Update 2020-03-24
Most of the hardware has now arrived. Time to focus on integration and layout.
Arrived
BMS
The Neptune 15 arrived and as @ningning notes above, it does not come with separate charge leads. I’m wondering if these were present on an earlier version of the BMS and have since been removed to simplify manufacturing. Also, the dimensions are not perfectly accurate. The BMS comes covered in a protective plastic which adds approximately 1mm to the width. Additionally, there are screw heads on the top of the metal enclosure which (along with the plastic cover) results in a height of 14mm (as opposed to the 10mm in the specs).
BF1 Fuses
For fusing the battery leads (see below). Spec sheet here
Motors
eLofty motors and trucks arrived- heavier than I expected. These came with the new adapters (fewer pins) and (1) set screw per can.
Battery
30Q 18650 cells arrived. As noted above, I will be adding fish paper rings to each before fitting into the N.E.S.E. cases.
New
Fuse the Battery Leads
I realized that in making the battery a removable component that I had failed to protect the pack when it was disconnected with the other hardware in the enclosure. After much research I landed on the BF1 fuses linked to above. The current plan is to integrate this fuse directly into the positive battery lead and then heat shrink for protection. This should give me a pack protected from shorts even when broken down to just the N.E.S.E. modules and the main leads.
For my 12s4p battery I need coverage to at least 50.4v, and these fuses work up to 58v. Additionally, they are available in multiple current ratings which should fit all variety of packs. In looking at the spec sheet I decided to go with 60A fuses since they last for 360,000s (or 100 hours) at 100% duty. Even if I were to hit that peak current every once in a while (I’m not going to run my 30Q cells to more than 15A each) it should give me a very long life. At 200% duty it takes 3s for this to pop, and at 300% duty it pops in 0.3s. So (if I have this spec’d right) this fuse should pop before the Antispark dies down stream and well before the cells hit thermal runaway.
An interesting thought occurred to me here: if you were REALLY paranoid you could use these fuses as replacements to the bus bars in an N.E.S.E. build where each P-group would be fused individually. Since the bus bars are 22mm for 18650 and 26mm for 21700 you would only need a little extra spacing between each P group to tie these fuses in directly.
Charge Connector Wiring
In looking at the limited space I have left in the enclosure and the lower current requirement for the charing leads, I decided to use an XT30 connector between the M16 charge connector and the charge / load leads on the BMS. These leads will use smaller 18AWG wire which should save space, be easier to solder to the M16 connector, and have a tighter bend radius.
Mounting Drop Through
I’m going to try to mount the motors drop through initially. It doesn’t appear that I will have any wheel scrub but the jury is still out on motor scrub.
Changed
BMS Battery Leads
Initially I was going to connect the battery leads on the BMS directly to the M5 bolts at the edge of the pack. After realizing that each time I needed to attach / detach the BMS from the battery would require unscrewing the main battery leads, I changed my mind. Instead, I have simply soldered the BMS battery leads to the leads of the “in” side of the AntiSpark. This optimizes a few things:
- It allows for disconnection of all electronics from the battery by unplugging (1) XT60 connector
- It allows for charging of the pack regardless of the state of the AntiSpark. This allows me to continue to monitor P-groups without the VESCs being energized
- It reduces the total amount of hardware inside the enclosure
Also note that this locks me into a charge only configuration for this BMS. I’ve really tried weighing the pros and cons of this approach, and I think I have an interesting solution (see the Research section below)
Balance Wire Routing
After looking at my original plan for routing the balance wires I realized that having all the wires encapsulated between foam pads would be a safer and more robust approach. In the process I realized that I actually needed 13 wires (and not 12) for the initial wire guides. Add to that the new fuse on the battery leads and the BMS parallel connection and the newly revised routing plan looks like:
Complete
Balance Leads
Really happy with how this turned out. My process:
- Cut and attach the foam to the bottom of the N.E.S.E. cases
- Put a ring terminal on each wire and run it through to the endpoint
- Measure the approximate spacing of where the BMS will land inside the enclosure, using that measurement to give a length to the balance leads
- Mark the balance leads with a Sharpie to provide an accurate length for each lead
- Cut the leads to length
- Cut the BMS ribbon connector to length
- Detech one ring terminal at a time to provide extra length in the lead so that the heat shrink would be far away from the soldering.
- Strip, solder, and heat shrink the leads (combining the ribbon wires where necessary to make it work with a 12p pack)
- Pull the completed balance lead harness back toward the N.E.S.E. modules and re-attached the ring terminals to the correct bolts.
Custom XT60 Splitter
I started by clipping the ends off of (3) already completed XT60 pig tails. Then I joined them together with a bunch of solder, sanded off the sharp edges, and isolated using electrical tape. After the 1/2" heat shrink arrives I will likely wrap the whole connector. This allows for the best routing of wires inside the enclosure
Extend VESC XT60
I needed to extend the XT60 connector on one of the VESCs in order to plug it into the splitter correctly.
What’s Next
- Adding the fuse to the positive battery lead
- Protecting the N.E.S.E. bolts
- Binning the 30Q cells for matching voltage per P-group
- Fish paper the 30Q cells
- Dropping the 30Q cells into the N.E.S.E. cases and screwing down the lids
- Plugging in the BMS to the balance leads for test bring up and programming of the BMS
- Soldering the charing leads to the M16 connector
- Test charging the pack
- Testing drop through mounting of the trucks and motors to the deck
- Finalizing foam placement in the enclosure
- Wire routing for the motor leads and sensor wires
Research
BMS Emergency Notification
The BMS is set to cut off “downstream” power if it enters a fault state. Since I will be using the BMS in a charge only capacity, this doesn’t directly affect board behavior
however
I thought if I could use the charge leads to light an LED, I would know that the LED would be lit only when everything was in optimal health. When the light goes out then I would know the BMS had detected an error, and I could safely slow to a stop and investigate. There are some very interesting debates on these forums over how much a BMS should intervene in the case of a problem. It seems that (at least at this point) a good way to know that the BMS sees a problem without throwing the rider might be a dummy light. Research will be focused on how to implement this or (even better) how to invert it so that the light only comes on when a failure is detected.