neg leads the full length of packs with 8awg and xt150 anti spark terminals. You are recommending we add a strip along the end of the pack also?

 

[spinningmagnets]

16P would be 160A at 10A per cell. You asked if that was continuous or peak. The answer relies on the duty cycle. If you acceleration phase is very short, you can go to a higher temporary peak, since the longer cruise phase allows the rest of the bus mass to absorb any of the heat that is generated, and to then shed it over time.

If this is for a racing battery, it only needs to survive one race, which means you can lean the duty cycle calculations towards the warm end. 160A is motorcycle territory, so...you posted in the E-bike technical section. There is nothing wrong with having a large battery on an ebike that can put out high amps. Doing that can mean that when running under a mild duty-cycle, the battery doesn't get hot, and it will last a very long time.

If this is a project where you are keeping the details confidential, no sweat, you can contact me in a PM. You can add a dozen temp probes on the pack, and then run it in the designed user-profile, then the data can reveal any hot spots.

If you find a hot-spot (high resistance, or normal resistance with too many amps at a bottleneck), then you can bridge that hot spot by overlaying it with copper wire. Warm is good (you can hold your hand/finger on that spot continuously), but hot is bad. Even if you are not concerned about the longevity of the pack, a performance pack should avoid voltage drop across bottlenecks.

What controller do you plan to use? and also, do you have pics of the pack and the 8ga wire on the pos/neg collectors, that you would be willing to share?

 

[Matador]

Matador's Conductor Strips Ampacity Charts:
(ratings are relatively conservative)




Enjoy,

Matador

 

[Ham]

Thanks Spinning Magnets, nothing is confidential by any means.

The pack is on an ebike yes and was built for pack performance and longevity as the small amp draw (relative to what it could theoretically pump out) is low.

I am currently running a Castle talon 120 and will be moving back to the HV160 at some point.

My initial question has come from hunting voltage sag in the system. I built the pack to try and eliminate sag and have only succeeded a small amount from hobby king lipo bricks, although overall performance is improved I still see more sag than I would have expected from this pack.

I have replaced all bike wiring for 8awg, connectors from xt90 to 150 and also have a bigger shunt now too. Sag looks to be caused by the pack itself now although the esc still has three internal xt90 connectors, I am yet to try the castle hv160 which is now with xt150 connectors throughout.



The image is of one of the 6s16p packs and does not offer the best view I understand, but hopefully you can see the copper 8awg wiring along the top running the length of the pack.

Thanks Matador for the table although I am a little confused as the nickel strip I used says 27x0.15 and your chart goes to 10mm? Would I be right in thinking then that I should measure the width of the narrowest section of mine?

 

[Matador]

Right,
27 mm strips is less typical of a width. Most use width of up to 10 mm. Your is almost 3 times wider (which is good).

But what matters is the conducting cross sectionnal surface area (in mm2)... In your case, 27 mm width x 0.15 mm thickness is 4.05 mm^2 CSA. In terms of gauge, 4.05 mm^2 is equivalent to about 12 AWG, that is, IF your strips were made out of copper.

But for nickel 4.05 mm^2 CSA (or 27 mm width x 0.15mm thick) is good for about 15 AMPS, with a resistivity of around 17.5 mΩ/meter of strip...

If these same strips 4.05 mm^2 CSA (or 27 mm width x 0.15mm thick) were made of copper instead of Nickel, they would be good for about 43 AMPS, with a resistivity of around 4.20 mΩ/meter of strip.

Matador

 

[Ham]

Thank you!

If I were to run a bead of solder across the series connections would that help or is that a silly idea?

 

[lionman]

They are not really 27mm though. That's the total width of the strip, but the interconnects between cells are more like 10mm. Measure the width of the interconnects and then refer to the chart above. I would say they are good for about 6A if they are 10mm.

I don't think coating the interconnects with solder will help much. Solder doesn't have the best current carrying capability. You would be better of doubling up the use of the nickel strips, or finding some thicker nickel strips.

To get 10A/cell you really want 10mm x 0.3mm strips.

 

[Ham]

Thanks Lionman,

I will see if i can borrow the spot welder again at some point and add more of these strips.

Would it be necessary to add them to all connections or just across the series cross overs?

 

[lionman]

Thanks Lionman,

I will see if i can borrow the spot welder again at some point and add more of these strips.

Would it be necessary to add them to all connections or just across the series cross overs?

