Need help with a few various parts.

therobby3

1 W
Joined
Feb 16, 2021
Messages
64
Hello, so I'm underway and still waiting on a motor and controller for an electric ATV conversion. I'd like to get everything ordered and ready so that when the motor and controller comes in, I don't have to keep ordering and waiting for more parts. I've got the motor, controller, and (still working on) the battery. So now I need to make sure I have any other odds and ends. I'm still new to electric, so please let me know if there's anything I'm missing. Here are some things I believe I will need below, so if you could help with any recommendations or info that'd be much appreciated. :)

Some info first, the controller at max should be pulling 300 amps from the battery and will be running 96v nominal. I'm assuming 96v nominal = 26 lithium ion cells, correct?

1. I imagine it would be a good idea to have some high power main cut off switch directly from the battery incase of emergency. Is there any one specifically I should look at getting? Again, battery will be putting out 300a max.

2. A main fuse? So I'm not misunderstanding, the main fuse would also sit in-line from the main positive battery terminal to the controller? Any ones specific I should get?

3. A BMS for 96v nominal and 300a peak.

4. Connectors that are rated for this kind of power? Are there any high powered connectors that I should be looking at to connect things like battery to controller, and controller to motor, etc?


And lastly, if there's any other major things I'm missing please let me know. I figured it'd be a good idea to ask so I don't miss anything and have to drag the project on longer than I want. And so I don't waist to much time and money on parts that aren't correct, or will fail and cause a fire or shock or whatnot.
 
Hey, I would really like to know more about your battery pack configuration! As your are going for massive 96v setup, I think it's important to make sure the battery pack handles it!
You have mentioned 26 cells, can you describe more? Like their type, how much in series and Parallel, company, etc!
 
300A at 96V means you will be pulling/pushing 30kW into a electric motor. that is HUGE. it equals to roughly to the power of a 800cc multi cycliner engine going full tilt from a dead stop. i personally think that if you have a motor that can put out that amount of force will just destroy your drivetrain instantly if you are even able to stay on long enough.

i seriously doubt your motor can actually absorb that amount of power without being liquified from the inside.

please share more about your specific engine, controller and planned battery setup. and what you plan to do with it.

ps: the fuse goes on the negative.... :wink:
 
I am using tesla 21700 cells and I'm going to be building a pack. The motor is a 8kw/30kw peak QS mid drive motor, so I think it should be able to handle it fine. The quad is a Raptor 700, so it's drivetrain should easily be able to handle it as well. It's mostly those various parts and such that I listed that I need help with.
 
without limiting the power you will destroy that drivetrain with that motor.

the amount of torque that the QS motor can put out is litteraly more then double of what the original petrol engine can put out at ideal rpm. think holding the perfect rpm and dumpting the clutch and double that. the drivetrain is NOT designed to handle that level of abuse.the QS motor can do that from a dead stop every time at a flick of your wrist without effort.

at full power you will turn your drivetrain into a really expensive maraca if you have too much grip or are stuck.

its going to be a HELL of a ATV with that motor but be prepared to either trash your drivetrain a couple times and/or turn down the power.

personally i would NOT have the bms process the current between the battery and the controller. use a contactor and hook that up to the 12V supply wich IS connected to the bms. so if the bms detects a problem it will cut the power to the 12V and close the contactor rendering the ATV dead. and have a 500A megafuse between the battery and the contactor.
bms is basically only on guard duty during operation and still in control during charging.

how much cells are you planning on putting in parralel? if its 10 because you read that the cell can do 30A then you are on the wrong track. you will need a 26S20P battery in order to get anywhere near close to decent power levels without causing exessive heating. dont forget that these cells have litteraly the best thermal management in the world when they are put in a model 3/Y and you dont have that in a ATV so you need to derate the cells accordingly. 15A max at the highest if you want to see some lifespan from them. wich i would hope you will considering the price tag of this project. its not really a question of what the bare minimum is, its what you can physically fit in the ATV. more = better. anything under 15P is basically pushing it and you will have to put a couple temp sensors in the battery to prevent the battery from overhating. these cells WILL get extremely hot under high loads. they are mare for capacity, not current. at least not without serious active cooling and heating to keep the temperature in the golden zone.

personally i think you would be better off with samsung 25R or example. much higher current capacity and also cheaper.
 
