HK Lipo Bricks - what do I choose and why?

[ElectricGod]

I have purchased from and have had others also order with good success from "Shenzhen E-fire Technology Development Co.,LTD" http://www.cl-rd.cn/

Friendly, flexible, good pricing. I would recommend asking for pricing from a couple of different reputable Alibaba suppliers. Key to success with ordering cells from China is the supplier getting the packaging right, if you get a quote for shipping dramatically lower than competitors I would be asking why. I've heard all sorts of horror stories of batteries being seized by customs due to deliberate mislabeling etc. Ms Little Bear (that's really her name) from Shenzhen E-fire was excellent and all the cells I have ordered were packaged individually and to the letter of the law regarding lithium shipping.

Thanks for the info on the seller. I have contacted them. I know what you mean about lithium batteries not getting shipped correctly. The LIPOs I bought a while back came out of China and they shipped the first batch OK, but everything after that was envelopes and randomness.

How many extras do you think I should get? I never expect 100% success. I'll put all the cells through some basic tests to make sure they are OK, but that's to cull the weak ones that show up right away. That's what I want the extra cells for.

 

[Hillhater]

Careful guys, you are not comparing apples etc.
You are comparing a single cell 18650 with a multi cell lipo pack fully connected and ready to run.
Run the figures again and try using the more commonly used 6s, 5.0 Ahr, 20C HK lipo ($40, 765 gms)
And also compare finished "pack" costs for say a 48v, 10Ahr or 20Ahr pack fully assembled, wired , ready to plug and go !
....and try to use an 18650 with a rating that would allow a 50 A continuous pack discharge !

Last post I expanded the scope a little to the pack level. Lets do another example using your defined requirements

'48v' packs actually seem to vary a little in terms of cell count depending on who you're talking to. This is really a vestige of the days of lead acid. For ease of calculation I'm going to go with a 12s configuration so I can use the 6s pack you mentioned.

8 of the 6s/5ah packs will give us our 12s20ah pack. Each of these packs deliver 108wh for a total of 864wh. The lipo pack will be 6.2kg including the required harness to parallel pairs of packs and then connect in series. Total cost including harness ~$340.

To equate this using Samsung 30q cells we will require 7 parallel cells for a total of 84 cells and 907wh. Including spot welded nickel, shrink wrap, kapton tape, nylon sheet, foam and some 10ga wires to terminate to connectors this equates to bang on 4.1kg. These cells are available in small quantities from reputable retailers for ~$5ea, from chinese distributors in large-ish (>100 cells) for around $3.80 last time I looked. Including other consumables I would estimate pack costs of between $340 and $445.

Using the upper estimate the wh/$ works out to be 2.54wh/$ for Lipo VS 2.04wh/$ - 2.67wh/$ depending on your final cell costs. Of course there's a pretty substantial weight saving, not to mention the volume difference. To some people that's worth plenty, as the bulkier pack simply won't fit where they need it to. For other applications not so critical.

Another comparison between these two options is to compare kwh delivered over their lifespan. As a rule I would expect 250 80% cycles from the lipo pack and 500+ from the 18650's, assuming no early failures etc etc. That's based on real world as well as some synthetic testing. The Lipo pack delivers 172kwh VS 362kwh for the 18650's. This is of course a gross over simplification and frankly it's probably too kind to the lipo based on failure rates, but it gives you another point of comparison that really makes a difference if you use these things frequently.

Both packs will comfortably deliver 50a continuous efficiently.

You are under estimating the real "cost" of assembling these packs to compare to the "plug and play" status of the lipo bricks.
There is several days work in assembling that size 18650 pack , with many issues to consider along the way.
For an experienced pack builder, with the correct tools, welder , etc, it may be no problem, but others may not find it so easy. If you put value on your time/ skills, you cannot ignor the input required.
An "off the. Shelf", ready to run, pack of these cells is around $600 from em3ev !
EDIT..
Lunacycle 48v , 20Ahr, pack prices..
Lipo..$480....14.5 lbs
30q...$739..... 10.5 lbs

And , whilst a 7p pack of 30q may be capable of 50A continuous, at that discharge the working pack voltage will sag by approx 6 volts and you would need to run down below ~ 36v (<3v per cell) to get the full capacity.
The Lipo , in comparison, would sag only 1 - 2 volts and would still hold 40+ volts fully discharged.
I like, and use , 18650s ..as well as lipo, you just have to be realistic with the pro's and cons of each.

 

[Allex]

This is my first 10s20Ah or 20s10Ah lipo nest back when I was a complete noob. I had to do Y harnesses for mains and balancing leads. Took about a day to complete. A similar 18650 battery would take about a day and a half.

