The Pain of My Range
Although I've not done that much riding on my Onewheel Pint (less than 300 miles) for the past three years of ownership, it is time to deal with the big reason I didn't do as much riding in the first place. At some point, I had a wild idea to ride it to work and actually tried it two times...yes, should have gotten an XR. However, I was stopped by an unavoidable problem. It is the well known thing called "range anxiety" and I have an obvious geographic issue that makes it so.
I live about 5.2 miles from my work as the crow flies and so the claimed range of my Pint should be close enough? The reality is that the path deemed safest to ride is also, no surprise, longer than the "as crows fly" range and my weight does not help (192lb) either. The best path work to home is 6.4 to 6.7 miles depending on which side of the street and how many intersections, however, all that and facing a steady incline of 125 feet elevation by the end.
So the home to work ride, in contrast, is not the issue. When I attempted it, the Pint was left with around 15% of charge remaining upon my arrival at the front door of work. Going home was a disappointing 5.3 miles of range with the board shutting down and more hilly terrain to go. In fact, the last 3/4 mile leg of the trip is a long climb of maybe 40 feet before a small down hill and again a climb around 30 more feet of elevation.
On the second try, I included a shortcut by way of a foot path that led to a pedestrian access stairs to reduce the roundabout bicycle path by nearly a quarter mile. This gave me a slightly shorter total and less than a half mile from home before the red light bar appeared.
How Much Is the Point?
For only $300 to $400 dollars I could purchase a high capacity pack from qualified third party builders and that would be that or for $1100 buy a Pint X, but then I'm cheap remember (actually I like to build stuff). So I looked into how much it might cost to build a high capacity battery since I already own the battery spot welder, load tester, power bench and have some knowledge about batteries.
Napkin sketch would add up to about $115 dollars and that is the batteries, nickel tabs, fish paper, glue, tape, wires and shrink wrap. If I'm not too far off, that is a savings of almost $200 at least. As well, I noted that If I count up the remaining supplies, another pack could be built for a little less than $100 dollars and maybe a third.
Research So Far
My first round of digging around for battery layout diagrams originally netted the very helpful, but limited article from "The Board Garage" which was what inspired me to even write about building this battery pack.
The only thing is that the article's details seemed just a little watered down, maybe to prevent legal issues I suppose? I noticed missing was at detailed battery wiring diagram or the information on the type of thermistors to use. Although, to be fair the article's intent was to repair an existing pack and not build one from scratch.
This minor omission may also be a measure to prevent anyone from just slapping together the parts to "manufacture" their own battery packs to sell, but not actually know what the dangers are if done poorly. However, from a dead pack starting point, the information would easily help anyone to piece together how to make a new pack as the original intention of the article was to reuse those existing key components.
From the article, an included pinout chart was a little tricky to understand on first glance. It's not immediately clear the orientation of the plug by the face or looking on, but it was a little more clear when I looked up the datasheet for the 26 pin connector (ZPDR-26V-S) and maybe when I disassemble the OEM battery box to be sure. The clues might be the locations of the unused pins and the images on the web shows that this chart was arranged in the manner to facilitate loading the connector pins (SZPD-002T-P0.3) into the plug during assembly.
Battery Planning
The original Pint OEM batteries are the 18650 size lithium Ion cells in the neighborhood of 2700 mAh each and arranged in a 15s1p configuration in a frame of eight over seven and the large capacity version to be constructed will consist of 21700 cells and not quite double at 4200 mAh each. All fifteen 21700s must be a careful arrangement to best fit the limited space of the original module case. The larger capacity cells claimed to deliver up to 15 miles of range, however, will effect the ability to read the power level bar on the board.
The lack of a diagram required me to come up with one based on images out of the article and other sources. A reasonable diagram of the battery cells looked something like the below layout, but not final. I would say this is the common arrangement of the cells. However, I'm not too crazy about the short positive output wire, nor the big jumper wire that is near the middle as well the next cell with a shorter jumper. There is the argument that I should not venture too far from the seemingly standard design, but I wonder about the jumpers found in the middle of the series of cells.
The "Tap" markers are for the BMS wires to be attached. However, from the information I could glean, one area was an odd arrangement between taps B6, B7, B8 and in my version I arranged it slightly different. Cell 7 and cell 8 from the article had their positive ends facing and require two longer connections to bridge to the next cluster, where I inverted cell 8 and only require one long connection and eliminated the odd bridge between cell 8 and cell 9.
At the time of this writing, I lack solid information on the type of temperature sensor or thermistors needed prompted a long search and looking at a few datasheets. However, the correct component must match the OEM version's resistance or Ohm range or the BMS will not accurately balance the cells. Rumored was that the sensor is NTCLE413E2103F102L and only after I build the pack will I know if this was right.
Tools I have
- Spot welder to attach the nickel strips
- Wire connection crimper tool for 26 gauge wires to .3 pins
- Temperature controlled soldering iron to attach the BMS wires to tabs
- Power supply bench for the charging via a built in battery charger mode to test
- Load tester to simulate use and inspect the performance
- Voltmeter to check all voltages of each cells during charge and discharge
- Heat gun to seal things up as well as other bit and tools.
- Lithium Ion cell charger that will take 21700 size cells to help assure that I'm starting out with all cells at the same levels.
For Sure the Firmware is Good
My version of the Pint is somewhat an early version known as a Gemini 5050. It will have the ability to use a larger capacity pack with no known limits. I understand that the later versions monitor the amperage and cut off the current ignoring the remaining battery power effectively treating all battery packs as a stock capacity. However, the lack of an accurate power level bar would be a small price compared to the gains.
The article did put me on the right track for the part number of the connector and I have the makings of a list of a few other components such as wire gauges and nickel connectors for welding.
When I actually get underway, I'll then have to show some images of the actual build.
As of so far (late 2024)
Battery to BMS socket ZPDR-26V-S $ .61 ea.
Pins for Socket (26 needed) SZPD-002T-P0.3 $ .10 ea.
Thermal sensor (4 needed) NTCLE413E2103F102L $ 1.68 ea.
Nickel .2mm x 10mm tabs $ 11.00
Fish Paper rings 21700 size $ 8.00
Fish Paper Sheets $ 4.00
26ga. Silicone coated wire (6 color set x 20 feet) $ 16.00
Molicel 21700 4200mAh (15 needed)INR-21700-P42A $ 3.95 ea.
Some Notes to Remember
- Unplugging the Power Pack from the BMS connector first and then the XT60 after.
- Arrange the 21700 cells in the battery tray and glue in the 3 groups for accurate fit.
- test possible insulation placement and the effects on fit.
No comments:
Post a Comment