The Barncat Battery- No weld/no solder

Barncat

1 kW
Joined
Oct 26, 2020
Messages
368
I designed and built these packs a year ago. After a couple thousand miles of testing and considerable internal debate, as a service to the DIY community and as a social experiment, I'm releasing the design publicly.

You use my design and any information contained in this discussion thread ENTIRELY AT YOUR OWN RISK! Lithium ion cells pose serious shock, fire and explosion hazards!

It should be obvious to anyone skilled in the art or who has given some thought to battery pack building how this design works. And I'll say right out of the gate this is best suited to being produced on a large 3D printer. I don't have that tech and thus built these 100% functional prototypes as a very close tolerance woodworking project.

Soldering cells together is messy, inconsistent, and potentially heat damaging; spot welding cells requires some specialized equipment and lots of nickel strip and shapes that are not optimally conductive and it's permanent and slow going... and with both of those DIY approaches you still have to heat shrink the whole thing and put it in some floppy bag or something and create some half-assed attachment to your ride.

I didn't want to deal with any of that. Nor did I want to use hundreds of nuts and bolts or springs or magnets or pressure foam or other inefficient types of conventional cell arrangements.

With my optimized end-to-end design you quickly load your cells into a rugged electrically non-conductive case, adjust them, and go.

I don't believe a BMS is necessary. I'M the BMS. There is instant DMM access to each parallel cell group down one side to check voltage and at the ends for pack V.

The compression adjustment is done with reversed Dremel slotted nickel plated lamp screws riding in carefully plumbed threaded brass inserts, and locked with simple hex nuts with a socket that has a short handle welded to it through which a small screwdriver is inserted. The adjacent external busses MUST be taped off during adjustment to avoid serious shock.

The cells are only free to move longitudinally. The clearance tolerance on the inside case width and height is +.001"- .005". Obviously there is room on one case end for adjustment which allows loading/unloading cells. There are .032" polycarbonate insulating sheets between each layer of cells.

My main concern during early testing was possible shoulder shorts. That concern proved unfounded. Scrupulous prep of used cells is required however and all overwrap on the cell negative ends was removed with a sharp razor knife to maximize electrical contact. 100% burr free hand sanded .025" copper strip was used to make the parallel connections in the case of the green 20s5p battery. The 15s5p gray 21700 pack uses 3/4" flat braided copper instead for much higher amp rating.

So there it is guys. Works great, looks cool. It's a design feature not a detriment. Only limitation is it's gotta be rectangular...

So I'll see how the reaction goes here. If you make one for personal use and want to kick me a small donation- great.

The real social experiment is who's going to start producing it for sale without giving me a dime. That's where the honor system comes in. I'd work with a commercial developer and provide additional details.
 

Attachments

  • 20220213_110531.jpg
    20220213_110531.jpg
    5.4 MB · Views: 1,516
  • 20220213_110600.jpg
    20220213_110600.jpg
    4 MB · Views: 1,502
  • 20220213_110627.jpg
    20220213_110627.jpg
    3.4 MB · Views: 1,495
  • 20220213_112441.jpg
    20220213_112441.jpg
    3.6 MB · Views: 1,496
  • 20220213_112619.jpg
    20220213_112619.jpg
    3.1 MB · Views: 1,496
  • 20220213_112519.jpg
    20220213_112519.jpg
    3.4 MB · Views: 1,497
  • 20220213_113112.jpg
    20220213_113112.jpg
    3.1 MB · Views: 1,507
  • 20220203_124143.jpg
    20220203_124143.jpg
    2.3 MB · Views: 1,507
  • 20220203_124156.jpg
    20220203_124156.jpg
    2.1 MB · Views: 1,501
  • 20220213_111440.jpg
    20220213_111440.jpg
    3.1 MB · Views: 1,503
  • 20210124_145944.jpg
    20210124_145944.jpg
    2.8 MB · Views: 1,507
footloose- yeah. Ace Hardware. Unfortunately I can't find any solid copper lamp screws, and they may not exist. Don't have my lathe on site to make any.