I am going to use an ANT bms from ic gogo on aliexpress for a similar setup.
(same motor and maybe controller, more powerful battery and 30s)

Flippys contactor idea sounds good, but I think it would be a little tricky to get to work well on that bms.

What have you been told about when the motor will be ready to ship?
 
@flippy I'm not really sure what drivetrain I would destroy with the motor. It's just the motor with a chain to the sprocket. The only thing it could really destroy would be to somehow shred the rear sprocket off the axle, which I highly doubt that would happen. I realize the motor itself puts out more torque, but by time you compensate for the gas engine's gear reduction (especially in first gear), I don't think it would necessarily be more.

Yes I was probably planning on around 20 cells in parallel, which would put it around 15 amp max draw per cell. The specific cells say they are rated for 15 continuous, so I'd think a blast of 15 peak would be ok.

@jbork Sounds good. Any idea on shipping time from AliExpress? I've never ordered from there and am just a bit hesitant. I don't want to be waiting another month or two for more parts. >.<

And I message qsmotor about my motor exactly 11 days ago, and they said it would be ready in two weeks. So that would mean in a few days it would get shipped out. Although they initially said it would ship in 3 weeks, and we're going on 5 now, so unfortunately I'm not confident in it shipping out on perfect time. =/ This is my first thing ordered from them, do they always take this long to "prepare" a motor, or is it just because this is a newer motor?

Also, the salesman did say this to me about the controller as well.

"The APT 96800 Controller support 96V battery (maximum 120V), we recommend full charged voltage no more than 110V. The maximum battery current is 300A, and the settings are 180A now.

Just incase you want to be sure 120v will be safe for the controller. I'm still semi new, but figured I'd share that.


Anyway, what I really need help with is exact specific parts, like links to them so I don't buy something wrong. I've bought these fuses, in 400 amps. I'm hoping these are correct ones that will work. https://www.ebay.com/itm/224057969036

Also, what would I do as far as anti-spark connections? I converted a small dirt bike to electric at 72 volts, and sometimes connecting the battery without an anti-spark connector would result in a nasty spark. I used an XT90 anti-spark connector though, which is obviously way to small for my ATV build. What is everyone doing as far as anti-spark connectors for their builds?

Still wondering about some kind of main on/off switch that I could put right at the battery as well, that I could just push off and would prevent the power from the battery from going anywhere in case of emergency as well.
 
therobby3 said:
@jbork Sounds good. Any idea on shipping time from AliExpress? I've never ordered from there and am just a bit hesitant. I don't want to be waiting another month or two for more parts. >.<

And I message qsmotor about my motor exactly 11 days ago, and they said it would be ready in two weeks. So that would mean in a few days it would get shipped out. Although they initially said it would ship in 3 weeks, and we're going on 5 now, so unfortunately I'm not confident in it shipping out on perfect time. =/ This is my first thing ordered from them, do they always take this long to "prepare" a motor, or is it just because this is a newer motor?

Also, the salesman did say this to me about the controller as well.

"The APT 96800 Controller support 96V battery (maximum 120V), we recommend full charged voltage no more than 110V. The maximum battery current is 300A, and the settings are 180A now.

Just incase you want to be sure 120v will be safe for the controller. I'm still semi new, but figured I'd share that.


Also, what would I do as far as anti-spark connections? I converted a small dirt bike to electric at 72 volts, and sometimes connecting the battery without an anti-spark connector would result in a nasty spark. I used an XT90 anti-spark connector though, which is obviously way to small for my ATV build. What is everyone doing as far as anti-spark connectors for their builds?

Shipping time, it can vary. They usually specify what can be expected. 2-6weeks depending on shipping method usually I think.

