Panasonic NCR18650B 2.7kWhr pack for my Stealth Bomber

Sigmacom

10 W
Joined
Aug 8, 2012
Messages
79
Location
Sydney, Australia
Finally got around to posting the details on my Panasonic NCR18650B pack for my Bomber.

Most of my riding is done in National Parks and after my first ebike (Hyena kit) I decided that you can never have too much range. Going into wilderness areas range and climbing torque is much more important to me than max speed. I'm happy with a max speed of around 45km/hr. I bought the 2nd hand Bomber with a dead battery as at the time it had the biggest battery compartment around, in a good frame. I wanted the cells with the highest energy density, so chose the Pana NCR18650B (3.4Ah per cell). I know these cells wont tolerate heavy discharge (or recharge) so I planned for a 14P 47.6Ah setup. My max pack current draw is around 60A so at 14P this worked out to 4.3A per cell or 1.2C peak. Before building the whole pack I built a 1S14P module and put it in series with my first ebike's LiPo pack, with a Cellog attached.

1s14p.jpg


After doing a couple of discharge cycles I found that the cells performed well under my typical riding conditions tracking the lowest (0.2C) discharge curve as provided by Panasonic.

Having had a few individual cell failures in my Lipo pack I also wanted to be able to pull the pack apart without too much effort. Since the cells are tab welded I decided that each 1S14P module would be the minimum replaceable unit. All of the nickel strip is double layered and spot welded. It turned out that having 3 of these modules side-to-side fitted perfectly in the Bomber's battery box! Here are the drawings I gave the tab welderer for the assembly of each 3S14P block:

3s14p.jpg


These 3S14P blocks were then stacked on top of each other (with insulating card in between) and the tabs hanging over each end folded down and soldered to the tabs on the pack below. Each series connection is carried by 4 double thickness nickel strips this way. As the cell spacers slide apart removing a 1S14P pack for replacement would mean desoldering these tabs then cutting 4 nickel strips on each side. Not pleasant but doable if necessary in the future.

5 of these packs were stacked end-to-end forming a 15S14P pack laying down on a sheet of aluminium cut to size. Balance leads and discharge leads were added, and a 60A breaker mounted on the end :

balanceleads.jpg


The whole thing was then shrinkwrapped. As seen below the pack is about the same size as the original Bomber battery (yellow) and only weighs about 500g more:

comparison.jpg


I hooked the pack up to my home made 3kW load and pulled around 50A for several minutes - cells did not even get warm to the touch!

load.jpg


There was still room in the Bomber's battery box in the top right corner just under the seat post to fit the additional 1S14P test pack in series for a total of 16S14P. I cut a square piece of aluminium which was screwed into place under this pack to keep it in place:

toppack.jpg


Since I plan to use my android app to monitor cells using a bluetooth cellog I'm also working on, there is no BMS. I also charge the pack using two UNA9 balanced chargers (same as BC168s but 9S each) through a HMC Molex connection where the CA used to be (idea stolen from Hyena). As others have mentioned on this forum these cells cannot be charged at 0.5C despite what the Panasonic documentation suggests! Doing so will rapidly reduce their capacity. I charge them at about 550mA per cell as the UNA9 max out at 8A. The pack can also be bulk charged via the usual charging port at the side.

molex.jpg


With the original 5404 motor and rim I was able to get 130km out of this pack on moderate trails and taking it fairly easy, with a top speed hot off the charger of just under 50km/hr. I've just fitted a TC65 in a 17" moto rim for more torque (see separate post).
 
Sigmacom, you did a very nice job!

Fitting 2.7kWh in so small space is awesome!

To improve your battery life and cell overheat I would suggest to improve the current share between the cells. The actual connections you make on each 1s group make the cell closer to the center to suffer more than the cell around the group.

1s14p.jpg



The nickel have high resistance and the cell closer to the main serie connection will suffer and be drained at higher rate than the rest.



here is a representation to explain and also some ex of solution I suggest to resolve that:

file.php


Doc
 
I thought that stealth bomber runs on 24s lifepo4. If so you would need 20s for the same voltage as the stealth bomber. I think the fighter is running on 48V pack.

So did you purposely run lower voltage on the bomber or do you have a stealth fighter?

BTW, nice battery build.
 
That's just is test 1s pack. His design shares the 7 wide 14p across 4 conductors and he's running relatively low current.

Using such questionable high capacity cells for ebike use, which means highly flammable and potentially explosive in that sealed quantity, what I would want is some type of significant thermal monitoring.

Great work Sigmacom.

While we're still waiting on a significant advance in the power density of sufficiently powerful batteries, unless you ride where there isn't any sun, I'd carry a portable solar charger and schedule some nice breaks around midday with the longest during peak sun. Ideally a remote yet secure location and a pack split in 2 would work best. Run on one half depleting it on the journey there and part of the early ride. Take that half off the bike and set up the solar charger to fill it back up. Get most of your best riding done with a much lighter load, and come back to a fresh pack ready for the trek home.
 
That is some really great work. I would like to make a similar pack for long distance touring.

I don't know how much the original Bomber battery weighs, so 500g more doesn't mean much to me. Out of curiosity, what are the dimensions of the pack overall and how much does it weigh?

