Cylindrical cell assembly concept

[tylerwatts]

Hello forum.
I have 'outside the box' ideas to problems, usually in the shower/bathroom time. Just showered before bed and had an idea I want to run by you. Please offer constructive feedback if you have experience.

I've been trying to decide the most practical way to build a battery for my oddly shaped space (multi-trapezoidal motorcycle trellis frame) and figuring a 18650 cell would be most flexible. Then the task of welding got me worried and I started looking at kits like Vruzend offer. These are square however and sacrifice some ~12% packaging space utilisation and every little counts.

So my idea was, picturing how an assembly of Vruzend caps linked together makes a big plate, one on top and one on the bottom of a pack (layer). And these now have clamping bolts holding the plates together to lightly compress the cells. And the v1.6 kit is limited to 3.5A per cell, 5A max burst discharge.

Could I make/have made a large PCB with high current traces linking the cell terminals in the correct patters, 1 top PCB and 1 bottom, with soldered in spring terminals for the positive cell ends. Use insulating plastic posts between the cells to stabilise and position them, having a smaller dia pin/dowel shape on each end that fits through a predetermined hole I the PCB and every 5th or so post is threaded and screwed to both PCBs too and bottom to clamp the PCB over the cells.

This would position, clamp and wire the cells all in 1 and could even include BMS circuitry. Potentially a multilayer PCB could be used sandwiched between layers or just separate between pack laters with an insulating rubber sheet or similar. It also significantly reduces the layer height for multilayer batteries giving the highest cell density possible.

I recognise this means bespoke printed PCBs are required but would that be so much more expensive than the Vruzend type kit? The PCBs could be printed in set sizes and configurations with a simple interlocking connector system between them on a layer larger than the individual PCB. And stacking would work the same either way...

Alternatively, the shed builder version is using a thin insulating sheet material such as perspex and attaching terminals through this in a similar manner/design. But more tricky, riskier for faults and not as efficient use of space.

I appreciate your feedback and experience and if anybody knows or has capability to help me build a prototype I'd be keen to discuss this.

Thank you

Cheers
Tyler

 

[john61ct]

Reinventing a new solderless no-weld, wow ambitions!

Note that G10 FR4 is the laminate (fiberglass / epoxy) that top quality PCB boards are made from.

Super dense and strong for the thinness, more so than metal I believe tensile / shear / maybe puncture too,

commonly used to hold rigging under stress to the hull in boat building.

Not cheap in large sheet sizes, but FYI.

 

[amberwolf]

just a data point based on the problems poeple have reported with various interconnect schemes here on es and elsewhere:

the non-permanent-interconnect methods have a major flaw: they require perfectly even tension across all contacts to get perfectly even resistance across all contacts. and that' just doesn't happen in most cases.

so you end up with some cells with great contact to the system, and others with poor contact, sometiems no contact as things distort and flex over time.

the great contact cells provide the majority of the current in the system, and the others provide less to little to none.

so cells get out of balance, wear unevenly, so you wind up with a pack that is less capable (sometiems much less) than a pack that has permanent interconnects.

sometimes the pack doesn't work right from the start, and sometimes it works fine. sometimes it seems to work fine, but later problems begin to appear.


the mechanical engineering and materials needed to make things precise enough to get perfectly even contact and thus resistance across all the contacts on all the cells may make a pack with a lot of wasted space that you could ahve used for cells. if space is at a premium, a welded pack is much likelier to give you better results in less space.

 

[tylerwatts]

Thanks guys for the feedback. As I said, it was just an idea that kept itching so I thought to ask more experienced persons for input. I appreciate your responses.

I had thought of most of those responses, not necessarily solved them though. Maybe welding is just better and I need to man up and give it a go...

Regarding compression and even contact, this is why I mentioned soldering a spring contact to the positive position of cells on the board to take up unevenness in compression.

Regarding rigidity of the boards, the case should be the appropriate overall size to lightly and evenly compress the pack and maintain even pressure even after each layer is clamped on the cells. As a whole unit.

I'm sure there are lots of other considerations too and yes, it is ambitious if I want to find an alternative solution.

Does anybody have experience with Vruzend kits or know of any others?

Cheers
Tyler

 

[cecopiperov]

N. E. S. E. Battery kit is also solderless and although I do not have experience with either kit Vruzend or NESE, I like NESE more. I have not considered price, size, weight and energy density, because the battery I need is not that big, and I do not have to squeeze it in some indescribable form of space. I would be happy to learn about other kits and their pluses or minuses.

 

[john61ct]

The ability to test/monitor individual cells

and to easily replace sub-par ones

would be a huge improvement.