 

[Hillhater]

You can buy the harness's for lipo packs.
You cannot do that for a 18650 build...you either have to have the skills/resources to DIY, or pay someone to do it.

 

[Ohbse]

You are under estimating the real "cost" of assembling these packs to compare to the "plug and play" status of the lipo bricks.
There is several days work in assembling that size 18650 pack , with many issues to consider along the way.
For an experienced pack builder, with the correct tools, welder , etc, it may be no problem, but others may not find it so easy. If you put value on your time/ skills, you cannot ignor the input required.
An "off the. Shelf", ready to run, pack of these cells is around $600 from em3ev !
EDIT..
Lunacycle 48v , 20Ahr, pack prices..
Lipo..$480....14.5 lbs
30q...$739..... 10.5 lbs

And , whilst a 7p pack of 30q may be capable of 50A continuous, at that discharge the working pack voltage will sag by approx 6 volts and you would need to run down below ~ 36v (<3v per cell) to get the full capacity.
The Lipo , in comparison, would sag only 1 - 2 volts and would still hold 40+ volts fully discharged.
I like, and use , 18650s ..as well as lipo, you just have to be realistic with the pro's and cons of each.

Absolutely correct, I did not account for any time or tools required. I would not recommend a complete EV noob to take on an 18650 build as their first project, but likewise I would prefer to steer that same person away from a rats nest of fragile lipo.

None of those assembled pack prices surprise me in the slightest, as you mentioned there's a substantial amount of time and some tools required to do it justice. Those packs will also include a BMS and in the case of Lunacycles from memory a charger is included too. Very good value IMO.

While you're very correct about the differing voltage discharge curve (this is one of many reasons why I would recommend to anybody that doesn't have absolute physical size/weight restrictions to fit the largest pack they can fit) you've massively overestimated the sag on the 30q pack - my calculator says 2.5v pack sag at 50a. 20c lipo will be about half that, at least when it's fresh.

 

[ElectricGod]

You can buy the harness's for lipo packs.
You cannot do that for a 18650 build...you either have to have the skills/resources to DIY, or pay someone to do it.

Probably most people who are reading this thread this far into it either have a much better than average attention span or else they are DIY die hards and are going to build their own packs anyway. I know a lot of people take R/C lipo bricks and combine them together via the built in connectors to get the voltage and amperage they want. It works, but IMHO that's not really DIY. It's just throwing a few connectors together to make a larger pack. It does have the advantage of making re-configuring your overall pack for more current or voltage, much easier since its just a matter of unplugging and plugging things together differently. Personally, I look at what I am wanting to accomplish and at how much money I want to spend and then plan my pack build to accommodate those requirements. If I bought LIPO R/C bricks, I would take them apart so I could get direct access to the cells and then solder them together into the larger pack that is my end goal. It's just my opinion...but the fewer cross connects and what not you have, the better. We have all heard of the KISS principal. I learned that back in the mid 80's while getting my EE. The LIPO packs I am currently running on were individual 8000mah cells that I soldered together into 12S packs and added 6S JST balance connectors and a 12S BMS. They have worked very well with 1 minor cell failure due to a faulty BMS channel. I'm a personal proponent for fully DIY (build up your own battery packs from scratch), but am willing to use R/C bricks as a starting point.