If you don't mind sharing a link to where you purchased the cells and the price, that would also be appreciated very much!

Thanks for posting!
 
Who'd you get to do the tab welding..

The welding was done by a cell wholesaler here in Sydney that my company already has a business relationship with. Unfortunately they normally don't do one-off jobs for the public.

Doc - as John points out that setup was only for my test pack. The main pack has a criss-cross connection of strip across all the cells in parallel and 4 strips forming the series connection between each 1S14P, and those 4 strips are spaced evenly across the 14 cells. When I loaded the pack to 55A I did do some voltage drop measurements on the double-thickness nickel strip and determined that this setup was ok. Also these Panasonic cells have relatively high internal resistance (between 50-70 mOhms by my testing) which reduces the impact of the varying (much lower) resistance of different current paths through the strip.

So did you purposely run lower voltage on the bomber or do you have a stealth fighter?

I intentionally ran lower voltage. For my style of riding on rough tracks 45km/hr is ok.

John - I take your point about temp monitoring. Having relied on cheap, large HK lipo packs in the past I'd consider these cells to be a safer option. I did get very close to building a pack using 20Ah EIG cells but physically they were going to be a lot heavier and not fit the Bomber space as well.

The total weight of the pack is 11.5kg and the main section measures 340 x 125 x 155mm which works out to about 235 Wh/kg.
 
Don't overlook the potential for a defective cell or tab weld that made one defective. Ideally I'd hit a pack with some cycles before assembly, but I end up relying more on self discharge and time to weed out weaker battery blocks. I always check balance upon arrival and mark them. Then I parallel everything and balance charge it before setting aside for a couple of months. Storing them full may give it a bit of hit to capacity, but compared to having to break down a pack to remove a bad one it's worth it to me.

11.5kg is absolutely awesome. I can't wait for the day that is cut in half and it's high power cells. I'm sure I will get to see that. I keep imagining our grand kids having 1kwh the size of a pack or 2 of cigarettes. Add in superconductors and magnets that don't need backing rings of steel and far better core material, and we'll have motors to match the batteries. Then we're talking airborne vehicles, but the ebikes sure will be fun too. :mrgreen:
 
The pack sat out of the bike for several months as the bomber went off the road for an unexpected protracted VBoxx repair, and lacing the TC65 etc. I did get to cycle the pack a few times and do the load testing then. Unless a cell goes completely short or open I'm not sure how easy a weak cell is going to be detected with 14 in parallel. This was always the trade-off for a pack like this vs a 2P setup with the EIG cells I was also considering. In the end I chose the higher energy density option banking on the fact that Panasonic cells have a pretty good rep. I'd love to know Tesla's failure rates on them.

I do run Cellogs across the whole pack in logging mode and with LVC alarms I have fitted to them.

I think 400Wh/kg will be not too far off and I can't wait! Mind you not too soon .. I have to get my money's worth out of this pack first!
 
maybe it would be a good idea to add a small waterproof sliding ventilation port that could be opened and closed on epoxy board case battery packs. perhaps a mini intake or extractor fan that could run off a mini solar panel & battery or using ram air pressure. ive always thought these epoxy board battery packs could do with better ventilation. condensation protection also
 
Welcome to the pure Battery case Engineering world! 8)

Assembling cells and installing BMS is easy! Make sure it is waterproof, that all cells will have the same temp to age the same over the entire battery life and have great air cooling... all that in a very small compact case are the real work!

Doc
 
These cells after few months of use won't show noticeable loss of capacity given 1.2C max current consuption. So i would say there is too early to judge something but another question is what we expect :p Li-ion ncr's are proven to be very reliable.
 
The pack is going great. The discharge profile still follows the 0.2C curve from Pansonic. There has been no detectable loss in capacity. I'm currently building a 'booster' pack out of the same cells - this will be 16S5P and only paralleled with the main pack on trips that require it. It will be stored in the rear pannier. Should extend range up to around 170km max.
 
I saw a battery company on alibaba that sells cell joiners for 3.7v 18650 cells that does not require welding and just clamps them together on the ends meaning quick connect and disconnect unlike spot welding. it looked like a cell holder bracket with terminal joiners built in. company says eliminating spot welding keeps the cells in better lasting condition by not marking the terminals and is better. makes cell replacment easier

if a cell was to fail after say 100 cycles is it ok to replace the failed cell and mix a new cell with used cells as battery companies say you should not do this mix new cells with old cells
 
Doctorbass said:
Sigmacom, you did a very nice job!

Fitting 2.7kWh in so small space is awesome!

To improve your battery life and cell overheat I would suggest to improve the current share between the cells. The actual connections you make on each 1s group make the cell closer to the center to suffer more than the cell around the group.

1s14p.jpg



The nickel have high resistance and the cell closer to the main serie connection will suffer and be drained at higher rate than the rest.



here is a representation to explain and also some ex of solution I suggest to resolve that:

file.php


Doc
What would be best Series current share connection with batteries side by side like picture ??
 

Attachments

  • 20160422_232737-480x853.jpg
    20160422_232737-480x853.jpg
    37.4 KB · Views: 3,034
Back
Top