I have speculative imaginings that conductive adhesive would work just fine,

combined with electrically insulating, shock/vibration absorbing

and maybe thermally conductive materials like

flexible foam, sealant potting compounds etc.

And per-cell **wiring** (not strips or plates) brought to the outside

for accessibility of the BMS, balancers, hobby charging etc

or even enabling changing the xPyS layout without physically dismantling the pack

 

[amberwolf]

Regarding compression and even contact, this is why I mentioned soldering a spring contact to the positive position of cells on the board to take up unevenness in compression.

the catch with springs is that if there is current flowing thru the spring itself, if it is enough to heat the spring, it could cause loss of spring tension. so you wind up using "better" springs, which probalby cost more.

this also significantly increases the size of the pack, as you have to add the spring length (compressed) to the lenght of the cells plus the board thickness, etc.

Regarding rigidity of the boards, the case should be the appropriate overall size to lightly and evenly compress the pack and maintain even pressure even after each layer is clamped on the cells. As a whole unit.

I'm sure there are lots of other considerations too and yes, it is ambitious if I want to find an alternative solution.

nothing wrong with ambition...i just wanted to point out some of the many challenges (and disadvantages) of such a system, so you know what you're getting into.

Does anybody have experience with Vruzend kits or know of any others?

some people have had problems with the vruzend (and other such kits), specifically the stuff i listed before. of course, many people probably have not had these problems, but it might also be that they simply haven't noticed them, or that their use case doesn't cause them problem.

there's a number of "solderless weldless" or "no weld no solder" systems, both diy and commercial, that have been posted on es. one simple search that finds some (but not all) of them:
https://endless-sphere.com/forums/search.php?keywords=*weld*+*solder*&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search


another more expansive search (but more irrelevant results)
https://endless-sphere.com/forums/search.php?keywords=*weld*+*solder*&terms=all&author=&sc=1&sf=firstpost&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search

 

[Dak77]

Thanks guys for the feedback. As I said, it was just an idea that kept itching so I thought to ask more experienced persons for input. I appreciate your responses.

I had thought of most of those responses, not necessarily solved them though. Maybe welding is just better and I need to man up and give it a go...

Regarding compression and even contact, this is why I mentioned soldering a spring contact to the positive position of cells on the board to take up unevenness in compression.

Regarding rigidity of the boards, the case should be the appropriate overall size to lightly and evenly compress the pack and maintain even pressure even after each layer is clamped on the cells. As a whole unit.

I'm sure there are lots of other considerations too and yes, it is ambitious if I want to find an alternative solution.

Does anybody have experience with Vruzend kits or know of any others?

Cheers
Tyler

I have been using Vruzend v2.1 kits on a little chinese scooter (Uberscoot 1000w etc) for a bit over a year and they work really well in it , mostly because the assembled battery has very little room to move and what dead space there is , I've - foam board- insulated and reduced shock and vibration to almost nothing. The "barrel bolts" that keep compression on the caps are probably essential to this working long term because of the tendency of the caps to back off and crack over time. I ordered an extra kit so that I could 1. have spare caps to replace any cracked ones and 2 .have more barrel bolts to spread the compression evenly over the whole battery . It's a viable alternative to welded packs only if you can limit movement because they won't maintain sufficient contact in a rough environment very long . The obvious benefit to these , other than easy building , is to be able to easily inspect/test/replace individual cells AND you can make removable leads for charging/balancing individual sections with cheap hobby chargers . I don't have to worry about a failing bms destroying my battery . They have their place for sure , but I probably wouldn't use them for a vertically mounted battery on a bicycle or motor bike. Too much lateral force on plastic parts.

 

[tylerwatts]

The ability to test/monitor individual cells

and to easily replace sub-par ones

would be a huge improvement.

I have speculative imaginings that conductive adhesive would work just fine,

combined with electrically insulating, shock/vibration absorbing

and maybe thermally conductive materials like

flexible foam, sealant potting compounds etc.

And per-cell **wiring** (not strips or plates) brought to the outside

for accessibility of the BMS, balancers, hobby charging etc

or even enabling changing the xPyS layout without physically dismantling the pack

Conductive adhesive is a fascinating idea, I hadn't thought of that. Someone must offer such a product but probably not cost effective for DIY.

Thanks John

Cheers
Tyler

 

[john61ct]

There are many, and IMO cost is not an issue.

Clamping and cure times more so.

 

[tylerwatts]

Thanks amberwolf

I didn't mean coil spring though, was thinking of the simple spring used by vruzend. A coil would heat up significantly.

I'll review the search results

Cheers
Tyler

 

[tylerwatts]

Thanks guys for the feedback. As I said, it was just an idea that kept itching so I thought to ask more experienced persons for input. I appreciate your responses.

