I imagine the lighter the cell the better but I was meaning what method you think works best.
One I made years ago that uses pressed copper sheet and rubber bands
1 Like
Ahh…
I’ve actually never seen a compression setup that can reliably handle the vibration/shock of PEV/LEV use. Either the cells bounce between the contacts, giving you intermittent connections (like NESE has done), or the “springs”/contacts eventually fatigue and snap.
All I can recommend doing is test…test…test…
2 Likes
I’m sure somewhere there’s a spec for maximum allowable compression. I imagine it’s pretty high - batteries are mounted in spring-loaded devices (flashlights, vapes, etc.)
But I keep my compression quite low. I figure the risk of damaging the cell and the resulting fire is higher than the risk of the cell momentarily disconnecting, especially in a parallel situation.
Also the cells are front-rear aligned on the board which is less likely to experience heavy shock loads unless you send the board down a set of stairs or ram a light pole.
I’ve never had an issue, but I’m a very calm rider.
Now I am going to retire my truck-based mountainboard for a suspension board because I think the constant rattle over gravel roads is just too hard on components. And a suspension board fell into my lap that probably only needs a new battery pack. 
1 Like
I just got a 3d printer with a big bed (sovol sv08) anticipating doing compression batteries for bikes. I’m planning to use foam and thru bolts and rubber bands. It’s been awhile and forget what foam is best. ?
I’d also like to know the amount of compression a 21700 can take.
I think with the rubber band approach it’s much less likely to lose contact from shocks as both contacts are pulled to the cell regardless…assuming the battery, bands, and contacts are all “floating” in foam. My band battery was losing contact before as it was clamped under a deck.
Magnets compressing copper foil is pretty well tested with bikes on endless-sphere. I bet it would work well on skateboards.
The can? Huge amounts of pressure across the entire top/bottom of the cell without problems. The can won’t buckle with any imaginable pressure levels from compression mounting.
But some cell top contacts (some Samsungs especially) are thin metal and on skinny “legs” and can easily be bent. Especially over time and with all that vibration/shock momentarily increasing pressure thousands and thousands of time.
Other top contacts, like Molicels, are really beefy. Test…test…test…
About the only risk from momentarily disconnecting is the arcing that will occur…thousands of times…and that builds up a layer of high resistance material that can cause lots of heat and really rob you of performance.
With some cells then providing more current than others you get more inter-cell current flow within a p-group and greater voltage sag along with shorter life for some cells. This will unbalance your pack faster and requires a verrrrry good BMS to keep control of and warn you (or disable the pack) when it can’t compensate anymore.
Battery_Mooch
About the only risk from momentarily disconnecting is the arcing that will occur…thousands of times
This is true.
But I think there are some mitigations, such as the disconnect time (really short) the current current draw (Hmm, maybe the amount of current currently being drawn), in my case quite low (My boards are limited to 60A).
Using the original springs which were these ones
I saw a signs of arcing, but with the new contacts on the few packs that I’ve disassembled I’ve found no signs of arcing.
So that’s good and bad. It means that arcing can occur, but it also means that with the right connector the chance of it happening can be reduced or eliminated.
In another project that is very high vibration but very low current draw I use these
on both the positive and negative side. Because the force required to compress the springs is very low I can bring them closer together, completely eliminating the chance of disconnect regardless (up to a point) of how hard they get hit.
But I do think there needs to be a bit of science done to find the optimal pressure of the contacts on the cells and then determine the amount of deflection you need on the spring contacts to achieve that. So that means precision distancing between the PCB’s which could be difficult in a home shop.
1 Like
The energy of the spark is determined a lot by how quickly the voltage changes (instantaneously for this…a worst case scenario) and the inductance of the circuit. That drives how much the voltage can rise to bridge the opening that formed when the connection opened and has a big effect on the spark’s energy. The circuit current definitely affects things too but other variables can really drive the spark hard.
The spark can last only nanoseconds but is still hot enough to create the classic “black dot” that indicates metal was melted and which creates the high resistance high spots interfering with good connection to the cell.
Excellent, that’s great to hear! So many people just ignore the tiny black dots but those are what cause the long term problems.
IIRC, for springs it’s 50% deflection as the target point. But during vibration/shock that can vary considerably so I guess testing is needed to see if the spring can reliably cover the range of motion experienced due to its stiffness (without disconnect and with decent metal fatigue life).
So much testing for this though!
The pressure on the cell contacts probably doesn’t matter on its own though, as a consideration or target range to stay within. The cells can handle a lot and what the springs/contacts do when bounced will be a lot more important IMO.
1 Like
if the battery is clamped with the connections free to move then just need to compare the mass and velocity (=force I think) of the connections alone in comparison to the contact pressure, as opposed to if battery AND/OR connections are how held, such as Doug did and commonly done, then the heavy cells themselves would be in the equation and much greater force needs to be put on the cells by the connections.
@DougM do your modules cause arcing?
cause we’ve cirrently came to the conclusion that it seems to be problematic if it aint spotwelded
Not that I have seen. The particular contacts I use are quite high tension and have a good amount of travel. So if you tension them carefully I think the chances of arcing are very very low.
I actually think the chances are zero if you tension them to just over half the travel distance of the contact. Then there’s no way they can physically disconnect. But i don’t do that because it is a tremendous amount of pressure and I never bothered to look up the maximum compression a cell can take.
1 Like
Ok, but now that I think about it if the arcing is caused not by momentary disconnect but by very high current demands and the connection between the cell and the contacts heating up that could be an issue.
These contacts are gold (flash), and the surface area is pretty good, maybe as much as if you added up all the spot weld points (unless you lost your spot welding mind)
But as I said, I top out at 60A which across 5P is 12A per cell. If you’ve got a 45A cell and you’re dumping the whole thing then you should probably bench test your assembly with a firebox handy.
1 Like
it’s gonna be 12S2P as the choochoo deck has very limited space
so guess it’s gonna be using higher Amps (?)
I don’t think there would be any burn marks if not arcing from a disconnect.
You’re probably right - overheating would show up in a different way than arcing. So a good archeologist would be able to tell the difference.