Panasonic 18650B 3400mah

redorblack

100 W
Joined
Sep 26, 2009
Messages
142
Amazing what a couple of years can do... when I got into the lipo game with 18650 cells for my flashlights the Trustfire (don't laugh) flame 2400 were the best bang (hope not) for the buck cell out there. The Tesla Roadster was being built with 2200mah Panasonics. The Tesla S is being built with Panasonic 18650A cells at 3100mah. From what I've read the 300mile range pack is going to be built with the now available Panasonic 18650B 3400mah cells. Elon gave a speech recently about capacity gaining about 10% a year (real world, not press releases for lab potential gains). On a Tesla forum, some of the Roadster owners are almost begging to get their packs rebuilt with the new 3400mah cells. For me this new cell means about a 1/3rd jump in real world useful time over my present cells, for the Tesla guys it would mean a jump to over 300 miles range in a Roadster.

Now to just find the best place to actually get the real cell without paying some premium because the supplier sells to police and survivalists. I see some selling them from about 8-9 a cell but the players are selling for 2-3x that amount per cell... plus shipping/handling fees. Are the counterfitters already faking these? Anyone have a good cheap source for real cells? Might buy 8-10 of them if the price makes it worth it.

[moderator edit to fix spelling in title]
 
I saw someone from russia on here that did a pack build with that at $6.50 a cell.

Think I paid $10+ for mine (only got two for the flash lights)
Nice and legit, would be good for a light commuter.

Anyone know what Wh/Kg density of the cells are?
 
About 210 Wh/kg from memory, and their volumetric density is impressive too - close to 410 Wh/litre, even if you lose volume to the cylindrical shape.
 
What you guys are not taking into account is that these batteries are good for about 1C, tops. Otherwise you end up wasting craptons of those precious watt-hours into sag and killing the battery in short time.

Tesla may use these kinds of cells, however in a car application, the high C draws come in short bursts, not continuous. The car will usually ride along drawing about 0.25C from the battery, which is fine and dandy.

If you want an ultra long range pack with the lowest weight/size, at the expense of a high cost, then these are the cells! even the last gen panasonics are good too.

34-ncr18650b-1.jpg


At 1C though, you are getting sag to the point where you don't get that 3.4AH, you get more like 3.2AH. Also the cell is more like a 3.4v-3.5v nominal at that point.

So if you want to use these, you'll want to design your system for more like a 0.5C draw ( or less ) if you want to see something that looks remotely like the 210whr/kg figure you're talking about. Otherwise it's 190-160 whr/kg in reality :(
 
The Orbtronics 3400mah batteries ARE the Panasonic 18650B cells... They also have a hybrid cell for high current applications that is good for about 10 amps draw per cell, but a lot less capacity at 2250mah, the CGR18650CH Hybrid IMR.

The 18650 format doesn't make sense for most people doing seriously high current draw applications like badass bike battery packs due to the number of cells required to get the voltage/amperage requirements. But for Tesla, they've got their pack design down for safety and with the fact that they are building for range as well as power, they can get away with these cells since they are building packs with 6800 cells to store 53kw. I love this old paper on their battery pack... http://webarchive.teslamotors.com/display_data/TeslaRoadsterBatterySystem.pdf

What they do that we as home builders generally do not is incorporate layer upon layer of safety systems after layer upon layer of safety inspections. Before they even get to building a pack, the cells are all coming from the supplier as a run just for them, Elon gave a speech about what making contracts for millions of cells can get you. There is a reason the 18650 is the cheapest size cell with good density... it is the standard now for laptops/power tools... and Tesla battery packs. If you contract to design a cell just for your application, it'll cost a fortune... just ask the auto manufacturers going for large format cells.

If Tesla was to do a swap out program for early roadster owners that wanted the 3400mah cells, the pack would go from a conservative rated 53KW to about 83KW bumping range from 200 to 300 miles with the same weight and performance. Considering Orbtronics wants $21.99 per cell... a bit ridiculous at $149K plus shipping/handling, but I have seen a link where someone is selling them at $8.50 a cell if you buy 20 at a time... and that is a reseller so I'm guessing that a Tesla owner in theory would need to buy enough that they could probably buy them directly from Panasonic with a delivered cost around half that, and KNOW they were getting fresh real cells. If they can get the cells at $3.50 in that sort of bulk, you are talking $23,800 for 85KW worth of batteries... I'm curious to know if anyone has a clue what those cells might cost Tesla by the hundred thousand or more, if they can squeeze them down to that $4 a cell or lower at this point. This being the safest, uniform, spec matching cell on the planet at this time... not a cheap knockoff. Of course the other issue for a Tesla Roadster owner would be someone having to disassemble their battery pack and rebuild it... holy cow that would take some labor... or is that another set of machine operations that Tesla has already worked out?
 
neptronix said:
What you guys are not taking into account is that these batteries are good for about 1C, tops. Otherwise you end up wasting craptons of those precious watt-hours into sag and killing the battery in short time.

Tesla may use these kinds of cells, however in a car application, the high C draws come in short bursts, not continuous. The car will usually ride along drawing about 0.25C from the battery, which is fine and dandy.

If you want an ultra long range pack with the lowest weight/size, at the expense of a high cost, then these are the cells! even the last gen panasonics are good too.

34-ncr18650b-1.jpg


At 1C though, you are getting sag to the point where you don't get that 3.4AH, you get more like 3.2AH. Also the cell is more like a 3.4v-3.5v nominal at that point.

So if you want to use these, you'll want to design your system for more like a 0.5C draw ( or less ) if you want to see something that looks remotely like the 210whr/kg figure you're talking about. Otherwise it's 190-160 whr/kg in reality :(

Or use them for an extender pack to back up a small A123 26650 pack ? Then you could use smaller packs without sagging them.
 
Wouldn't be a good idea to mix the different chemistry.....


I really do heart the shit out of 18650 or any cylindrical form factor, real winner over pouched in regards to china manufacturing reliability.

I don't believe that commercial ebikes should have ended up so heavily on pouch based batteries, I mean it would have been fine if they had invested some time in case design/padding instead of 4mm wall with ridged edges that seamlessly transfers impact...

I got some of those cheap 18650 ABS battery holders for 4s (shame they won't come in any higher sizes)

curious to build a small pack for a commuter with spare cells, like the idea of simple interchange of degraded cells
 
etriker said:
Or use them for an extender pack to back up a small A123 26650 pack ? Then you could use smaller packs without sagging them.

3.2v cells will have a different voltage curve than 3.6-3.8v nominal cells, so that is not the best idea. Ultimately you would be pairing some 120-130whr/kg cells with some 210whr/kg cells but not getting the full discharge capability from either.

Your best bet is to use the right cell in the first place, or at least don't mix and match lifepo4 with other chemistries.

Yeah, panasonic does sell a high discharge variant of their 18650 cells, but those cells are very unimpressive.

The best bang for your buck, in terms of the lightest and smallest battery per dollar, is probably the dow kokam 'high energy' cells, which have 1-5 C ratings on them, and extremely long cycle lifes. The cost ends up being $1000/kw-hr though! I believe the panasonics end up being more expensive than that but i haven't done the math.

You can for sure get good current out of an 18650 pack, you just need to take some things into consideration with making the pack, and you need to oversize the hell out of it. So if you have a 40 amp peak load, you want a 80AH pack if you want to extract the maximum watt hours out of these cells. Figure out the cost of that pack and... yeah, you will suddenly understand why nobody is using these ;)

If we could only get them at the Tesla price, they could make a lot more sense, but the pack construction is rather tedious anyway.
 
Some of the Roadster owners are almost begging to get their packs rebuilt with the new 3400mah cells.


Hmm.. I would love to see the "old" 2.4Ah cells of all updated tesla battery becoming availlable for the crowd... just like the A123 20Ah pouch become from the Fisker karma :mrgreen:

I would love to get one of these 11x 5kWh sub packs like the that the tesla battery is using!
They are watercooled too

Doc
 
I have been doing this for a while now so I know it works.

Here is the thing.

If I need to go for a shorter ride like I do lots of times I just use the A123 pack.

If I want to go for long rides I parallel the laptop cell pack and get lots more miles. That is why I am calling it an extender pack.

The most basic design is 11s A123 and 10s laptop cells.

I also do not like huge laptop cell ebike battery packs.

This is not something new.

6 months ago I posted this.

My first parallel battery ebike test rig. :)

http://www.bikeforums.net/showthread.php/796000-Homemade-li-ion-ebike-battery-pack-from-recycled-laptop-and-Dewalt-battery-A123-cells?highlight=

neptronix said:
etriker said:
Or use them for an extender pack to back up a small A123 26650 pack ? Then you could use smaller packs without sagging them.

3.2v cells will have a different voltage curve than 3.6-3.8v nominal cells, so that is not the best idea. Ultimately you would be pairing some 120-130whr/kg cells with some 210whr/kg cells but not getting the full discharge capability from either.

Your best bet is to use the right cell in the first place, or at least don't mix and match lifepo4 with other chemistries.

Yeah, panasonic does sell a high discharge variant of their 18650 cells, but those cells are very unimpressive.

The best bang for your buck, in terms of the lightest and smallest battery per dollar, is probably the dow kokam 'high energy' cells, which have 1-5 C ratings on them, and extremely long cycle lifes. The cost ends up being $1000/kw-hr though! I believe the panasonics end up being more expensive than that but i haven't done the math.

You can for sure get good current out of an 18650 pack, you just need to take some things into consideration with making the pack, and you need to oversize the hell out of it. So if you have a 40 amp peak load, you want a 80AH pack if you want to extract the maximum watt hours out of these cells. Figure out the cost of that pack and... yeah, you will suddenly understand why nobody is using these ;)

If we could only get them at the Tesla price, they could make a lot more sense, but the pack construction is rather tedious anyway.
 
That graph is rather interesting.

From doing crude block manual integration to calculate energy given from the ~0.15C (nearly 7hr discharge) discharge rate, the energy output is ~12.38Wh.

From the ~1.5C discharge curve, you're down to ~10.52Wh of energy per cell. That means you're throwing about 20% of your packs energy just into heating the cells up when you're at a measly 1.5C discharge. That's pretty astounding really.

Say you wanted to make a typical consumer ebike pack from those cells, say 10.2Ah and 36v (that's about typical for store-bought stuff right?) and you have a controller that draws say 700w (that's fairly typical right? I normally deal in 10's of kW's, so I'm slightly out of touch, but I think that's ball-park for normal ebikes.)

You're looking at 2C discharge for that setup. The graph shows us the cells' Ri averages ~80mOhm/cell. This means losing roughly 1/4 of your packs energy into heating itself rather than moving your vehicle forward. An imaginary data-sheet pack capacity of 372Wh, but a usable pack capacity (even in the very mild ebike setup described above) of ~280Wh.

That's a pretty brutal amount of energy to go just into heating your cells.

Just to extrapolate for some datapoints, we will look at difference C-rates for a 10s (37v) battery made from these cells, note capacity is irrelevant when using C-rate to look at sag, because if you made it 4p vs 2p, the discharge current would also be doubled so it balances. (this is part of why folks use the unit C-rate)

So, these are for a 10s battery from these cells (~80mOhm) at various C-rates.

0.01C= ~37v average
0.5C = ~35.64v average
1C = ~34.28v average
2C = ~31.56v average
3C = ~28.84v average
5C = ~23.4v average
6.8C = ~18.5v average (this would be the power peak discharge point for this cell, half the cells energy is going into heating itself, half the cells power is exiting at the terminals)
8C = ~15.2v average (note: Once your sag is over 50% of resting voltage, you are losing output power as you increase current on the cell, so in this cell, even for a dragster or something that just has to survive a few seconds of load, discharging past 6.8C gets you LESS power out even though the number of amps is greater.)


I would love a set of those cells for my laptop battery, but not for my ebike. Just to match my current deathbike's battery power needs (780amps off the battery at 116v) I would require about ~2860 of them ( for a 3C pack), at 46g/cell, that's 131kG in cell weight alone... Just to do what ~20kG of Nano-tech's are doing for me.

It makes me super happy every time we see another higher capacity 18650 though, and I look forward to my future laptops running longer, and I think it's awesome Tesla and some others found a way to use them in EV's.
 
cool
we can buy it from taobao
http://search8.taobao.com/search?q=Panasonic+18650B&commend=all&ssid=s5-e&pid=mm_14507416_2297358_8935934
 
Yeah luke

90% (random bs figure) are 36V10Ah.

The 18650 cells packs are mostly 4P 10S, which for the stock legal market <500W is totally fine.

These are the latest generation of rack batteries, big with the euro market (no comment of the diffrence between international markets :lol: and there various 'style' acceptances )
http://i00.i.aliimg.com/img/pb/515/039/485/485039515_673.jpg (image via alibaba.com)

Internal shot (mind my grubby fingers :) )
https://dl.dropbox.com/u/24071298/4p.png

And the best part is most of the components on the BMS are 'molex' (not sure if molex branded) connected to the rest of the compontes (keys/switches/leds) and they actually sunk the FETs :mrgreen: huzzah for quality

I'd like to see a decent headway rack based battery with a high quality BMS :cry:

Still need to see exactly how Tesla pack build with 18650 cells, gah

But yeah I'm the same, nano-techs for the bike and 18650s for the mai lights :mrgreen:
 
I'm looking to use the EiG pouch cells in my CRX. They pack 175 Wh/kg, 370 Wh/l and are good for 10C. This means a 37 kWh pack will take up about 100 litres and weigh 225 kg. I'm hoping to get well over 200 km range under normal conditions. Not bad for a 950 kg car (I hope) with awesome aero. Unless there is a better chemistry out there, I will go with these I think.
 
iperov said:
cool
we can buy it from taobao
http://search8.taobao.com/search?q=Panasonic+18650B&commend=all&ssid=s5-e&pid=mm_14507416_2297358_8935934
Is that a site like Alibaba, just all in Chinese? I can't tell what they are offering, much less who from.
 
yeah does it have an english version that is easy to use like alibaba express?

wouldnt mind order some
 
redorblack said:
iperov said:
cool
we can buy it from taobao
http://search8.taobao.com/search?q=Panasonic+18650B&commend=all&ssid=s5-e&pid=mm_14507416_2297358_8935934
Is that a site like Alibaba, just all in Chinese? I can't tell what they are offering, much less who from.
taobao is full chinesse local china market, number 1 of popularity in China.
 
$4.25 a cell now in 2015 in lots of 200.
 
http://www.aliexpress.com/item/200pcs-lot-freight-free-WEU-NCR18650B-3400mah-for-18650-panasonic-battery/2027764035.html

http://www.aliexpress.com/item/200pcs-lot-free-shipping-NCR18650B-3400mah-for-18650-panasonic-battery/2009099270.html
 
Say you wanted to make a typical consumer ebike pack from those cells, say 10.2Ah and 36V (that's about typical for store-bought stuff right?) and you have a controller that draws say 700W (that's fairly typical right? I normally deal in 10's of kW's, so I'm slightly out of touch, but I think that's ball-park for normal ebikes)

Nine out of ten North American E-bike users (that don't actually post about their experience on ES)
36V X 20A (from a 10-Ah 0.5C pack) = 780W max performance

Average ES poster
48V X 25A = 1200W (mix of lithium 18650's, and the most affordable 20C LiPo)

Average ES hot-rodder (the 2%)
72V X 70A = 3600W (18S/20S LiPo)

(This is pure conjecture based on completely unscientific data compiled by lots of reading. Nobody should believe anything I say, and you should do your own research before spending any real money)
 
liveforphysics said:
That graph is rather interesting.

From doing crude block manual integration to calculate energy given from the ~0.15C (nearly 7hr discharge) discharge rate, the energy output is ~12.38Wh.

From the ~1.5C discharge curve, you're down to ~10.52Wh of energy per cell. That means you're throwing about 20% of your packs energy just into heating the cells up when you're at a measly 1.5C discharge. That's pretty astounding really.

Say you wanted to make a typical consumer ebike pack from those cells, say 10.2Ah and 36v (that's about typical for store-bought stuff right?) and you have a controller that draws say 700w (that's fairly typical right? I normally deal in 10's of kW's, so I'm slightly out of touch, but I think that's ball-park for normal ebikes.)

You're looking at 2C discharge for that setup. The graph shows us the cells' Ri averages ~80mOhm/cell. This means losing roughly 1/4 of your packs energy into heating itself rather than moving your vehicle forward. An imaginary data-sheet pack capacity of 372Wh, but a usable pack capacity (even in the very mild ebike setup described above) of ~280Wh.

That's a pretty brutal amount of energy to go just into heating your cells.

Just to extrapolate for some datapoints, we will look at difference C-rates for a 10s (37v) battery made from these cells, note capacity is irrelevant when using C-rate to look at sag, because if you made it 4p vs 2p, the discharge current would also be doubled so it balances. (this is part of why folks use the unit C-rate)

So, these are for a 10s battery from these cells (~80mOhm) at various C-rates.

0.01C= ~37v average
0.5C = ~35.64v average
1C = ~34.28v average
2C = ~31.56v average
3C = ~28.84v average
5C = ~23.4v average
6.8C = ~18.5v average (this would be the power peak discharge point for this cell, half the cells energy is going into heating itself, half the cells power is exiting at the terminals)
8C = ~15.2v average (note: Once your sag is over 50% of resting voltage, you are losing output power as you increase current on the cell, so in this cell, even for a dragster or something that just has to survive a few seconds of load, discharging past 6.8C gets you LESS power out even though the number of amps is greater.)


I would love a set of those cells for my laptop battery, but not for my ebike. Just to match my current deathbike's battery power needs (780amps off the battery at 116v) I would require about ~2860 of them ( for a 3C pack), at 46g/cell, that's 131kG in cell weight alone... Just to do what ~20kG of Nano-tech's are doing for me.

It makes me super happy every time we see another higher capacity 18650 though, and I look forward to my future laptops running longer, and I think it's awesome Tesla and some others found a way to use them in EV's.

I am using these cells for a year now and LFP is right with hes numbers. I have 10S5P, 350w hub motor, it sags right at 1c mark. Nothing special, except for a laptop.

DSC09477

5s5p brick:
20130513_172054
 
rsilvers said:
http://www.aliexpress.com/item/200pcs-lot-freight-free-WEU-NCR18650B-3400mah-for-18650-panasonic-battery/2027764035.html

http://www.aliexpress.com/item/200pcs-lot-free-shipping-NCR18650B-3400mah-for-18650-panasonic-battery/2009099270.html


Thanks!
 
Back
Top