I had thought of most of those responses, not necessarily solved them though. Maybe welding is just better and I need to man up and give it a go...

Regarding compression and even contact, this is why I mentioned soldering a spring contact to the positive position of cells on the board to take up unevenness in compression.

Regarding rigidity of the boards, the case should be the appropriate overall size to lightly and evenly compress the pack and maintain even pressure even after each layer is clamped on the cells. As a whole unit.

I'm sure there are lots of other considerations too and yes, it is ambitious if I want to find an alternative solution.

Does anybody have experience with Vruzend kits or know of any others?

Cheers
Tyler

I have been using Vruzend v2.1 kits on a little chinese scooter (Uberscoot 1000w etc) for a bit over a year and they work really well in it , mostly because the assembled battery has very little room to move and what dead space there is , I've - foam board- insulated and reduced shock and vibration to almost nothing. The "barrel bolts" that keep compression on the caps are probably essential to this working long term because of the tendency of the caps to back off and crack over time. I ordered an extra kit so that I could 1. have spare caps to replace any cracked ones and 2 .have more barrel bolts to spread the compression evenly over the whole battery . It's a viable alternative to welded packs only if you can limit movement because they won't maintain sufficient contact in a rough environment very long . The obvious benefit to these , other than easy building , is to be able to easily inspect/test/replace individual cells AND you can make removable leads for charging/balancing individual sections with cheap hobby chargers . I don't have to worry about a failing bms destroying my battery . They have their place for sure , but I probably wouldn't use them for a vertically mounted battery on a bicycle or motor bike. Too much lateral force on plastic parts.

Hi Dak77

That's very helpful info, thanks. Yes, I would want the layers in light compression in a battery case also, even in the Cruze d kit they can move a bit and that would worry me.

Cheers
Tyler

 

[tylerwatts]

There are many, and IMO cost is not an issue.

Clamping and cure times more so.

Oh ok I'll do a proper search later, thanks. Enjoying the sunshine.

Cheers
Tyler

 

[Dak77]

I hope my post didn't come off as discouraging the use of Vruzend kits. They are actually fantastic in the right application. I've had no corrosion in over a year of high moisture exposure. It's insulated but not airtight. The contacts and nickel plated copper parts are still shiny and new. When using extra barrel bolts, a 14s7p brick is very rigid and stable. If you can put them in a non conductive enclosure, not shrink wrapped , they will serve you well in almost any use case. -- Now back to the OP - I like your concept a lot. I was thinking about your approach , and I think it would work, but by the time you get enough material to create the sufficient strength and rigidity, it will add quite a lot of bulk and weight. On the rear rack of a bike or a bike with a lot of space under the downtube, who cares eh .?.

 

[tylerwatts]

I hope my post didn't come off as discouraging the use of Vruzend kits. They are actually fantastic in the right application. I've had no corrosion in over a year of high moisture exposure. It's insulated but not airtight. The contacts and nickel plated copper parts are still shiny and new. When using extra barrel bolts, a 14s7p brick is very rigid and stable. If you can put them in a non conductive enclosure, not shrink wrapped , they will serve you well in almost any use case. -- Now back to the OP - I like your concept a lot. I was thinking about your approach , and I think it would work, but by the time you get enough material to create the sufficient strength and rigidity, it will add quite a lot of bulk and weight. On the rear rack of a bike or a bike with a lot of space under the downtube, who cares eh .?.

Yes you are correct I think. The trick is keeping everything stable and compact. Welding is both but it's almost permanent.

This adhesive idea has my interest though. Bond the cells to sheets in the correct arrangements and enclose them effectively and it should be golden. I'd use hex fasteners to hold the cells together and in position of course, not just adhesive.

My first concern would be heat is conductivity Vs current draw as heat usually destroys synthetic bonds that aren't chemical etches (fusion basically of dissimilar material, like welding). Then vibration resistance is key. Something like a double sided polymer adhesive tape would be ideal if it is conductive and heat resistant.

Cheers
Tyler

 

[john61ct]

My thinking is the conductor bonding to the cell ends is completely independent of, not relying on the physical compression for the integrity of the electrical contact.

That latter combined with a flexible sealant or foam, maybe potting compound, to cushion against shock / vibration.

Thus the cells do not need to be held in any particularly precise position, so a variety of case sizes & shapes can easily be accommodated, no standard layout needed.

 

[tylerwatts]

My thinking is the conductor bonding to the cell ends is completely independent of, not relying on the physical compression for the integrity of the electrical contact.

That latter combined with a flexible sealant or foam, maybe potting compound, to cushion against shock / vibration.