SlowCo- I just use cheapo Yarlan 3A, 4A chargers and keep an eye on them. I usually unplug when groups are around 4.15v and rarely discharge below 3.3v or so. Groups on both packs end 30 mile rides within .00-.02v of each other. On charger a group or two might try to run a bit high particularly past 4.15. I have a 50W power resistor with alligator clips for that if needed. One could easily add whatever BMS to this design, I just keep it simple...
 
I started to read about the battery pack. But then saw those mid-drive bike hell is that what they're called Barncats ? I want to ride one of those the silver one.
I will get back to the battery thread and read it.
 
Barncat said:
I don't believe a BMS is necessary. I'M the BMS. There is instant DMM access to each parallel cell group down one side to check voltage and at the ends for pack V.

I'm the same... it's just not that difficult to manually watch what is going on.

The BMS seems for the person who really is not into this hobby very seriously.

Love the design.

In fact a few weeks ago on one of these threads I suggested this same idea of doing cells the long way in Series and putting spacers to do the Parallel.

Very good work.
 
SafeDiscDancing said:
The BMS seems for the person who really is not into this hobby very seriously.
Presumeably, ”This hobby”.... being diy battery building ?
Most users of , Ebikes, and batteries generally, ..dont have the knowhow, equipment, time, or inclination to monitor battery condition.
 
Hillhater said:
SafeDiscDancing said:
The BMS seems for the person who really is not into this hobby very seriously.
Presumeably, ”This hobby”.... being diy battery building ?
Most users of , Ebikes, and batteries generally, ..dont have the knowhow, equipment, time, or inclination to monitor battery condition.

There is clearly a "Consumer eBike Owner" who wants to buy something and not really know much about it but enjoys riding it.

Then you have the DIY crowd who builds everything and pretty much can handle manual battery supervision and care. Like this guy in this thread, he obviously is smart enough to build and manage everything. But he is pretty rare as far as number of people similar to him.

Every cheap Chinese thing I have bought has turned out to be complete crap and so I just decided that if you can just not buy those things you evade lot's of problems.

If you know what you are doing you do not need any outside "robot" to protect your battery... you can do it yourself.
 
That's pretty neat Barncat- what do you have between the cells to conduct power? Or did I mis-read, and each longitudinal "string" ISNT the series connection?

SafeDiscDancing said:
If you know what you are doing you do not need any outside "robot" to protect your battery... you can do it yourself.

Quit your Gatekeeping boy. 99% of this fanbase doesn't have the time or desire to individually test every cell- the needed entry-level of knowledge into the eBike space is high enough as it is without this behavior.
 
The Barncat wrote:

I don't believe a BMS is necessary. I'M the BMS. There is instant DMM access to each parallel cell group down one side to check voltage and at the ends for pack V.


All I'm doing is agreeing with him that a moderately interested hobbiest has no reason to resort to a BMS.

Think of it like Spark Plugs in a car.

Some cars had as many as 12 cylinders and that meant you had to routinely check 12 Spark Plugs to be sure everything was working correctly.

With an ebike you might have 12S or 12 cells in series.

We do the same things now... but instead of checking a Spark Plug for color and other features now we check voltage for a cell.

At the end of the day it's the same type of interest.

But using that same analogy today we ask:

"How many people even know HOW to check their Spark Plugs today?"

I'm the same way with my car and I have replaced the Spark Plugs and even recently replaced a power steering hose which was a real pain to do.

Some people are just naturally into their toys and hobbies... others just use their toys but don't really get into them that much.

The trend is towards FEWER people with hands on skills. (which some of us think is sad)

Don't we sort of just "assume" that any guy with the genius to create his own No weld / No solder pack solution would also be the guy who wants to know everything going on inside?

It would seem "out of place" if he strapped a BMS to this.
 