LIPO R/C packs have the advantage of reasonably high capacities (5000 to 10000 mah) in 1S to 8S sizes for reasonably low prices. As a result they are easy to work with and meet a lot of peoples power requirements with little or no effort. For people that are picky, whiney engineering types (me) and need/want lots of voltage, current and capacity well that pretty much means get a battery pack built for you or DIY. For those kinds of people it might be worth it to build up a pack out of individual LIPOs or 18650's and to get a spot welder. It takes effort, planning and time to build your own pack, but it also means you know exactly the quality of the end result. A company that makes custom built packs will probably not have the advantage of having each cell individually tested before they are integrated into the larger pack. The cost and time involved would be prohibitive. A DIY pack on the other hand is probably built in your garage or den and you throw some cells on your testing rig before going to work and they are "tested" by the time you get home again. A company can't afford to spend the time to test every single cell like a DIY pack builder can. That is one of the true values for a DYI pack build. Since I tested every one of my cells before ever integrating them into a pack, I had reasonably good confidence that only good cells were going into the pack. Just about all manufactured packs are built with the expectation that a cell has a nominal voltage for that type of cell so therefore it must be good. That might be the case, but then it might not be. It's kind of like saying a 100mah cell is equivalent to a 10,000mah cell because they have the same output voltage. Ummmmm OKAY BUT one cell has 100mah capacity and the other has 10,000mah capacity!!! That's a huge difference! I don't know if a 10,000mah cell can actually get bad enough to only have 100mah of capacity ( I would have stopped using it long before then!), but they can definitely lose half their capacity or possibly never have it to begin with. I want to KNOW the quality of all the cells before they are ever built into a pack and that requires up front testing...no matter the cell type or chemistry. It's a total guess since the time you may spend as compared to the time I may spend varies widely and depends on the cell capacities and other elements, but each one of my cells got several hours of time on them before I considered then "good enough" for a pack build. Fortunately, I was able to test them in quantities of 6 cells at a time!!! I bought 60 8000mah LIPO cells, used 48 of them, had 8 that were marginal or not "good enough", 2 cells that were flat out unuseable crap and 2 more that I considered "good enough". 100% of the cells made into packs are running pretty close to the same with the exception of one cell that is always showing lower than average charged voltage due to a marginal channel on the BMS. 48 cells in 4 packs and approximately a 2% failure rate and that due to a BMS fault is pretty decent in my opinion. Anyway, testing cells before they ever get built into a pack is soooooooooooo worth it to me and that probably only happens for a DIY pack build.

 

[Hillhater]

That's a very thorough, professional , approach to pack building Egod.
However, we don't all want to spend the time or effort to "reconfigure" the Lipo pouches to optimise pack connections, and are happy to run with a premade or self made harness, and basic testing of capacity hold, capacity matching, IR matching, cell balance , etc .
One of the advantages of retaining standard Lipo "brick" modules, IR that if/when a cell fails or starts to show signs of problems, it can be swapped out quickly and cheaply. ( individual cell transplants can then be done on the brick if necessary, on the workbench). To some extent its effectively a modular pack, with disposable modules !
This is also what an 18650 pack could be like if we can find a reliable, workable, cheap, "no weld" pack assembly design.

 

[ElectricGod]

When is someone no longer considered a noob? LOL! I'm up to 10 watts now! hahahaaha.

I got a quote back for samsung 30Q cells at quantity of 180 for $975 including shipping. That's $5.42 per cell for a mere 3000mah...ouch! This is the kind of price trade-offs I did when I chose to go to LIPOs last time around. I paid $8.60 per cell for the 8000mah LIPOs I bought and got better than 2.5X the capacity on average. The LIPO's to be "good enough" had to be better than 7000mah when load tested and most were much better than that. They are losing capacity after 7 months of use so from the perspective of longevity they are not doing as well as I would like, but that is party due to the fact that they are low quality LIPOs and my cheap BMSs often charge to 4.2v or higher per cell which is killing cell longevity. I should have used a power supply that was closer to the maximum charged voltage. That would have prevented the BMS from over charging cells. My power supply put out 56 volts and I only needed 50.4 volts to get all 12 cells to 4.2 volts. That's .47 volts extra that could possibly get to each cell on average. The BMS should have handled that little bit of extra voltage and kept it away from the cells, but it didn't do a very good job of it. I don't know how well my current LIPO packs would be doing if they had always charged to 4.15v per cell. I'm still looking at 18650's, but unless I can get 3200mah at around $3 per cell and something like 20C discharge, LIPO is still looking better. when it comes to capacity vs price. Weight and maximum charge density for the volume of the cell are not necessarily my primary concerns. Unlike a lot of people space for my battery pack is not a huge concern to me. If I build into the battery bay only, then space is limited, but I have other options that make space for batteries more or less ridiculous (290 cubic inches).

 

[Hillhater]

You can get 3400mah 18650's, for under $2 cell....but their discharge ability is poor. (1-1.5c)
However, it may be possible to combine capacity from lightweight, cheap 18650's, with "power" from a
Lipo "booster". Pack !
See thread.."Hybrid Lipo/18650 pack for discussion.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=75705

 

[ElectricGod]

That's a very thorough, professional , approach to pack building Egod.
However, we don't all want to spend the time or effort to "reconfigure" the Lipo pouches to optimise pack connections, and are happy to run with a premade or self made harness, and basic testing of capacity hold, capacity matching, IR matching, cell balance , etc .
One of the advantages of retaining standard Lipo "brick" modules, IR that if/when a cell fails or starts to show signs of problems, it can be swapped out quickly and cheaply. ( individual cell transplants can then be done on the brick if necessary, on the workbench). To some extent its effectively a modular pack, with disposable modules !
This is also what an 18650 pack could be like if we can find a reliable, workable, cheap, "no weld" pack assembly design.