Thus the cells do not need to be held in any particularly precise position, so a variety of case sizes & shapes can easily be accommodated, no standard layout needed.

Yes I'd expect so also, though we should minimise tension force on the bond joint.

I'm not sure about potting or packing the cells. I'd want some air flow cooling as the battery will need to put out some higher peak current in my application simply because I can't fit enough battery in to minimise this. Well, I say so, depends how dense a pack I can design.

I had a lazy day on the lawn in the sun playing cards and picnicking with my girls today so might have a play in the garage and mock op battery shapes and sizes to work out space in the frame.

I'm liking the adhesive idea in principle. Permanent mechanical joint but still repairable.

Cheers
Tyler

 

[goatman]

im using 3/8 diameter 1/8 thick neodymium ring magnet with a 1/8 hole looks the one in this link but not exactly
https://endless-sphere.com/forums/viewtopic.php?f=14&t=104405&hilit=ring+magnet&start=25#p1526465

its the same size as the positive button and I can swing the battery around by the 12 gauge copper wire that goes through the hole but im looking for solid 8 gauge copper to act as an electrode, theres a smaller magnet that will take a 16 gauge solid copper electrode but I wont have them til may. then seal with a coat of plasti dip as the battery ends insulation. im trying to find a ebc-a20 tester then ill do a 1000 cycle test at 7amp discharge to see how the magnets hold up

the ones im waiting for are like the one in the picture but are 1/4 diameterx 1/16 thick x 1/16 hole. theyre light and cheap

 

[john61ct]

should minimise tension force on the bond joint.

Yes, wires not pulled away from cell ends, zip-tied with some slack.

> I'd want some air flow cooling

No, I'm going for weather resistant if not waterproof.

Thermal conductive & electric insulating materials, aluminum outer case like a heatsink.

 

[Dak77]

im using 3/8 diameter 1/8 thick neodymium ring magnet with a 1/8 hole looks the one in this link but not exactly
https://endless-sphere.com/forums/viewtopic.php?f=14&t=104405&hilit=ring+magnet&start=25#p1526465

its the same size as the positive button and I can swing the battery around by the 12 gauge copper wire that goes through the hole but im looking for solid 8 gauge copper to act as an electrode, theres a smaller magnet that will take a 16 gauge solid copper electrode but I wont have them til may. then seal with a coat of plasti dip as the battery ends insulation. im trying to find a ebc-a20 tester then ill do a 1000 cycle test at 7amp discharge to see how the magnets hold up

the ones im waiting for are like the one in the picture but are 1/4 diameterx 1/16 thick x 1/16 hole. theyre light and cheap

This is why I love this forum. So many good ideas

 

[Dak77]

My thinking is the conductor bonding to the cell ends is completely independent of, not relying on the physical compression for the integrity of the electrical contact.

That latter combined with a flexible sealant or foam, maybe potting compound, to cushion against shock / vibration.

Thus the cells do not need to be held in any particularly precise position, so a variety of case sizes & shapes can easily be accommodated, no standard layout needed.

If you're familiar with the compound they used in Ford ignition modules from the mid 70s - late 80s, I wonder how well that would work, if available ..

 

[john61ct]

That seems to have been tough to remove, maybe melts with heat too.

 

[AngryBob]

All the conductive bonding agents I have seen discussed, most did not conduct very well, most would not survive heat very long, most would crack with age and vibration. All had significant problems. Adhesive copper tape was tried with poor results. When used adhesive side UP, and folded around a magnet, kinda tweaked my idea.

Good conductors make poor springs, good springs make poor conductors, poor conductors get hot, heat kills a spring's tension.

Poron foam is non-conductive, but maintains spring pressure best over time.

Magnets maintain compression, but are vulnerable to lateral movement due to vibration.

Some compression packs used dimples in the copper strip for connection points. My first thought was that you need thicker foam, over the dimple, to maintain pressure, then I thought just fill in the dimple, next, how about a magnet in the dimple, sealed in place with some silicone or other just to completely fill the dimple, with poron over the top. The dimple keeps the magnet from changing position, the foam compression is a backup.

If you make a million bucks off the idea, send me a cheeseburger.

 

[tylerwatts]

Thanks is Angrybob. Don't expect to make anything off any ideas just now. Thinking of how best to build my pack. It's an odd shape and size so tricky to configure and fit enough juice into. I'm getting closer to the decision welding is the only feasible option to maximise energy density. Or perhaps I should say cell density. Chemistry will have to suit power requirements Vs range.

Cheers
Tyler

 

[john61ct]

Overall pack density is what counts in the end.

Power output is rating for amps, peak vs continuous.

Energy density is Ah stored, which determines range.

And weight is a different factor than volume.

So really, four measures of the different density types.