I am one of the “human BMS” ebikers....i dont want the complication of unreliable BMS circuits to fault find.
However, with commercial battery packs, i believe they should have all possible safety systems included....charge, plug, and play !
Many conventional ICE cars are now fitted with Lithium 12v starter batteries, as are caravans and camper trailers etc.
Those users will not expect to have to “learn” how to monitor and check these battery packs each time they use them, so automated charge control, bms monitoring, and safety shut down systems are essential.
Likewise for all those cordless power tools, and household appliances that are now commonplace.
Human battery monitoring and repairs are only for the likes of a few dedicated battery hobbiests like us.
These days even the cost is no barrier to simply buying a suitable size pack for our toys.
Replaceable cell packs are a hangover from the days of unreliable cell supplies.
 
Anyone is welcome to put a BMS on this design or not as suits their needs, so it's all good.

There is just a niche market for this, but clearly most of us on this forum build our own custom batteries, and mine solves basically all the no weld no solder issues in a very simple and effective manner, unless you need a triangle (a stepped triangle is possible but but then you're talking considerable additional complexity and potential failure modes...)

A printed case would look very slick, and there are a few minor refinements that could be made using that technique. Not going to divulge all of that at the moment.

The s/p layout is flexible from essentially 4-6s per layer and up to 7p. Imagine how quick and easy it would be to load your favorite cells into this premade case, turn a few screws, and bolt the pack on your ride. Plus it's not physically possible to make shorter more direct series connections... And you can quickly take the whole thing apart whenever desired to check, examine or swap cells.
 
Barncat said:
There is instant DMM access to each parallel cell group down one side to check voltage and at the ends for pack V...

...And you can quickly take the whole thing apart whenever desired to check, examine or swap cells.

Two very worthwhile features!

And, no soldering or spotwelding!

Bravo! :thumb:
 
Barncat said:
footloose- yeah. Ace Hardware. Unfortunately I can't find any solid copper lamp screws, and they may not exist. Don't have my lathe on site to make any.

For future iterations -- maybe these?
https://www.ebay.com/itm/373719571616
https://www.ezlok.com/ezknife-insert-400-M5
Screen Shot 2022-02-14 at 9.43.23 AM.png
Screen Shot 2022-02-14 at 9.43.34 AM.png
 
file.php


Before I start to add ideas let me say again... THIS IS AWESOME. :bigthumb:

One thought that pops into mind is to make this more like LEGO blocks that can be easily put together.

Just looking at those boxes makes me think you could do each "layer" individually.

Maybe it could be some injection molded plastic in a shape that allows for 5S and 5P in a single layer.

Then you close up each "layer" and stack them on top of each other and do all the connecting from the outside.

* Note : Adding genuine copper bolts is another plus.

Just ideas..

-------------------------

Another totally wild and off on a different tangent idea is a "front loading" frame centric battery.

In this bizarre idea the battery is loaded from in front of the handlebars and the fork must be of the triple clamp variety as is common for down hill mountain bikes.

The trick would be to design the frame and forks to be wide enough to allow the battery to be inserted and the fork tube would be REMOVED and only the triple clamps would have the bearings external to the frame. (and the forks tubes should be spread out a bit more too for extra clearance)

It might be a bit difficult to imagine this but sort of like that electric motorcycle with the air passage through the center is the general idea.

This means the frame ends up as the "stressed member" surrounding the battery which gets inserted from the front.

Sorry if I've gone too far into "wild imagination" mode... it happens for me...
 
footloose- thanks my friend for the research and links. M5x25 hex bolts would certainly work, a bit of overkill perhaps but the current iteration uses 8-32 thread, the equivalent of M4. Too bad copper has become practically a semi-precious metal... and a month wait from overseas.

I'd hasten to add that there's nothing wrong with the nickel screws and of course they won't oxidize or tarnish, but we all know copper is king of the heap in affordable conductivity. I suspect the result of swapping in copper would be slightly higher efficiency and a wee bit less sag on hard acceleration. Given my pursuit of high performance I may have to upgrade...