I agree...quick pack/cell failure swap outs are easy to do with ready-made R/C packs. And I have done R/C pack rebuilds for this reason too! It is a serious consideration to take into account when deciding to build packs rather than just connecting bricks together. I like clean battery packs and all those extra connections are adding mess and complications and possible points of failure.

Thanks for the compliment.

18650 quick cell swap out option. Have you seen these 18650 battery holders? I bought several of them so I could build 6S battery modules for charging and testing cells, but there is no reason why you couldn't use them for a large pack build. They hold the cells super tight. You can't just pull the cells out with your fingers. You need to pry the cells out with a small screw driver! I think the cells would stay put in the holders in an EV.... even in a mountain bike solution that had no shocks. The only concern is the extra space the battery holder takes up. Take two of these 4 cell holders and solder them back to back so you have a 2P pack. All the wiring would be on the sides of the holders. Put together whatever number of 4 cells in series to get the voltage you need and then go as wide as you need to get the capacity you need. Hmmmm...I might have just convinced myself to do this!
http://www.ebay.com/itm/371342603518?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

BTW...those holders are designed for a bare 18650. I have a few with a protection board on them and they won't fit.

 

[ElectricGod]

You can get 3400mah 18650's, for under $2 cell....but their discharge ability is poor. (1-1.5c)
However, it may be possible to combine capacity from lightweight, cheap 18650's, with "power" from a
Lipo "booster". Pack !
See thread.."Hybrid Lipo/18650 pack for discussion.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=75705

I don't consider a cell that has less than 3C even viable. I suppose if you build assuming you will always use less than 1C, then it won't matter, but who does that?! LOL! I sure don't. I want to typically run at 1 or 2C, but then have the ability to run much higher if needed. It's my opinion that a hybrid solution with lipo and lion cells is just a case of poor planning. If you need a "booster" pack, then probably you just didn't properly consider your current needs before building your pack.

 

[Hillhater]

I think you have not seen the possibilities..
Potentially, Less weight, $$'s, and size, than either all Lipo, or all 18650 pack of similar capacity and performance.!

 

[ElectricGod]

I think you have not seen the possibilities..
Potentially, Less weight, $$'s, and size, than either all Lipo, or all 18650 pack of similar capacity and performance.!

Yeah that's possible...or I'm too much of a whiny purist to consider a "mixed" solution. I want to do it right the first time or not at all and then over build significantly. IE: as a result I never see the need to create a hybrid solution. Truthfully, so I don't lose the investment in my existing LIPOs, I will likely pull those packs apart and then build them into 20S packs and then add more of whatever other solution to make up the rest of the capacity. That sounds rather like a hybrid solution to me! Hmmm...thinking about it more...I have 50 or 60 2600mah 18650's from various laptop batteries. I could probably get everything I need from them and my existing LIPOs to get me to my next voltage, current and capacity goals. LOL! Sounds like I'm going to be a battery whore after all!!! hahahaha! Use those battery holders I mentioned earlier and when a few of them fail, just pop in a new cell as needed. Yeah!!! I LIKE this idea. I'm a cheap bastard at heart. ;P

Thanks for mentioning the hybrid option...at first glance it sounded less than ideal, but now that I think about it again...well you are right...not a bad option.

Probably the laptop 18650's I have are all 1 or 2C cells. I sure don't know! it's all a mix of cells from "harvesting" over the past few years. If anyone cares, harvesting cells from laptop batteries gets you a lot of worthless 2600mah cells. Usually when a laptop battery dies, there's a 40 to 60% cell failure rate or worse. I consider anything that won't do 80% of its rated capacity of limited value and 50% is going in the trash. 1300mah is just such a low capacity to me that it is of virtually no value for other than a light weight LED flashlight so I just don't keep them around.

 

[Hillhater]

An "off the. Shelf", ready to run, pack of these cells is around $600 from em3ev !
EDIT..
Lunacycle 48v , 20Ahr, pack prices..
Lipo..$480....14.5 lbs
30q...$739..... 10.5 lbs

Absolutely correct, I did not account for any time or tools required. I would not recommend a complete EV noob to take on an 18650 build as their first project, but likewise I would prefer to steer that same person away from a rats nest of fragile lipo.

None of those assembled pack prices surprise me in the slightest, as you mentioned there's a substantial amount of time and some tools required to do it justice. Those packs will also include a BMS and in the case of Lunacycles from memory a charger is included too. Very good value IMO.
.