This is not the usual waiting time from qs, but this motor has not been produced before. I think it is usually a week or two before they ship.
It has been "almost ready" for at least a year, but it has not been possible to order before. I think I ordered it in January.

That they only recommend 110V was more bad news.. I have heard that the 96600 dont deliver max amp at full voltage, so I guess that is probably the case here too.
The size and limited to 300 battery amps already made me hesitate, if it is "recommended" to not go over 110V..
I hope to be able to run at least 120V, maybe 125V.
I want a controller that can do 1000phase A and 500battery A, and 125V.
But it is not easy to find anything suitable.

Anti spark? Dont work work on live high voltage. Turn off the bms before you connect stuff, no need for anti spark connectors.
 
Ok, good to hear the wait time from QS isn't always this long, incase I buy some other things from them in the future.

Yea, I'm hoping the controller is still plenty powerful though, so I guess we'll both see how things go. :)

Maybe I'm confused on something. I did a quick search and am not easily finding any BMSs that are 26s and can support at least 300amp. I would think, since my build will be 300 amp peak, that I should look for a BMS a little above that, correct? Like 350 or 400amp BMS? Surely it can't be this hard to find a BMS? ...I'm also curious as to why literally everything is from China. >.<

So your saying with a BMS, that I wouldn't need an anti-spark and would just turn off the BMS? I have a small BMS for a battery that I am working on for a small dirtbike, but as far as I know that BMS has no on/off switch or anything like that. Your saying these bigger ones likely would? Also, I thought the BMS would be stored internally with the battery once completed.
 
therobby3 said:
@flippy I'm not really sure what drivetrain I would destroy with the motor. It's just the motor with a chain to the sprocket. The only thing it could really destroy would be to somehow shred the rear sprocket off the axle, which I highly doubt that would happen. I realize the motor itself puts out more torque, but by time you compensate for the gas engine's gear reduction (especially in first gear), I don't think it would necessarily be more.

you are miscalculating the forces involved, as i did many years ago with one of my first conversions. learn from my mistakes.

the highest tourqe that the motor can dish out happens at 0 rpm. your standard ATV chain will snap like dry spagetti and you will probably bend the rear sprocket and axle or a bit too loose chain will just strip the teeth off. at those power levels SOMETHING will move or give way. usually something expensive.
drivetrains are not made to handle such forces from a dead stop. this is why electric vechicles do NOT have gearboxes and insane reduction drives with sprockets you normally find in 18 wheeler gearboxes and diffs.
you need to double up the chain (offset by 50% of you can) or get some pHat aftermarket sprockets that can take a wider chain.

i managed to utterly destroy a drivetrain of a converted BMW E38 with a (utterly beautiful) M73B54 V12 engine that i replace with a DC motor from a forklift. the tourqe from the forklift motor was roughly the same as the original V12 but it still twisted that driveshaft like a piece of overcooked fusilli when i put my foot down. after replacing that and locking the gearbox in 3rd to keep the torque down it imploded the gearbox at half trottle. after that i removed it all and just bolted a steel driveshaft of a truck directly between the DC motor and the rear diff. wich ofcourse also exploded like a bomb within 20 miles.....
and that ENTIRE drivetrain was originally for a 500nm 350bhp V12 engine and had done 150k miles in 10 years without any issue and it would not even last for 10 miles with a 100bhp 500nm DC electric motor.
your new motor can do DOUBLE the torques and match the power output of what the original petrol motor could ever do at 3~5000rpm from a dead stop.

drivetrains CAN handle power, if you allow then to make LOTS of rpm like the petrol engines MUST make before the engine produces some torques. but a electric motor does the opposite, ALL the torques and none of the rpm. this is why a tesla reduction gearbox has a (roughly) 10:1 reduction ratio and STILL has 2 inch thicc gears inside. wich is roughly about the same width that gears inside 18 wheelers have inside their gearboxes. so truck gears in a 250bhp compact EV just to handle the forces.
this is why there is basically no EV on the market with a conventional gearbox. the gearbox would need to be lifted out a 18 wheeler in order to handle the power.