One thing I did do is set the brass inserts slightly proud of the Starboard (an HDPE sheet material) end piece such that they all contact the external copper buss plates- hence maximizing electrical contact. I could just make a new end piece with the M5 inserts and screw it back in place. Though that part and the insert installation requires pretty high precision to insure flat contact with the cell ends. One of the reasons my design works as well as it does is that the compression force required to get good series connections down the line is not that great, and once the nuts are locked they stay locked, and the lexan lid clamps everything down. The packs have handled street riding with all the vibration and bumps with zero failures.
 
SafeDiscDancing- thanks, and nothing wrong with thinking outside the box. You were on the right track the other week re linear layout of cells if you're not doing a welded pack. As I've mentioned above, my design lends itself to variously sized and/or multiple pack per machine configurations, but minimal weight and materials use will be achieved with a single case housing all your cells where possible. I would not try to incorporate this concept as a stressed frame member. It makes a great gas tank replacement.
 
Barncat said:
I would not try to incorporate this concept as a stressed frame member. It makes a great gas tank replacement.

Have you seen the "Stark" electric motocross battery design?

They use magnesium as a shell casing to hold free battery cells into place (no welding) and the overall battery with it's magnesium becomes an actual stressed member.

https://youtu.be/eGhaZDUKzdI

...I'm just say it is "a thing".
 
Love the simplicity and availability of the parts. 3D printing is nice and all but it's nice to see a design without the need for one. Copper is a softer metal no? Would a copper screw not strip or bend more easily under the clamping stress than nickel or another material?

Also, my unrequested 2 cents on BMS: I believe that you get a secondary or tertiary duty of the cell monitoring during load or the logging capabilities for amperage and troubleshooting. Choosing a $20 no-name piece of trash is obviously a gamble but some of the more well known ones have additional features for the hobbyist that wants to have the behavior of their machine and its powerplant in graphs and raw data. What is often left out when people talk about not needing the extra electronics of a BMS is the cells they choose are massively overkill for the application in most cases. Sure, if you have brand new cells known to be able to handle 5C or 10C, you have individually verified capacities and your vehicle never even surpasses 1C or 2C for charge or discharge, then a BMS may very well not be needed in that instance. But one could argue you shouldn't need a bank of fuses for the additional components run from your power source or a mechanical braking system if you have regen set up properly.

If you ride with a multimeter and can check each group voltage after a particularly bumpy section of riding to ensure no loss of conductivity or stick a thermometer in the middle and aim a gopro at the readout as you ride your most demanding route, you could adjust accordingly after the fact. The BMS is supposed to throttle or kill power when a damaging situation happens when you aren't watching (I personally can't watch the battery when I am riding). But hey, if you ride on your own property, charge in a shack with no valuables and no neighbors, manage your belongings however you like.
 
I'm not saying monocoque battery construction is impossible- it's just not applicable to my design.

Solid copper screws would be more than adequate for the light stress in this application and would be better conductors than nickel plated steel.
 
MorbidlyObeseKoala said:
If you ride with a multimeter and can check each group voltage after a particularly bumpy section of riding to ensure no loss of conductivity or stick a thermometer in the middle and aim a gopro at the readout as you ride your most demanding route, you could adjust accordingly after the fact.

Let's think of it in terms of pressure.

Just to pull numbers out of the air let's say there is 5 pounds of pressure holding the cells together in the series string.

What happens when those cells experience a sharp bump?

My "guess" would be that the metal to metal contact points would move slightly and you would see a "roughing up" of the metals at the point of interface.

Now the funny part about that is if there was corrosion before it more likely would get CLEANER as far as connectivity after the bounce.

So you need to imagine the reverse scenario.

What if active bouncing actually reduces resistance and improves performance?

-------------------

An example.

When you rewind a motor the ends of the magnet wire need to be cleaned and then soldered to the phase wires to go outside the motor.

I just did this recently and at first that connection was not perfect and the resistance was pretty high.

But then as I rode the motor the current gradually broke up any resistance and now it's better than at first.

-------------------

So my "guess" would be the harder you pushed this pack and the more bumps you went over the BETTER it would get as you more or less "broke it in".

