Another No Solder/Weld 18650 Build (Updated 03/09/2014)

spinningmagnets said:
agniusm, I think those bus-plates are wonderful, thanks for posting them. I do have one suggestion. The small finger that sticks up at the top center, the part that the slide-on connector attaches to? I would make it "T" shaped. I would slide a steel crimp sleeve over a wire that has a stripped tip. Lay the bare copper strands from the wire-tip onto the center of the copper "T", fold over the two arms so that the T is now an "I". Then slide the steel crimp sleeve over the folded copper arms and crimp it down.

The copper would carry the current, and the steel crimp sleeve would provide long-lasting clamping force. I haven't tried this yet, so I don't know what issues might come up, but...just a thought.

edit: unless, of course, that connection is only for the low-current balancing wire. In that case, I'm sure it will be fine.

Thanks, your edit is correct. Its for balancing wire only. It was posted long time ago on my thread, (https://endless-sphere.com/forums/viewtopic.php?f=14&t=63190&hilit=18650). Its a slow moving project, but hopefully it will materialize to something purchasable.

liveforphysics said:
It looks great on a datasheet, as to most silicones.

I personally avoid silicones when possible, due to crack-propagation and notch-sensitivity, meaning an otherwise tough and robust piece of silicone (or glass which is basically the same building blocks as silicones), can get a tiny surface scratch, and now the crack propagates right through the whole material (glass or siloxane polymers alike). The other funny thing silicones do is transport vapors through them though they seal liquids fine, they are not gas barriers (may or may not matter depending on pack and application). They also tend to be awful to get finished polymerizing siloxane surfaces (like silicone sheets) to bond with adhesives to anything well enough to stay located in situations where the clamp loading alone can't control the position of the material from extruding out and escaping it's clamp loads over time.

I do still use silicones once in a while (sometimes it is the best option for the application), but typically only as a last resort if no other material is suitable due to the above material issues.

I cant see it escaping clamping position in my case:

FB_IMG_1457974169894.jpg


but i need the sheet to be with adhesive on both sides and i don't know how good they are with silicone. Anyways, i like Poron, it cuts good with laser if you follow Rogers guidelines on laser cutting.
 
How about you try measuring the force (in grams) that the silicon has left after a certain time has passed, and compare it to the initial force (both in compressed status, maybe try different clamping forces/compression ratios to see how far you can go without damaging it/losing to much compression "power")? As this imo is the figure that matters and it's easy to measure.
 
twikis said:
How about you try measuring the force (in grams) that the silicon has left after a certain time has passed, and compare it to the initial force (both in compressed status, maybe try different clamping forces/compression ratios to see how far you can go without damaging it/losing to much compression "power")? As this imo is the figure that matters and it's easy to measure.

There is a unit for what you described, it's called 'compression set'.
 
Ah okay! I could only see the percentage thing mentioned here. So that means the force left after some time is exactly the same percentage as the recovery of the bumps immediately after unclamping?

I just thought measuring it this way may be more accurate, as he mentioned that the bumps stick to the surface he peels them off from, so this way may give him a more exact number.
 
Hi. You list your bus bars as being .308 (inches wide) by .010 (inches thick?). Is it possible that you meant to list the thickness as .10 inches (or a bit more than 1/16 inch thick)? To me that kind of looks more like the thickness portrayed in your pics. Also, any suggestions on where I might purchase preformed copper strip of these dimensions? I had no luck with McMaster-Carr, which carries 101 copper strips at only 1 inch or wider and 110 copper at only 1/2 inch or wider.
 
Hi msebold...

Here is a place in USA specialized also in EV power... (I have seen them at the Novi, Michigan Battery show in September 2016)
https://stormpowercomponents.com/
Storm Power.JPG
If you are interested you can look at the exhibitor list on the next Battery show coming in September 2017... you will find
many interesting companies in the EV industries...
http://www.thebatteryshow.com/exhibitor-list
 
Would these be acceptable for using on a pack?

https://www.aliexpress.com/item/50pcs-SMD-gold-plated-pins-2-Pack-18650-SMD-battery-holder-size-7-1X39-80X14-90mm/32640791799.html
 
JGaspar said:
Would these be acceptable for using on a pack?

https://www.aliexpress.com/item/50pcs-SMD-gold-plated-pins-2-Pack-18650-SMD-battery-holder-size-7-1X39-80X14-90mm/32640791799.html
Yes, if you want to push 1A or less per cell. Those springs are usually steel and at some current, spring will heat up and melt the holder, apart from resistive loses inefficiency.
 
