Pint Battery So Far

Almost Ready to Build It

WARNING!   The following article is from a tinkerer who can't do anything the easy way.   The information provided is based on many assumptions and should not be followed too closely without your own research and testing. DO YOUR RESEARCH, TAKE PLENTY OF PRECAUTIONS , HAVE A PLAN, ALWAYS ERR ON THE SIDE OF SAFETY.  Any injury, critical malfunctions, explosions and death due to following elements of details or procedure contained in these articles are at the risk of anyone reading these article and attempting to follow this information without a reasonable amount of knowledge in constructing Lithium Ion batteries and a strong knowledge of electricity is doing so at your own risk!  Constructing large Lithium packs require an understanding and respecting of the dangers as well taking plenty amounts of precautions in constructing even a small pack.  Do not attempt constructing any type of battery packs without proper prior instruction or training and again, having basic electrical knowledge, properly layout the pack plans are important.  This is a project that I am researching and the total amount of equipment I own would make this worth trying.  The actual expense if one were to start from scratch would out strip the amount in savings that is listed in my blogs.  The information within these blogs are starting points for what is required to construct a battery pack and does not have all the answers, such as, technical details of voltage drop issues, internal resistance considerations, charging methods or structural techniques and so on. Those details depend on the application of the battery pack and exceed the scope of these articles.  

Model: Pint (not Mint, but the later sand color rails)

Hardware: 5314

Firmware: Gemini 5050

 Concept version 4

Coming from the general layout from the previous post, I have taken some liberties on the cell path or the "flow".  Here is my version of the 15S1P battery pack and one big or controversial detail is the start and end of the pack (#1 and #15) are right next to each other.  I admit I've not looked at many pack setups or access to any, but I will not hide my concern that this may pose a short circuit danger could somehow happen if the two cells touch or wires come loose and contact some other part.  This is a serious subject of, if the pack were to receive high current draw that could melt insulation or a shock that moves the cells around, there would be little to stop a unwelcomed contact and then a thermal runaway situation.

Current Facts and the Figures

The XT60 connectors are rated for a nominal around 30 amps and I believe the 12 gauge wire leads will be the same maximum at 30 amps as well.  The purported Pint amps is 10-15 during nominal movement and a max of 30 amps if in high torque situations.  So it is safe to say that the margin is narrow and the extra shielding is warranted.  The .2x10mm nickel strips were claimed to have around 25 to 35 amp handling and the limiting factor is the length of the the strips also the number of welds per contact point.  Six is recommended and I would say that is the minimum.  I read somewhere that a single spot weld would carry about 9 amps

If I were to proceed with the build in the above arrangement, the negative wire will have shielding along the length of CELL #15 and as well the positive wire secured in the same manner.  However, in the extreme scenario of failure, I would have both output wires passing along a curtain short-circuit.  And the perfect storm is the constant 30+ amp draw for a critical length of time.  The saving point is that I'm looking to race the Pint nor do I intend to go very fast.  The published peak amperage is 28 amps at 16mph while on an average maybe 7% grade.  Sadly, I my ride may have up to 12% and some small spots or at 100 feet is a elevation of 12 feet and a potential max amps of 30.

Feeling Out the Flaws

The concept of the placement of the positive lead along the side of a cell was to reduce vibration damage, as reported in a few forums.  The run of the negative lead is now longer as well and would run along the same cell, but towards the lid of the module.  A concern is the increased wire resistance due to the length, which I am very aware of even it the amount is considered negligent (0.00132 ohms per inch) or an increase from .0013 to about .003 ohms.  If I keep the positive wire short, I'll remove the potential issues of heat, but still have the issue of damage from vibration and it is true I could make a wraparound lug to solve that.  The potential heat that could melt the battery cover or even the Pint module is a consideration to solve sooner than later if it does exist at all.  This is why I will do a load test to prove these concern one way or another.

Example of a short wire to the XT60 - not to scale, but you get the idea.

Another design decision is the cells are grouped as eight cells on the upper with shielding to prevent wear and abrasion damage against the group of six cells beneath.  Cell number 7 is the odd cell due to the location and its need to jut out at the bottom of the pack that will cause some loss of rigidity of the overall pack.  

Prior to receiving the 21700 cells, I decided to modeled the pack as accurately as possible to get a sense of gap and placement of connectors and insulation.  The cells that meet end to end have folded connectors (Cell 1 to 2, 3 to 4, 5 to 6) and will benefit from a compression tape strategy to give increased surface contact for more stable amperage transfer.  

The last concern is the one dangling cell number 7 and although the same compression can be applied to stabilize it, the needed jumper to the lower group of six is a weak point.  The hope is to use a set of strips of nickel to form the connection instead of a wire.  This allows the connection to be flat against the cells and pack cover and removing a failure point where solder spots could push into a cells jacket.

Still pending for when the cells become available, a possibility of staggering the batteries before spot welding and reducing the width be a tiny 2 to 3 millimeters.  This will only be necessary or useful if the batteries are too tight.

Latest Supplies List:

• Nickel Tabs .2mm x 10mm
• Fish Paper rings for 21700
• Fish Paper strips for high abrasion spots
• Kapton 1/2", 1", 2", tape to secure wires
• Hot melt glue to secure cells and positions
• 15 count cells - Molicel P42A 21700 4200mAh 45A Battery
• 4 count thermal sensors- NTCLE413E2103F102L
• 1 count 26 pin plug - ZPDR-26V-S
• 26 count - SZPD-002T-P0.3 terminal pins
• 14 feet - 26 gauge silicon coated wire 
• 24 inches - 12 gauge silicon coated wire (12 inches red and black)