therobby3 said:
Yes I was probably planning on around 20 cells in parallel, which would put it around 15 amp max draw per cell. The specific cells say they are rated for 15 continuous, so I'd think a blast of 15 peak would be ok.

yes, that would be a good starting point. but you do understand that tesla cells have aluminium tops and are nearly impossible to weld properly?

therobby3 said:
Anyway, what I really need help with is exact specific parts, like links to them so I don't buy something wrong. I've bought these fuses, in 400 amps. I'm hoping these are correct ones that will work. https://www.ebay.com/itm/224057969036

megafuses are great. just get the original ones from bussman for example. trust me, those ebay ones are NOT trustworthy.
just search for ANL-400 bussman and get a supplier like mouser,digikey,farnell or whatever.

and yes, real fuses are that expensive. i know.

therobby3 said:
Also, what would I do as far as anti-spark connections? I converted a small dirt bike to electric at 72 volts, and sometimes connecting the battery without an anti-spark connector would result in a nasty spark. I used an XT90 anti-spark connector though, which is obviously way to small for my ATV build. What is everyone doing as far as anti-spark connectors for their builds?
Still wondering about some kind of main on/off switch that I could put right at the battery as well, that I could just push off and would prevent the power from the battery from going anywhere in case of emergency as well.
you dont need a anti spark connector, you connect to the contactor and that has a resistor across it to precharge the controller.
you put a fail safe rope thingy you put on your wrist and put that on the wire feeding the contactor. pull the cord and the contactor opens and it instantly kills the controller/drivetrain but keeps your 12V intact for lights and stuff so you remain visible.


therobby3 said:
Ok, good to hear the wait time from QS isn't always this long, incase I buy some other things from them in the future.

Yea, I'm hoping the controller is still plenty powerful though, so I guess we'll both see how things go. :)

Maybe I'm confused on something. I did a quick search and am not easily finding any BMSs that are 26s and can support at least 300amp. I would think, since my build will be 300 amp peak, that I should look for a BMS a little above that, correct? Like 350 or 400amp BMS? Surely it can't be this hard to find a BMS? ...I'm also curious as to why literally everything is from China. >.<

So your saying with a BMS, that I wouldn't need an anti-spark and would just turn off the BMS? I have a small BMS for a battery that I am working on for a small dirtbike, but as far as I know that BMS has no on/off switch or anything like that. Your saying these bigger ones likely would? Also, I thought the BMS would be stored internally with the battery once completed.

with currents like this you dont run the power over a million mosfets. thats just dumb and unreliable. once 1 fet blows you get a cascade and you blow up your entire board. forget that style.

you get 2 outputs on your battery: 1 with 96V "low power". that ones connects to your 12V and the charger. that actually goes over the bms but will probably only see a dozen or so amps during charging.
the other output is the "unregulated" power uitput that has a ANL-500 fuse on the inside of the battery as your "last resort". your ANL-400 (i would prefer a 350) is mounted outside the battery BEFORE the contactor. the contactor is powerd by your ignition key wich gets its 12V from a 12V converter that is always on from the low power output of the battery. the reason is simple: your ignition key is NOT rated to swich 110V DC. it will arc and kill itself and probably weld itself closed so you cant turn the ATV off. so you make 12V and switch that. you can ofcourse have a separate swich that can take 110DC to power off the 12V.

ps: get the JK-B2A24S-20P bms. its a LOT better then the ANT bms. https://www.aliexpress.com/item/4000529723243.html
it has active balacing and all that goodness.

ps: for connectors you just use anderson SB175 for the main unregulated side and a smaller SB50 for the regulated side that goes over the bms for charging. donr forget to get some of those rubbber boots for them to prevent muck from getting in.
 
Ok thanks for the info! I suppose I'll give the drivetrain a try and see how it goes first. I don't want to change too many things to soon if I don't absolutely have to.