But... you also probably need this "break in period" where you check the screw pressures to be sure they are in the range you want.

This is where a spring could make things better because it would maintain the pressure as a constant.
 
If you have worked with cells before,...even in flashlights....you will know that the ends deform (compress) with bumps or shaking.
That is why most “dry assembly” packs include some form of spring or flexible member to ensure constant contact pressure.
 
Hillhater said:
If you have worked with cells before,...even in flashlights....you will know that the ends deform (compress) with bumps or shaking.
That is why most “dry assembly” packs include some form of spring or flexible member to ensure constant contact pressure.

In this guys design he uses a wire mat rope like parallel cell connector which would act like a sponge to soften the interaction between the cells.

That is a "wildcard" addition to the usual bag of tricks and might prove to be a really good idea.

It does seem like the ideal is to do spring loaded ends rather than bolts in that you could know they will remain constant in pressure overtime.

Best to insert a copper contact "on top of" the spring, however, because current passing through a spring will ruin it quickly.

The flashlight spring works because the current is near zero.

-----------------------

Let me add that I have completely abandoned "cylindrical cells" in favor of "pouch cells" long ago and would never go back to all this connection nightmare nonsense.

With pouch cells you simply pick the ah size you want and then bolt them together.

It sort of erases all concerns about connectivity.

And use real bolts with self locking nuts... torque them hard when you install them.

No pop rivets... the aluminum will loosen and it fails quickly. Use steel bolts and nuts and washers.

And if you really want to get a solid connection add some copper plates in the mix.

For compression I always wrap each pouch with a single thin layer of fiberglass which eliminates the puffing.
 
One reason I believe some type of spring element is needed in pressure-contact packs is because the cells expand and contract from heating and cooling cycles.

The negative end can dimple slightly and still have a reasonable amount of contact, especially if the contact is soft enough to deform slightly and match the shape. However, the positive nipple is vulnerable to significant dimpling and partial collapse, leading to a contact that is not flat.

The pressure that holds the pack together should be concentrated on the shoulders of the cell, top and bottom. Whether the spring is some type of thin Poron closed-cell foam or a metal leaf spring. If using a leaf-spring or coil-spring of some type, spring-steel is the best material, which is horrible for a conductive contact.

Therefore I believe a copper sheet is the best contact. It can be nickel-coated or zinc-coated for corrosive environments. Both coatings are easily done in your home garage.
 
I've not done any experimenting with pouch cells. They're heavier and you still need to safely contain and attach them to your ride. Last year I thought I'd just buy some big lipo bricks but they were unavailable and/or way $$$...

(Edit- spinningmagnets, we wrote our posts simultaneously)
Re my reversed screw contact system: it works, so there's no arguing with that :). Sturdy, optimal surface area/electrical contact, and it utilizes the inherent but subtle "springiness" of the positive cell cathodes. Fortunately they can be safely compressed in series without passing (or probably even nearing) their yield point. This works both with the thin copper paralleling strips or the even more vibration damping flat braid copper. Trying to use springs instead would introduce unnecessary complexity and poor electrical contact.

The sequence is: load all cells and parallel strips in the case, screw the clear polycarbonate lid down snug but not tight, tape off adjacent busses so you don't shock the hell out of yourself, turn a screw til resistance is felt then back off and repeat a couple times to preload the series string- then turn an average of 3/8 further and lock the nut. This takes skill and is done strictly by feel. When that process is complete screw the clear lid down tight.

The whole very close tolerance case, of Starboard, poplar sidewalls, and lexan is excellent at vibration damping. There should be nothing but microscopic cell movement in use on the street.

If you're going to build one of these it's not rocket science, but it's a lot harder than it looks to hold the tolerances required. Don't try it without vernier calipers. And don't load a cell backwards :flame:
 

Attachments

  • 20220215_124844.jpg
    20220215_124844.jpg
    2.8 MB · Views: 1,062
  • 20220215_124936.jpg
    20220215_124936.jpg
    3 MB · Views: 1,061
Back
Top