Just the series connections. The parallel connections shouldn't see any more than an amp or so in a well constructed pack... They should only be subjected to more if there was a bad cell or cell connection.

What is the width of the series connections on those strips do you know?

 

[Ham]

Well I just pulled the spare roll I have and it is no more than 7mm at any one point...seems I have found my sag issues!?

 

[Neanderthul]

196 cells is a lot. You haven't mentioned the voltage or how many cells are in parallel.

When using the ladder-shaped nickel bus material, the series and parallel current paths are both the same width and thickness. This adds physical reinforcement to the pack, since nickel is very stiff. That being said, the parallel paths can be much thinner and narrower, because the current they carry is very low under all conditions. The issue of concern is all of the series current paths.

If you are using 14S then the P-count is 14P. At 10A per cell, that's 140A. That ladder-bus strip can handle that, but the two ends of the pack cannot. You will need to overlay the positive and negative ends with thick copper as a collector.

Hello,

This busbar design is beautiful. Would you be able to share the process and materials used to create those plates? I'm particularly interested in how the copper bar was attached to the (nickel?) plate. I have experience with resistance welding copper and it was a nightmare.

Thanks in advance.

 

[spinningmagnets]

Here's a link to a thread where snath shows how to make a dimpling jig for nickel.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=57810

The thick copper bar looks like it used eight rivets to connect to the nickel bus-plate. I don't know how I'd connect them, but I'd experiment with several methods.

 

[iampaulpease1]

Resurrection of the thread! Let's say you have a 10A cell discharge in a 6p configuration. Does that mean I need strips to handle 10A between P cells and with 6 cells in parallel I would need 60A between series connections? At 14S 6P I would need 60A? (Assuming 0 resistance). How do they make 30S packs? Some cells can discharge 30A per cell and 10P would be 300A. What can you connect the series connections with that can handle 300A? What would you even connect the P strings with for 30A?

 

[amberwolf]

What you need for P strips depends on how you connected the S strips, and how much current you will ever *actually* draw (not just what cells are capable of).

If you have S connections across the *entire* current path from every P cell to the next S group of P cells, then the P strips don't carry the current, just the S strips do.

If you use only certain points to make S connections, then the P strips have to carry the current from the non-S connected cells to that point, and so the P strips have to be capable of whatever that current is.

Example:
Let's say you will have an absolute max of 60A ever drawn from the pack by the system (60A current limit in the controller).
Let's say you have a P strip across each row of 6 cells.
Let's say you have two S strips from about the middle of each 3 cells in those over to the next S row of cells.
Those S strips have to then handle *at least* 30A each.
The P strips have to handle *at least* however much current flows from the two cells on either side of the S strip connections, assuming perfect current distribution that should be 10A...but it probably won't be perfect as cells and interconnect resistances vary, so actual loads will vary with them, making it better to up that by at least whatever percentage those will vary.

Let's say you have the same setup, but with S-strips that go from straight from each cell in one P group to the corresponding cell in the next S'd P group.
Now all the current from each cell will flow right thru that S strip and not have to go thru the P strip at all. The P strips are "backups" in case of an S strip interconnect failure, and to connect all the P group cells together to enable/ensure balancing / enable cell-group monitoring.

Let's say you have a non-rectangular layout, where some groups connect to the next with different widths of S connection (different numbers of S strips, etc). Now you have to use thicker strips on some than others, both for P and S, depending on which ones are connected in what physical arrangements.


So...it depends on how you physically build your pack.

 

[iampaulpease1]

What you need for P strips depends on how you connected the S strips, and how much current you will ever *actually* draw (not just what cells are capable of).

If you have S connections across the *entire* current path from every P cell to the next S group of P cells, then the P strips don't carry the current, just the S strips do.

If you use only certain points to make S connections, then the P strips have to carry the current from the non-S connected cells to that point, and so the P strips have to be capable of whatever that current is.

Example:
Let's say you will have an absolute max of 60A ever drawn from the pack by the system (60A current limit in the controller).
Let's say you have a P strip across each row of 6 cells.
Let's say you have two S strips from about the middle of each 3 cells in those over to the next S row of cells.
Those S strips have to then handle *at least* 30A each.
The P strips have to handle *at least* however much current flows from the two cells on either side of the S strip connections, assuming perfect current distribution that should be 10A...but it probably won't be perfect as cells and interconnect resistances vary, so actual loads will vary with them, making it better to up that by at least whatever percentage those will vary.

Let's say you have the same setup, but with S-strips that go from straight from each cell in one P group to the corresponding cell in the next S'd P group.
Now all the current from each cell will flow right thru that S strip and not have to go thru the P strip at all. The P strips are "backups" in case of an S strip interconnect failure, and to connect all the P group cells together to enable/ensure balancing / enable cell-group monitoring.

Let's say you have a non-rectangular layout, where some groups connect to the next with different widths of S connection (different numbers of S strips, etc). Now you have to use thicker strips on some than others, both for P and S, depending on which ones are connected in what physical arrangements.


So...it depends on how you physically build your pack.

So, I build the pack as shown below and it was drawn by someone on another thread I asked a question on. It is certainly not the easiest way nor the best way. (The way you described above seems much better for controlling amperage). I have been using this method because I can identify bad cells easier and just replace the string of p cells pretty quick and easy if needed. Unfortunately this doesn't lend itself to the method you mentioned above since every p string is individually wrapped to ease replacement. So I guess I am pulling the whole 50A (5p) through each of those p strings and that's why I had to use 10mm x 0.15 nickel plus some 20AWG wire to connect those p cells. I used 18AWG and double layered the nickel on the s connections but I guess that wasn't necessary since the Amps are going through every single p string and not adding up through the s connections. Is that correct?

 

[amberwolf]

The S connections *always* carry the full current of the whole pack.

If you have only one S connection between groups, then every single S connection in the pack carries *all* the current of the *entire* pack, all the time it's flowing. So if you have a 50A current draw, then every single S connection carries 50A.

If you have two S connections per group, paralleled, then each carries half (assuming perfectly equal resistances, which doesn't really happen). Then the pair of them in total carry the total full current of the pack, on every group's S connection.

If you have three, then each carries one third, but the three in total carry the total full current of the pack, on every group's S connection.

Etc.

 

[iampaulpease1]

The S connections *always* carry the full current of the whole pack.

If you have only one S connection between groups, then every single S connection in the pack carries *all* the current of the *entire* pack, all the time it's flowing. So if you have a 50A current draw, then every single S connection carries 50A.

If you have two S connections per group, paralleled, then each carries half (assuming perfectly equal resistances, which doesn't really happen). Then the pair of them in total carry the total full current of the pack, on every group's S connection.

If you have three, then each carries one third, but the three in total carry the total full current of the pack, on every group's S connection.

Etc.

Ahhh, I think I'm starting to understand why I keep burning through my connections. I have to redesign this based more on usefulness rather than convenience.

 

[amberwolf]

Ahhh, I think I'm starting to understand why I keep burning through my connections. I have to redesign this based more on usefulness rather than convenience.

That's an unfortunate problem that's pretty common. There are too many systems out there designed for other factors than whatever the critical one(s) are for it's usage scenario, so they don't do the job they're built for very well....

As long as you have the option to fix that for your system, I'd take that opportunity to look around at the many battery builds and see if there are easy-to-service options that also perform well enough, that are within your ability to construct. There are quite a few ways to connect cells up that have been tried out.

Depending on your specific physical constraints, and method / material / budget limitations, there may be an already-tested solution.

 

[thendless]

I am building a copper pack and have some confusion with the Matadors ampacity values:
I have a single piece of 0.15mm 35mm wide copper that will be used. It will be 35mm wide between series connections.

When I calculate the ampacity based on his table I get varying values depending on width of strip I use:
5mm (QTY 7)--> 12A x 7 = 84A
7mm (QTY 5) --> 16A x 5 = 80A
10mm (QTY 3) + 5mm (QTY 1) --> (21A x 3) + 12A = 75A

Shouldn't these be equal?

 

[amberwolf]

When I calculate the ampacity based on his table I get varying values depending on width of strip I use:
5mm (QTY 7)--> 12A x 7 = 84A
7mm (QTY 5) --> 16A x 5 = 80A
10mm (QTY 3) + 5mm (QTY 1) --> (21A x 3) + 12A = 75A

Shouldn't these be equal?

If the strip is wider but same thickness it can carry more current because it has a larger crosssection, just like using thicker wires.

If the total crosssection is the same, the current carrying ability is the same, except:

If the current has to get to other strips via the welds, then the welds are the limiting point for how much current can flow to those strips, along with whatever the resistance is of the surface contact between strips wherever they are actualy touching (they won't be touching on their full surfaces as they won't be completely totally flat against each other).