As for the tesla cell, I've already spot welded a handful of them for a smaller battery I'm building for an electric dirt bike. They actually spot weld very good and the welds are on there super tight. Maybe it's because they are "repackaged" cells, I'm not sure.

Ok thanks for the info about a contactor, will have to look into one and wrap my head around how to set it up.

So your saying the BMS will not monitor discharging then, and the discharge wires will just go straight out from the battery and not be fed into the BMS? If this is the case, I don't need to look for any super high amp rated BMS, correct? Since it would only balance when charging?
 
therobby3 said:
As for the tesla cell, I've already spot welded a handful of them for a smaller battery I'm building for an electric dirt bike. They actually spot weld very good and the welds are on there super tight. Maybe it's because they are "repackaged" cells, I'm not sure.
if you have button top tesla cells then they are modded because they are from a crashed (and used) model 3/Y. PLEASE take off the wrap of 1 cell and make a couple pictures. many of them are fake, especially if they were cheap.

do you happen to have a IR meter and a device to measure the capacity of the cell?

therobby3 said:
So your saying the BMS will not monitor discharging then, and the discharge wires will just go straight out from the battery and not be fed into the BMS? If this is the case, I don't need to look for any super high amp rated BMS, correct? Since it would only balance when charging?

correct. but you do need a bms with a couple features like high balacing current. the model i linked to will suffice for your battery and has some nice features like resistence meters for each group.

for current measurements on the go i would recommend using a shunt and a ANALOG meter to keep an eye on the current you are pulling. dont use digital. a needle is way more easy and faster to read at a glance. same goes for a temperature meter for the battery and motor.
 
Yea, I should probably do a drainage test to make sure the batteries actually have the capacity they say before I go and order a whole mountain of them. I do have an IR meter and a device to measure capacity fortulately. I'll have to do that soon. In the meantime, here is a cell I remove the shrink wrap and cap from.
 

Attachments

  • 20210409_131638_HDR.jpg
    20210409_131638_HDR.jpg
    141.7 KB · Views: 869
  • 20210409_131623_HDR.jpg
    20210409_131623_HDR.jpg
    165.7 KB · Views: 869
Hello therobby3 and group,
I have a lot of these tesla 3 cells - directly from a car. My cells look the same as in your picture. I tried to spot-weld them. The +pole ist out of aluminium and it the spot-weld-connection doesn`t work. One battery lost some fluid (looked like water). It came out of the + pole. :shock:
Which cap did you remove? I found on aliexpress some caps - on which you can spotweld: https://de.aliexpress.com/i/33047464647.html?spm=a2g0x.12057483.0.0.18881119ksO6U7 Does it look like this?

Or - How did you spotweld them?
I really would like to use my batteries. :|
 
@Mammutbaum My cells came wrapped and with a cap pre-spot welded to them (however they did that). The pictures I took were with the shrink wrap cut off and the cap removed. Here is what my cell looked like before I did that: https://bigbattery.com/products/model-3-2170-cells-w-cap-box/
 
@ therobby3
thanks for the answere - so they somehow spot-welded the cap on the battery - and I am not able to spotweld the nickel strip to the battery - so I`m most likely not able to attach this cap (?). :roll:
I have some shrink wrap - but I am relatively shure it is not stable enough to connect the cap with the shrik wrap to the +pole (?).
In your detail foto of the cell I see about 20 very small points in a ring. The ring looks a little bit treated. I would like to know how they connected the cap on the +pole of the battery. Otherwise my batteries are worthless somehow :/ - I put so much money and time in them to clean them. :roll: :(
 
therobby3 said:
@Mammutbaum My cells came wrapped and with a cap pre-spot welded to them (however they did that). The pictures I took were with the shrink wrap cut off and the cap removed. Here is what my cell looked like before I did that: https://bigbattery.com/products/model-3-2170-cells-w-cap-box/

EW, Uncycled Model 3 Cells are harvested from assembly defect modules. All cells have been professionally buffed, tested, re-wrapped & laser welded with a cap for safety when spot welding. These cells are perfect for building DIY projects such as powerwalls, any lead-acid to li-ion conversion, mobile solar generators etc.