I believe it's the em3ev packs that have a charger included, . It's optional with Lunacycle.
And yes, both packs include a BMS, so that's a wash.
Looks like Egod couldn't crack the <$5 per cell for those 30Q cells, so I suspect most of those pack price differences is still in raw cell price.
Ohbse, you did some tests last year on using different cells in a parallel configuration.
..did you ever take that further with any more "Hybrid" pack builds or tests.?
I am still thinking there could be a good combination of Lipo and cheap 18650 cells to save both weight and cost.

 

[ElectricGod]

Looks like Egod couldn't crack the <$5 per cell for those 30Q cells, so I suspect most of those pack price differences is still in raw cell price.
I am still thinking there could be a good combination of Lipo and cheap 18650 cells to save both weight and cost.

I think people that are pack builders are buying many thousands of cells so their volumes of cells get them a much better price break than I'm going to see. Any seller knows they have hassle factor involved. Probably the hassle for selling 200 cells to me is very similar to selling a pallet of cells to a pack builder. There's just not very much profit in selling small quantities so you or I pay more.

I'm going with a mixed approach. Reuse my LIPOs into 20S packs and then possibly buy LIONs to make up the difference after I use up my harvested laptop 18650 cells.

 

[ElectricGod]

I found 4 cell 18650 battery holders on alibaba for $1.20 each so I bought 70 of them. It's not the most compact way to build 18650 packs, but it has the advantage of 100% easily replaceable cells in a modular format. I intend to make up 20S2P battery holders with them. I'll add a 4" thick box to the back deck of my scooter to put them in. I'll post up how well they work out in active use once I get there. The space in the battery bay will hold the 20S LIPO packs made from my reconfigured packs. These holders look exactly like the ones I bought on ebay for $2 each so if they hold cells as well as those ones do, then the cells will be pretty secure. You have to bash a holder really hard on your palm to dislodge the batteries. I pry the cells out with a small screw driver by their negative end because they are held in so tight. I put two of them top to top with cells in them and there is a 1/16" gap between the cells. If I butt the packs close together, they will hold each others cells in place. I just measured my back deck and its 17 by 12. I can easily fit a 20S length (16 inches) of these battery holders in that space. Build them back to back in a 20S2P configuration will get me lots of capacity out of free/used laptop 18650's. Later I can reuse them for new 18650's or whatever.

The battery holders are 3.18" by 3.18" by 3/4". Two of them back to back are slightly over 1.5" thick. If I assume a 1/4" thick wall to my box, then that leaves me 11.5" internally for packs across the width of the platform or about 7.5 rows of these things. If I leave a 1/8" gap between the back to back holders, all wires can be routed between the holders or across the top of each 20S2P battery holder. I'll add 6S balance connectors for balance charging and cell monitoring. I'll go with 7 rows of 20S2P packs so that leaves a lot of room for padding or thicker walls on the box or whatever. That's 280 cells in a single layer which will make a decent amount of capacity even with used laptop cells. Now to get lots more 18650's for free, but I have some ideas for that!!! I work for a large software company (think of your favorite 2 letter company and you might be right.) and we have tons of laptops. I'm guessing close to 10,000 of them. There's a good possibility of there being a few bad packs out of that many laptops. OK..quite a few possibilities!

Here's my 18650 bulk battery harness.

 

[Hillhater]

Are there any connections between each cell in those holders ?...series or parallel ?
Or do you have to make all connections yourself ?

 

[ElectricGod]

Are there any connections between each cell in those holders ?...series or parallel ?
Or do you have to make all connections yourself ?

You make all the connections between cells. You could bend the solder tabs over and bridge them together with a short length of wire. I want all my cells oriented so that plus is up and minus is down so that means wiring across the back of each 20S pack quite a few times. If you look at the pictures I posted, the individual battery slots are labeled + and - in the indented circles. I intend to use that for battery orientation. I should be able to do the wiring on just one side and then solder the other side onto the pins. Both sides of the 20S2P pack will then have all the cross connects and balance connections. Once I get the battery holders, I'll take pictures as I build up the wired side of the 20S2P pack. This isn't rocket science. You just need to understand what you are doing. Here's a schematic for my 12S LIPO packs. Leave out the BMS and you have what I will do with these packs.