Awesome !!!!

The connection bussbar remind me a bit of this :mrgreen:
https://endless-sphere.com/forums/viewtopic.php?f=14&t=60364&hilit=nickel+plated&start=25#p1235640

And that pack I made with bolts and bussbars a while back :
https://endless-sphere.com/forums/viewtopic.php?f=14&t=57810&start=150#p1213147
 
That looks good to me. What is the price and availability of that kit?
otherDoc
 
Vruzend looks nice! But nothing to make sure there is good pressure and contact, just zip ties? And really thin wires, maybe you can take two of those connectors at the end. What are the dimensions?
 
Snaths thread has become an index of sorts for the "loose cell" no-weld / no-solder packs (the first post), so this is the thread where I'd like to post a suggestion. One of the issues that has kept previous builders of these boxes from selling a product to the public...is the fear of customer error. Meaning, in spite of clear instructions, the customer inserts the cells wrong and there's a fire, or at the very least, a bunch of fried cells. I thought of two strategies.

One, have series bus bars that connect rows of cells from one side of the pack to the other...this way, all cells must only be inserted in the same direction. The polarity of the entire pack could be forwards or backwards, but if all the cells face the same way, the customer can't get it wrong?

Red is copper busses, green is insulation, blue is compressive poron foam

BatteryCellBusses3.png

Two, use the common method to keep the series busses short and compact, flip each paralleled row of cells so pos to neg is right next to each other. The trick is to have a hole in the center of any cell position on that bus so that...if a cell is placed into the slot backwards, then the positive "nipple" will not make contact with the bus?

Green is an insulating strip, blue is the poron closed-cell foam to apply pressure. The red is the cross-section of the 4P paralleling busses in copper, or nickel-plated copper...

pic below is 4P





If anyone is interested in this idea, I am stating in public that this idea is free to use to everyone without any compensation to me. I'm not saying this is a good idea, but if anyone likes it, please use it. The ebike world needs some examples of DIY loose-cell packs.
 
Swe said:
Vruzend looks nice! But nothing to make sure there is good pressure and contact, just zip ties? And really thin wires, maybe you can take two of those connectors at the end. What are the dimensions?
I got quick answer from them now:
The cost of each kit is $19

The cost to ship to Sweden will be

India post $16 (20-25 days)

Speed Post India $22 (15 days)

Express channel will be more expensive but will take 4-6 days. I can
find out if you wish.

Each cap is of 2cm x 2cm in dimension with a height of 2.5cm.

So if you join 10 in a row it will measure 20cm.
 
I'm not sure if there are spring action in this, but we could mabe use BCC-ND spring between plastic part and terminal plate....
These kinds of springs : https://endless-sphere.com/forums/viewtopic.php?f=14&t=60364&hilit=nickel+plated&start=25#p1237600
 
I e-mailed the seller. He had very detailed info and answers to my questions. He seems to know what he is working with. Very impressed.

It is not copper or nickel-plated copper, nor pure nickel.... but great design anyhow! I'm giving it a shot for sure, ordering 4 kits to experiment with it.

The 55 connecting strips are 8mm wide x 0.5 mm thick steel (he wrote stainless steel in the email).... So my guess is a resistivity of 173 milliOhms per meter of strip according to a previous post of mine :
https://endless-sphere.com/forums/viewtopic.php?f=14&t=84333&hilit=Ampacity#p1240370
Keep in mind, many of these strips are in parrallel so resistivity decreases...

The contacts are also SS. The seller is very knowledgeable and aware of the DCIR question, but the Steel metal is the only way to offer spring action of contacts (cantilever action I guess) without loosing spring recoil propreties (copper would just be inappriate for that matter).