BigBattery’s Model 3 2170 Cells will trump any and all 18650 designs with 4800mAh Per Cell – There is nothing better!

gotta love that marketing bullshit they just slap in your face.

just a few notes:

no defective modules leave the tesla factory.
modules only leave the factory when bolted to a car.
the only way to get a module is to take it from a car.
its not economical to buy a car to harvest its cells.
the only economical wat to get a tesla battery module is to take them from a totaled car.
a module only contains cells, so how can you get good cells from a module if its defective?
what is professionally buffing? how does that differ from unprofessional buffing?
how can you test a cell if its not cycled?
re-wrapping cannot be done as the cells were never wrapped. its wrapping, not re-wrapping.


english is not my first language, what is their excuse? the re-wrapping stuff reminds me of pre-heating an oven. that is impossible. an oven is either heated or unheated.
 
@flippy Haha, fair enough. Yea, there is definitely a lot of fancy bullcrap marketing that companies like to use on their products.

@ Mammutbaum Yes, it looks like there are a dozen or so spots enough the cap that are linked to the cell. I'm not sure how exactly, if they were done by spot welding or what. Unfortunately, I'm not sure what your best bet would be. They had to attack them somehow, finding out how would probably be your best bet, there's gotta be some way.
 
@ flippy thank you for answering.
You are wright, tesla is hyped and most likely overrated in my opinion but I have to admit that I didn`t read this marketing speech bevore I bought them (in spring 2020 - a time where theses cells were not on the market.

I chose these cells because of
1. The brand: Lots of people I heard described some 18650 Sanyo/Panasonic cells to be very reliable
2. They are technically new - Tesla wouldn`t use outdated and short-lived cells for a car which should last for 10 years (image damage would be the downfall of the whole company
3. good technical values: 5000 mAh and 15 A - I think this is very nice. Not godlike - clear but in early 2020 "state of the art".
4. Upcycling thoughts: The cells I bought were used from a crashed car - so I think nearly no environmental and human exploitation was accepted in the producing (?) - like buying expired food...
5. somehow cheap these days (2,5 € to 3 € for a cell of 5000mah).

To point 5: If I had known that i had to "clean" and "cut" every cell for about 10 minutes from stone-hard epoxy, wrap the cell (I know, it is the first time), isolate... I would never had bought them. And now as I got a spotwelder and noticed that I couldn`t spot weld them - never never ever :|.
I think I was ripped off a little bit by the seller.

But in my opinion, if I would find the Panasonic NCR21700A for 3 € a piece - wrapped, with good cap (nice to spotweld) and without damage - I think I would probably buy these cells - because I can`t see any better and cheaper cells on the market... :)


@ therobby3 thank you for your hopeful words :) - probably I can find a solution for my problem :confused: and am still interested in your test of your cells :D
 
Mammutbaum said:
I chose these cells because of
1. The brand: Lots of people I heard described some 18650 Sanyo/Panasonic cells to be very reliable
2. They are technically new - Tesla wouldn`t use outdated and short-lived cells for a car which should last for 10 years (image damage would be the downfall of the whole company
3. good technical values: 5000 mAh and 15 A - I think this is very nice. Not godlike - clear but in early 2020 "state of the art".
4. Upcycling thoughts: The cells I bought were used from a crashed car - so I think nearly no environmental and human exploitation was accepted in the producing (?) - like buying expired food...
5. somehow cheap these days (2,5 € to 3 € for a cell of 5000mah).

1: correct, its still not possible to match 2170 cells in reliabliity and performance with 18650.
2: no, they cannot be new unless the cells they harvested are ripped from a car that they wrapped around a tree 1 mile from the tesla dealership when the original owner bought his model 3/Y. you simply dont know how old the cells are. (or someone broke into a tesla factory and carted out thousands of cells with nobody seeing)

this needs to be more clear when buying 2170 tesla cells:

model 3 production started in mid 2017. nearly 4 years ago.

you simply dont know if the cells are 1 week old or 4 years with hundreds of cycles. and ofcourse with the marketing they claim all cells are freshly picked from the vine by nude virgins. just like a used engine always has just 20k miles on it....

4: robots make the cells and batteries. no humans are exploited in the making of a tesla pack.
5: 3 euro for a -used- cell with a unknown life. for all you know its been supercharged its entire life.
 
Hi flippi

1 - I don`t understand - you say that the 18650 cells are better? why? For the handling and spotwelding process the bigger cells would be better because you need less connections and the area to spotweld is bigger.
2 - the design was new - not outdated...
I knew that the cells were not new. The seller bought a whole tesla battery (weight: 500 kg) to build a energy storage for the house for himself. The cells I bought was left from his project.
In EU they started to sell the tesla 3 in February 2019 - he bought a battery with around 10.000 km he said. Maybe someone crashed a nearly new tesla. Could happen with a fast acceleration or another car crashed into it. I don`t think that he shipped the 500 kg battery over the Atlantic :lol: but you never know :roll: I really don`t know if he lied to me but I think not. I will try to test the cells some day.
4 - my concerns regard not the last production step but the harvest process of the rare elements in Kongo for example. Really never heard from the problems?

But probably you are wright - probably they sell very old batteries(?). I don`t know at the moment. But @therobby3 will check the remaining capacity and I also will try. Probably Panasonic produced some batches with only 4800 mAh instead of 5000 mAh and they can`t use them for the car battery because they are too weak for them. Who knows (?).
Clear - if they are used the company who sells them has to specify it correctly.

From an environmental (nature) perspective I think it is very good to sell and buy them again (in case they work somehow properly) -
but I know that many people are not interested at all in this :|
 
Mammutbaum said:
1 - I don`t understand - you say that the 18650 cells are better? why? For the handling and spotwelding process the bigger cells would be better because you need less connections and the area to spotweld is bigger.
as long as you make solid welds there is no difference. more welds are only a problem if you don make them right. and then it does not matter how many you have, you are doing something wrong.

the less-welds-is-better-matra is not something i personally agree with. generally less is indeed more. but considering you are making more welds in a smaller surface you are spreading the loads (physical and electrical) over more welds making the load over each weld less and making the whole pack stronger. but if your construction is relying on weld strength to keep together you are doing something wrong in the first place.

my main point is still that the 18650 form factor is tried and true. quality is the best you will ever see and you can get a cell that specifically and exactly fit your use case. and they are cheaper. you can buy new cells for less money then what a -used- "sorta good" cell for your application. in the end its always a compromise. and as long as you cant buy the tesla cells brand new and with a offical datasheet from panasonic i would simply not use them. its always used stuff that you get without a known history and you still have to sink in a considerable amount of money and you are still not getting exactly the cell you need. all kinds of problems you dont have with 18650's.
 
@ flippy: I understand your point of view. For ebike etc. the 18650 have a good size. Yes more cells distribute the loads better than less cells - I also think in this way.
The 21700 user are somehow early adopter, the benefits of most of the cells seem to be small and the risks high. The propability to receive something that does not meet the needs is higher.
In my case I would have to pay about twice as much to get new Sanyo 18650 cells. I trusted the seller - to get nearly new 21700 cells - probably I was wrong but I will see. The cells required a lot of work and frustration and time - new cells of a good brand are save and tested. For me it would have been the better choice but now I try to do the best I can with the cells I already have.
 
dot worry about that too much, hindsight is always 20/20.

for me its simply a no go to even consider aliexpress or whatever. my job basically consists of "fixing" aliexpress battery jobs.
not to mention i have to pay the bills with it. so i cant spend hours and days just grinding off old weld nubs on each cell when i can litteraly buy better new ones for less money. and i can actually stand by what i made with warranty.
i dont need someones house burning down because i just happen to hit a batch of crappy fake cells.
 
Ok I understand. Interesting job!
I wouldn`t buy the cells on aliexpress. Here in eu we have a quite cheap shop called "nkon".
 
Back
Top