You will need a bunch of these balance cables if you want to balance charge the cells while they are in the packs. Each one covers 6 cells so in my case with 20 cells that's 4 per pack. The other option is pull the cells out and charge them in a lion battery charger or build a separate 6S balance charge block like I did. In my opinion, include the balance cables in the pack build and then you can balance charge and check cell status without having to remove cells from the pack. As a general statement, you will not be pulling cells out of the pack all the time so just do the pack wiring with everything the first time and never have to think about it again.
http://www.ebay.com/itm/10-PCS-6S1P-Balance-Charger-Silicon-Cable-Wire-JST-XH-Connector-Adapter-Plug-/231608533530?hash=item35ecf1c21a:g:CnsAAOSwLVZVkf0X

Here's the 12S schematic. The balance connector is a JST-XH connector for 6S balancing. You should notice that cells that are the end of one balance cable is also the start of the next balance cable.

 

[HamsterPower]

You make all the connections between cells. You could bend the solder tabs over and bridge them together with a short length of wire.

I would love to see more detailed pictures and assembly progress!

 

[ElectricGod]

You make all the connections between cells. You could bend the solder tabs over and bridge them together with a short length of wire.

I would love to see more detailed pictures and assembly progress!

I'll post pics and progress in here and also on my scooter thread.
https://endless-sphere.com/forums/viewtopic.php?f=35&t=75177


I just purchased some 18awg silicon wire and 30 6S balance cables. Considering that Samsung 30Q cells are only good for 5C, 18 awg silicon wire will have me covered up to 22 amps. I'll do 18 awg cross connects on both "sides" of the pack. The main leads will be 14 awg silicon wire. 14 awg is good for like 50 amps so that should work well for the main leads and still be overkill. I doubt I will ever pull anything close to 50 amps per pack (probably closer to 6 - 15 amps) so the cross connects and main wires will never get close to being taxed. Don't forget that there will be 7 of these packs and also 2 or 3 LIPO 8000mah packs. At 1C for everything, ideally that's something like 42 to 50,000mah at 72 volts. I can live with that! I am assuming 2600mah LION cells since that is what my scavenged cells are. And don't forget that I said "ideally" so my expectations are more like 70-80% of the ideal. Right now I know I have 50 18650 cells of various brands and ages that all do 2400mah or better. That's enough for a single 20S2P pack and 10 spares so I'll need to "find" lots more cells.

 

[Allex]

I don't know, I see lots of problems using the holders, to many connections, voltage drops and heat. And then we have your 18AWG:
http://electronics.stackexchange.com/questions/38515/what-is-the-max-current-for-awg16

 

[ElectricGod]

I don't know, I see lots of problems using the holders, to many connections, voltage drops and heat. And then we have your 18AWG:
http://electronics.stackexchange.com/questions/38515/what-is-the-max-current-for-awg16

18 awg times 2 and that's for 6000mah at 1C. A single 18 awg silicon wire will do 22 amps and I'm doubling that. I don't think it will be an issue. Even at 5C for 30Q cells that''s still just 15 amps so I'm well within a single wires abilities and I'll have 2 of them in parallel. I want to keep the gap between the two halves as small as possible...hence the dual 18 awg wires for all the cross connects. Remember that I will also have 6 more 20S2P packs and 2 or 3 8000mah 20S LIPO packs too. One or two of these packs will be typically seeing 1/10 of the total load at most and that is assuming 60 amps total which is what my Kelly controller is currently set to. Even then, things deliver what they can. The overall current draw will be balanced out across the various current sources. No single 20S2P pack will see anything close to 20 amps on it. I said earlier more like 6 to 15 amps and 15 amps will be uncommon.

Heat will be minor...lots of air space around each cell and if I feel like it gets warm in there, adding a temperature controlled fan is an easy addition and there is room for several anyway.

The possibly questionable connections will be the clips touching each cells poles and they are really strong spring clips. You have to pry the cells out with a small screw driver AND the cells will be trapped so they can't come out. My LIPO packs have the same number of connections as these will have and that is the minimum possible. Every cell needs to be connected to the next cell via a wire or just soldering the tabs together or spot welding a strip of nickle or whatever. All those are connections of similar reliability. The only other connection is again the minimum needed and that is the main power leads coming out of the pack. The balance connectors wont be used except for balance charging and monitoring so they are irrelevant to the operational reliability of the pack. The ideal pack solution is indeed soldering or welding all the cells together and NOT using battery holders at all. However, and this is the over riding reason for going the to battery holders, is that it allows for easy replacement of dead or weak cells. It also allows for closer to 1S monitoring which is ideal. In a large pack made of many 18650's in parallel, you can't tell if a single cell has failed or not and once the pack has been built it is a huge undertaking to replace a bad cell even if you did know you had one. Is there better cell to cell connection reliability with spot welding or soldering as compaired to a battery holder? I would have to say definitely so, but how much better? Is the difference sufficiently small to not matter? I honestly don't know! I'm willing to invest my money and time to find out and then report back for the benefit of everyone else.