I think the best of both world woud be brass terminal contacts, without cantilever legs (for spring action), but a springsteel spring behind it (sping between brass terminal contact and plastic holder... for good pressure), and nickel-plated copper for bussbar strips. The plastic holder seems perfect as they are !!
 
Matador said:
I e-mailed the seller. He had very detailed info and answers to my questions. He seems to know what he is working with. Very impressed.

Impressive you fellow. 24 years old, extremely communicative and anxious to provide good service. I'm going to also try these kits. I remain fascinated by the concept and believe it would open bikes to even more users with lower priced and repairable batteries. After all the budget batteries now sold are just glued, welded, and screwed. Claims of high end BMS are soon proven to be somewhat of a myth once we start opening packs. Perhaps we can afford better and more efficient BMS when self building. My welders work nicely but it requires far more skill and many more materials to build a welded pack. I did try the kits from the fellow in Florida and they're quite clumsy, in visual comparisons. It'll be fun to have them side by side. I'm still hoping for a design like Agniusm's to come to fruition. BUT that's been a long wait. Understandable but still an exercise in patience. The NTS WORKS bike and pack was interesting and the China modular that required trimming up the 18650 wrappers was interesting, but both are unontanium.
Vruzend- DIY Battery pack assembly kit for 18650 lithium ion cells without spot welding
https://www.youtube.com/watch?v=6M6Mm6jWqSk


Battery blocs

batteryblocmontage.jpg
 

Attachments

  • Screen Shot 2017-03-31 at 2.44.16 PM.png
    Screen Shot 2017-03-31 at 2.44.16 PM.png
    82.9 KB · Views: 4,704
I also ordered mine (4 kits to experiment, so 400 caps for 200 cells... might just start playing around with laptop cells to compare with my previous experiment using laptop cells and springsteel nickel plated....) Not sure yet if i'll try it on my VTC4.

In the meantime, I made some calculations to estimate total resistance of pack with Vruzend kit.
All conductive parts are stainless steel....
Made measures from video. I estimated that worst case scenario, the sping contact post are 3 milliohms each.
The interconnecting bussbars (2.5 cm long hole to hole ; 8 mm wide and 0.5 mm thick) are also stainless steel.

So if I make a 14S10P pack out of 140 Sony VTC4 cells I get the following (I don't want to get into details of calculation.... bit too complicated to explain, but using ohms law, resistance in series law, resistance in parallele law, etc...)
Each VTC4 cells is 23 milliohms DC internal resistance.

14S10P in ideal world where connection would have no resistance would yield at total pack resitance of 32 milliohms.
With Vrizund kit, that would come to 64 milliohms (14S10P with each 10P pack with 2 rows of 5 cells, and 5 strips to connect each 10P subpack to the next one).
With Vrizund kit, if I change the Stainless steel for copper interconnecting strips (keeping exact same strip dimensions), that would come to 41 milliohms.

Is this 14S 10P configureation (each 10P 2x5 cells with 5 strips interconnecting each subpack), the contact terminals are responsible for adding up 8.4 milliohms to the pack, while the inteconnecting strips add a total of 23.4 milliohms to the pack, but only 0.567 milliohms total if swapped for copper interconnecting strips of same dimensions (25 mm long x 8 mm wide x 0.5 mm long).
So pushing my BBSHD controller to maximum 30amps, id' see a 1.94 voltage drop on the original Vrizund kit with that 10P pack... Compare that to 1.23V drop with copper bussbars or with 0.96V drop in an ideal world (using supraconductors if a could, zero resitance !). Quite acceptable a priori.

I wont go into details. While this might seem like adding a lot of resistance to the pack, it shows because my cells are of low intrinseque resistance valures.

For instance, do the same with Panasonic GA cells (around 38 milliohms each cell https://endless-sphere.com/forums/viewtopic.php?f=14&t=73701#p1112249) and you get for 14S10P in this architectural configuration :

53 milliohms ideal world (resistance-less connection - zero resistance) --> 30A drop of 1.59V (10P pack)
85 milliohms with original Vrizund kit --> 30 A drop of 2.55 V
62 milliohms with Vrizund kit but swapped bussbars for copper instead of stainless steel. --> 30A drop of 1.86V (10P pack).

Pushing the example to the extreme, using old laptop cells (let's say 150 milliohms each cell), it becomes even less apparent :
210 milliohms ideal world (resistance-less connection - zero resistance). --> 30A drop of 6.3V (DONT PUSH 30A on laptop cells !)
242 milliohms with original Vrizund kit --> 30A voltage drop of 7.26V
217 milliohms with Vrizund kit but swapped bussbars for copper instead of stainless steel. --> 30A voltage drop of 6.51V.

Ordered 4 kits.... Dunno if it will be okay at 30amps, but it will most probably work wonders with lower power eBikes too (10-20A).

DISCLAIMER : These numbers are estimations I made. I migh be pleasently surprised by better conductivity. Yet, there still the question of proper contact pressure to be explored in practice. But anyhow I have 2 uses for 18650 : eBikes and small (<100Ah) Power banks...

BTW, I just notice, with my 14S10P VTC4 cell pack, if my 10P subpack are 1 row of 10 cells instead of 2 rows of 5 cells, and IF I then put 10 interconnecting strips between each subpack instead of 5, then the SS stips of the original Vrizund kit would add 11.7 milliohms instead of 23.4 milliohms. Fot copper bussbars, it would be 0.284 instead of 0.567 milliohms
That would mean :
32 milliohms ideal world (resistance-less connection - zero resistance). --> 30A drop of 0.96V (drop inherent to cell IR only)
52 milliohms with original Vrizund kit --> 30A voltage drop of 1.56V
41 milliohms with Vrizund kit but swapped bussbars for copper instead of stainless steel. --> 30A voltage drop of 1.23V
 
tomjasz said:
Matador said:
I e-mailed the seller. He had very detailed info and answers to my questions. He seems to know what he is working with. Very impressed.

I did try the kits from the fellow in Florida and they're quite clumsy, in visual comparisons.
batteryblocmontage.jpg

The idea of battery block was awesome, but I'm thinking neodynium magnets have very high resistivity, even compared to stainless steel... I remain skeptical, but would be open to beta test... They are on the expensive side though.

I just got sold on the Vrizund though. Pure indian genius... I find they are very ressourcefull and ingenious.
 
There is also the Tyva Moduloo....
Neat looking, but expensive !!!! (as expensive as the cells thermselves).
Also, Idk if it can stand much amp load.

[youtube]6vvHlFcwnvA[/youtube]
 
Matador said:
There is also the Tyva Moduloo....
Neat looking, but expensive !!!! (as expensive as the cells thermselves).
Also, Idk if it can stand much amp load]
What is the module price?
 
tomjasz said:
Matador said:
There is also the Tyva Moduloo....
Neat looking, but expensive !!!! (as expensive as the cells thermselves).
Also, Idk if it can stand much amp load]
What is the module price?

http://www.tyva-energie.com/en/moduloo-en/products/evaluation-starter-kit.html

I'm not sure since they only give the price on request. IIRC they used to display prices.... But they have starter kits with 1, 2 or 4 module (each module is 8 cells).... They also say on their page "STARTER KITS FROM 120 € ExVAT DELIVERY in 48 Hours". Dont know how many modules you get for 120Euros, but it's either 1, 2 or 4 modules (so 8, 16 or 32 cells worth of modules). Either way 120 Euros (128 USD) is way too expensive in my view... You still need to buy the cells !
To me it's not cost-effective for a 0.5 to 2 kWh range battery, although I like the design.

Also have a look here : http://www.conrad-electronic.co.uk/ce/en/overview/0501190/Battery-Clips-Holders/TYVA+moduloo

Tyva Energy is a small company stratup (less than 10 employees) based in Annonay (Ardèche) in France. The CEO is Thierry Claudel.
 
It seems that Some tyva modulo config for high power can give 240Amp for 10 second burst !

http://www.tyva-energie.com/images/Sales_Brochure_9.0__FR.pdf

Bussbar are nickel plated aluminium
Curved connector (vertically) stacked modules) are nickel plated copper.
 
liveforphysics said:
There is a unit for what you described, it's called 'compression set'.

Any idea which polymer has the least amount of compression set? The stuff O-rings are made from seems to be pretty good.
 
Back
Top