In between time check this out. Ask and you shall receive...and that was todays harvest. One of our IT guys took me to a locked closet where there were easily 250 laptops sitting there waiting to go back to our corporate office. He had a stack of them that had been wiped and were ready to be recycled. These were the batteries from those laptops, minus 2 or 3 that he kept back for a week or 2. He said once the next batch were ready to be recycled in another couple of weeks that he would have 10-20 more battery packs that I can have...and so on...and so on...and so on. Let the battery harvesting begin!!!!

 

[Allex]

My 30cm 12AWG wires at 10A continious, gets warm, around body temperature. When things get warm this means energy loss, something that is better to waste on the motor power or leave in the cells, not waste on cable/connector heating. Just something to think about...

"18650's in parallel, you can't tell if a single cell has failed" It is pretty easy to see actually and a little easier to replace compared to a lipo pouch where they are glued together and when separation them you pretty much ending upp destroying the good ones.
https://endless-sphere.com/forums/viewtopic.php?f=6&t=62650#p936926

Replacement:
https://endless-sphere.com/forums/viewtopic.php?f=6&t=62650&start=25#p954292

Now one should note that this is a very rare case and was caused due to a bad welder(punctured the can) and not the cell itself, when you take a new and geniuine 18650 and build up a pack from it you can almost forget about swappings. However if you build a pack from those laptops, you are just asking for a lot of trouble or a lot of your precious spare time if you will.

 

[ElectricGod]

My 30cm 12AWG wires at 10A continious, gets warm, around body temperature. When things get warm this means energy loss, something that is better to waste on the motor power or leave in the cells, not waste on cable/connector heating. Just something to think about...

Now one should note that this is a very rare case and was caused due to a bad welder(punctured the can) and not the cell itself, when you take a new and geniuine 18650 and build up a pack from it you can almost forget about swappings. However if you build a pack from those laptops, you are just asking for a lot of trouble or a lot of your precious spare time if you will.

Regarding warm wires...is it possible they are near warm objects such as batteries and speed controllers? 12 AWG silicon wire is good for something like 70 amps (I'm guessing, but 14 awg is good for 50 amps so that seems about right.). I'm thinking that there has to be another reason for your wires to get warm since you are using around 15% of the wires current carrying capacity. I do agree that one should use the largest wires possible for several reasons. Larger conductors have lower resistance and larger conductors carry more current. My 4 LIPO packs have 14 awg silicon wires. The packs are made of 8000mah cells. At 60 amps draw that means each pack is seeing 2.5C or 20 amps. The packs and speed controller generate heat so I can't tell if my wires are getting warm of and by themselves since they are all contained in the same space as everything else when I am riding. If I run off of one pack with the back wheel off the ground and the lid up I draw 6 to 8 amps at full speed and capture no heat. Nothing gets warm anywhere...including the battery wires. I've tried this with back brake applied to simulate more load and my wires don't get warm even though current draw easily doubles to 12 to 16 amps. I can carefully apply throttle so that I always pull less than 30 amps and ride with one LIPO pack. After I had the packs all constructed and tested on my bench, I put each pack in my scooter by itself and gave it a good 5 minute ride. I took each pack to 30 amps and held it there for 20 or 30 seconds. I watched for voltage drop-off (sag) and other signs of failure. After each pack was tested this way, the all went back to my bench to look for weak cells. I never noticed my wires getting warm despite being in a battery bay that was otherwise mostly free of other nearby heat generators. The packs were warm. the speed controller was warm, but the wires were the ambient temperature of the battery bay.

Yes...new cells have the advantage of being less likely to age off as quickly as a bunch of old cells that may have already lost half or more of their viable usefulness. You are absolutely right in that regard. My scavenged cells cost me nothing, but time and I would have spent just as much time testing brand new ones as I would the used ones so I don't see that as a a loss in either case. OK...the failure rate for new cells is going to be significantly less. And that brings up a super good point. With cells in battery holders, I don't care if a cell dies quickly or not. Pop it out and pop in a new one in a couple of minutes time and the problem is fixed. My worst case is I have spent around $100 on battery holders, some wire and some balance connectors that allow me to swap out dyeing cells quickly and easily and I have no other $$ cost involved. Testing cells is a menial task. I run them through my charger on several charge/discharge cycles and watch for the cells that charge too a low capacity. I use a skyRC d100 charger. Plug it into a USB port on your PC and you get lots of pretty real time graphs and charts for each cell. If I bought new cells, then I would have spent at least $800 for the cells (More like $1000!!!). My cost savings are 100% free cells. I'm willing to live with a higher failure rate when the cells are free. There's no way I would build a pack for someone out of them and claim anything other than "WARNING!!! You are buying a pack made out of old laptop batteries. Your run time and cell failure rate will vary and will be worse than new cells!!!". If I had a load of cells tested that I felt were going to hold up pretty well despite their used state and someone came along and asked if they could buy them from me, I would ask for a dollar per cell or some thing low like that. And then that's just to cover my testing time. I think considering my free battery source that battery holders are a probably the best option.

On the subject of replacing bad cells in a LIPO pack. I agree that taking apart an R/C brick possibly means damaging the cells around the failing cell since they are glued or double sided taped together and you have to pull the cells apart somehow. That is an excellent point for people to consider when they use pre-made R/C LIPO bricks. Up front before ever buying cells or bricks or whatever, I was considering things like this and as a result it weighed into my decision on how I would build my LIPO packs. I looked at R/C LIPO bricks, but eventually went with individual cells. In my packs, each cell is separated from every other cell by a thin piece of high density hobby foam. I wanted to arrest movement between cells and give them a buffer if there was movement. The foam has worked flawlessly. There has been zero cell movement that I can detect. I did take one of my packs apart because a tab broke off of a cell. All the cells looked perfect. The foam acted like the tape or glue between cells, but still allowed me to remove a single cell once all the Kapton tape encasing everything had been removed. There's a few places that Kapton is applied directly to the cells, but even that is minimal and only enough to keep the cells all together and not moving. Then I wrapped the entire pack in more hobby foam and that's where 99% of the Kapton tape is applied....onto the foam. It has worked super well and the cells are pretty safe inside their foam enclosures. There is literally zero evidence of a problem from my construction and binding methods. I will build future EV LIPO packs the same way! Even heat isn't a huge concern. I typically run at 1 or 2C so on a hot day (90 degrees F) my packs are running at 95 degrees or there about on average. That's barely warm IMHO.

Regarding connectors...despite my packs being limited by the BMS to 30 amps, I use XT60 connectors on them. I started with Deans connectors, but lets face it they pretty much suck for just about everything. All my battery connections go into my fuse block via an XT60 each. The power coming out goes to my speed controller via 2 8mm long bullet connectors. Another XT60 goes to all my auxiliary items such as lights, horn, monitoring, etc. It makes for a great way to isolate all that stuff with one reliable connection break. The only place I still use a deans connector is for the monitoring wires coming off my shunt. They are 22 awg which is uber overkill. That way I can isolate my watt meter from the rest of the system. Those 2 wires see a couple of milliamps at most. The LION packs will terminate inside the battery bay to an XT90 and feed in via 10 awg silicon wire. I'll add a multi fuse block inside the LION battery bay for safety reasons. Each 20S2P pack will get it's own fuse and maybe a status LED. It's super easy to make a small circuit that "detects" an open fuse and turns on an LED. All I need is some wire, LED and a current limiting resistor in parallel with the fuse. The fuse shorts out the LED and resistor so they don't "see" any current. When the fuse blows, all the current flow (which will be limited to 20mA) has to go through the LED which turns it on. An open fuse will be as simple as looking for the glowing LED. THe output side of the fuse block will all aggregate into the 10 awg wire. Each 20S2P pack will get an XT60 connector so it can be isolated from the system easily and reliably.

Well that about covers it in my opinion. Keep bringing your concerns and ideas. I am interested and obviously I am willing to state why I feel differently (if I do) or to agree with you. Thanks again to who ever it was that told me about hybrid battery solutions. This is all your fault. hahaha. See how you have saved my something like a thousand bucks!!! How dare you bring up a good idea that will cost me $100 instead! Haahahaha.

 

[Hillhater]

On the subject of replacing bad cells in a LIPO pack. I agree that taking apart an R/C brick possibly means damaging the cells around the failing cell since they are glued or double sided taped together and you have to pull the cells apart somehow.

With practice , care, and a few tricks, cells can be separated cleanly without damage.
A few drops of IP Alcohol and use of a fishing line " cheese cutter". , will be a good start.
Trickiest part is separating the tab connectors (crimped or soldered) and remaking them for the replacement cell.
Again, if you choose the common brick sizes, they are so cheap, it's hardly worth fooling about transplanting cells.

Thanks again to who ever it was that told me about hybrid battery solutions.

..your welcome !
Just goes to prove,...even a god can still learn something on ES. :lol: