[afidegnum]

I'm planning to build myself an electric car. Due to the cost of Li-ion batteries, i would like to build myself a hi-density rechargeable, battery or fuel cell.
So far, i have read about carbon based battery, salt, direct methanol fuel cell etc... which one would you recommend ?

 

[OR-Carl]

I hope you meant a scale model of an electric car! Modern batteries are sophisticated; they can not be made at reasonable scale in a home shop. You would be lucky to make something robust enough to power a remote control car. The good news is that major car makers are already making batteries for you, you just need to find a crashed EV car and salvage the used pack.

 

[afidegnum]

how do i find salvaged EV car batteries?

 

[Mustang101]

You can make your own pack if you don't want to gamble with used batteries that have been in a wreck. That's what I did.
18650 Cells are the way to go.

I'm actually sick of all the people on here telling people not to make their own packs. It does take time and depending on the salvage yard, might not be any cheaper but you'll get new batteries that haven't suffered an impact.

Attached are some pictures of the packs I've made for my 94 Camaro. It'll have 8 packs total when I'm done, 4 in the front and 4 in the trunk.

 

[remy_martian]

Most cells/modules are unaffected by an impact unless they get dented/pierced, which is rare because the cell has a Roman Candle "indicator."

Nice looking packs. Are they sealed for dust and getting water into them?

 

[DrGee]

You can make your own pack if you don't want to gamble with used batteries that have been in a wreck. That's what I did.
18650 Cells are the way to go.

I'm actually sick of all the people on here telling people not to make their own packs. It does take time and depending on the salvage yard, might not be any cheaper but you'll get new batteries that haven't suffered an impact.

Attached are some pictures of the packs I've made for my 94 Camaro. It'll have 8 packs total when I'm done, 4 in the front and 4 in the trunk.

Woah! These look really nice! I've dreamed about building my own packs for quite a while. How much did you pay for the 18650s? Are your packs sealed against dust and moisture?

 

[Duncan]

The trouble with buying "new cells" is that they are either very expensive or low quality - or both!
The ones the OEM's use have gone through a very thorough quality control process
The ones we can buy have either not gone through that process or have been the ones rejected

 

[Mustang101]

Duncan,

That's not true. You'll have to make phone calls and talk to dealers but my NCR18650PF are new from Japan. They went through a quality control process and of the +2000 batteries I bought very one was within a tenth of a volt of 3.4v.

3.4v is what they ship them at because it's better for storage.

DrGee, yes they are sealed from water and dust and I have a thermal management system I'm trying out with water cooling. The black tubes you see are standard quarter-inch sprinkler tubing.

I paid around $1.75 per cell.

 

[OR-Carl]

Hey Mustang, those are nice looking modules you have put together. I may have misinterpreted what the OP was asking, but based on the chemistries he mentioned, I thought he was talking about making the actual cells - which I do believe is a non-starter.

Building your own modules out of commercially available cells is another matter entirely. I am curious what you estimate your cost per kwh will end up being when all is said and done. Did you invest in a spot-welder to make all the those interconnections? Also, 18650 modules generally have cell-level fusing in case one of those 2000 cells develops into a short circuit - are you at all worried about that? I dont know what the odds are of a cell failing shorted, but with that many, it seems like eventually you might find out. Have you done any testing with that black plastic tubing to see how much heat it will actually transfer away from your cells? It seems that the low thermal conductivity of polyethylene is going to limit its effectiveness somewhat.

I feel like building modules is a great learning experience, but there is something to be said for leaning on the engineering and production acumen of the big car companies. If price is not a factor, and learning is more important than longevity, then I think building modules is probably perfect. The OP did not really make his priorities very clear, but hopefully some of this info is relevant...

 

[Mustang101]

OR-Carl,

Agreed making cells would be ridiculously difficult.

The price per kilowatt is around $15. That's including nickel strips and copper wiring and Anderson connectors. I haven't estimated the final kilowatt/hour.

I built my own spot welder with a car battery, PIC microcontroller, and some power MOSFETs. It wasn't that difficult and cost around $80 including the battery ($40).

I only spot weld the bottom of the cells. The top I do solder a small fuse and then connect that fuse to a copper bus bar.

I did run some tests with the thermal management system. I had one cell sitting in my garage in Las Vegas this summer with 115-degree heat. The outside temperature in the garage was around 101 but I had the controller keeping the packs average temperature to 85 degrees, never got above 87. I'm using standard automotive coolant with a Peltier system and pumps to run the coolant through the packs. I think the thermal conductivity of the vinal tubing is perfect for not stressing the Peltier elements but still cooling the pack to a nominal temperature range.

 

[brian_]

The price per kilowatt is around $15. That's including nickel strips and copper wiring and Anderson connectors.

My guess is that this was supposed to be the price per kilowatt-hour (of stored energy), not per kilowatt (of power output).

I haven't estimated the final kilowatt/hour.

Perhaps this means the final kilowatt-hour capacity hasn't been estimated? "Kilowatt per hour" doesn't mean anything.

 

[OR-Carl]

It looks like you have a pretty good setup there, Mustang. It is interesting to me to see what actually goes into building up a module from the ground up. I bought some tesla modules that also use 18650s, and I paid about 130$/kWh, going off the nameplate rating (in reality one would want to figure 80% DOD, plus some derating for the age of the cells to get a real-world price per usable kWh). Someday I hope to actually measure what they are capable of putting out, but my project is on summer break. It is nice to see that with enough time and ingenuity it is possible to make your own modules, although if I have learned anything so far it is to not underestimate the complexity of an EV conversion!

 

[Duncan]

Hi Mustang
You are aware that a tenth of a volt is a HUGE HUGE tolerance for a new cell! - 10 mV - or a 100th of a Volt is more usual

I've been on this site for a number of years - people who buy NEW cells - even expensive ones - have had a failure rate over 5%
Including me when I bought new cells

Cells from a crashed production EV? - I'm not aware of ANY problems

2000 batteries and $1.75 each so - $3500 - you claim $15/kwh ??
So 233 kwh in that dinky little pack -

I call BULLSHIT!!

That looks more like a $10 kwh pack - and that makes a lot more sense - $350/kwh -

Compare that to a Volt pack - like the one I bought for $1800 - that was $112/kwh
Just a wee bit LESS than your $350/kwh

AND I pull 1200 amps from my pack - 360 Kw - only for a few seconds

It's great to see people making their own packs - but expect to be severely disappointed compared to using a pack from a production EV

 

[MattsAwesomeStuff]

The price per kilowatt is around $15. That's including nickel strips and copper wiring and Anderson connectors. I haven't estimated the final kilowatt/hour.

$15 / $1.75 = 8.57 cells per $15.

Amps = 1000 watthours / (3.7 volts x 8.57 )
= 31.5 amps per cell.

Panasonic NCR18650PF 2900mAh 10A Battery

Panasonic Battery - Panasonic NCR18650PF 2900mAh 10A Battery

www.imrbatteries.com

"Discharge Current: 10A Maximum Continuous"

You're going to get 31.5 amps out of 10 amp batteries?

And, do you really care about power out of a battery?

You've already got 2000 of these you say... that's already 2000 * 10a * 3.7v = 74kW, or 100hp. Or, supposedly, 31.5 amps or 300+ hp short term.

And you're planning on having a lot more than 2000 I presume?

Power isn't really your limitation, energy will be.

I think the thermal conductivity of the vinal tubing is perfect for not stressing the Peltier elements but still cooling the pack to a nominal temperature range.

So... what happens when PVC tubes, in a vibration-heavy environment, through sharp bends, in heat and cold... break?

You're going to hose your batteries with water and antifreeze?

I like your solution but, I'd worry about imitating it. I am also doing 18650s, but I have no cooling solution other than perhaps forced air and a prayer (my are square packed, which does leave a fair amount of airspace for cooling).

 

[OR-Carl]

2000 batteries and $1.75 each so - $3500 - you claim $15/kwh ??

Yeah, this must have been a typo - as it seems to be off by a factor of 10. If they were indeed 2.9Ah cells (there are other capacity ratings available, I believe) they would hold something like 11 Wh/ cell if charged to 3.8v. So $1.75/11Wh = .159$/Wh or 159 $/kWh. Part of me feels like 5% failure seems like a lot if you are indeed getting fresh cells from a reputable supplier, but I do not have any actual experience to back that up. Also, with so many small fused cells in parallel, a few failures here or there is going to be less of an issue than with larger format prismatic cells. It does make me wonder how much of the capacity loss over time in a Tesla module could be explained by a few cells dying off completely vs degredation of the individual cells.

The use of plastic piping in the cooling system seems like a bit of a gamble to me too. I do think thermal managment is important, though, so I dont think it is wise to forgo it entirely. If price was not an object, I would think of using malleable copper tubing that was maybe rolled or hammered out into an oval shape? 1/4" runs north of a dollar per foot, though, so it would add up. I suspect all the auto manufacturers use aluminum, but getting set up to make aluminum water tubes with tight radius bends would not be easy for the DIYer I suspect.

If one were to try air cooling, what about putting aluminum bars between the rows of cells, and then putting in a 18mm aluminum tube in place of every 5th cell, say? Your pack would get a bit bulkier, but it might actually give you a fighting chance of forcing some air through. Or make it into a short-circuit-prone fire-waiting-to-happen?

 

[Mustang101]

Hi Mustang
You are aware that a tenth of a volt is a HUGE HUGE tolerance for a new cell! - 10 mV - or a 100th of a Volt is more usual

I've been on this site for a number of years - people who buy NEW cells - even expensive ones - have had a failure rate over 5%
Including me when I bought new cells

Cells from a crashed production EV? - I'm not aware of ANY problems

2000 batteries and $1.75 each so - $3500 - you claim $15/kwh ??
So 233 kwh in that dinky little pack -

I call BULLSHIT!!

That looks more like a $10 kwh pack - and that makes a lot more sense - $350/kwh -

Compare that to a Volt pack - like the one I bought for $1800 - that was $112/kwh
Just a wee bit LESS than your $350/kwh

AND I pull 1200 amps from my pack - 360 Kw - only for a few seconds

It's great to see people making their own packs - but expect to be severely disappointed compared to using a pack from a production EV

This is why I don't put out numbers, you guys are asses (looking at you Duncan). Yes I made a typo

 

[Mustang101]

$15 / $1.75 = 8.57 cells per $15.

Amps = 1000 watthours / (3.7 volts x 8.57 )
= 31.5 amps per cell.

Panasonic NCR18650PF 2900mAh 10A Battery

Panasonic Battery - Panasonic NCR18650PF 2900mAh 10A Battery

www.imrbatteries.com

"Discharge Current: 10A Maximum Continuous"

You're going to get 31.5 amps out of 10 amp batteries?

And, do you really care about power out of a battery?

You've already got 2000 of these you say... that's already 2000 * 10a * 3.7v = 74kW, or 100hp. Or, supposedly, 31.5 amps or 300+ hp short term.

And you're planning on having a lot more than 2000 I presume?

Power isn't really your limitation, energy will be.



So... what happens when PVC tubes, in a vibration-heavy environment, through sharp bends, in heat and cold... break?

You're going to hose your batteries with water and antifreeze?

I like your solution but, I'd worry about imitating it. I am also doing 18650s, but I have no cooling solution other than perhaps forced air and a prayer (my are square packed, which does leave a fair amount of airspace for cooling).

I'll let you know how vibrations cause issues with the cooing if any. I do have a temperature, voltage, and moisture sensors in each pack. I'm not worried

 

[Mustang101]

$15 / $1.75 = 8.57 cells per $15.

Amps = 1000 watthours / (3.7 volts x 8.57 )
= 31.5 amps per cell.

Panasonic NCR18650PF 2900mAh 10A Battery

Panasonic Battery - Panasonic NCR18650PF 2900mAh 10A Battery

www.imrbatteries.com

"Discharge Current: 10A Maximum Continuous"

You're going to get 31.5 amps out of 10 amp batteries?

And, do you really care about power out of a battery?

You've already got 2000 of these you say... that's already 2000 * 10a * 3.7v = 74kW, or 100hp. Or, supposedly, 31.5 amps or 300+ hp short term.

And you're planning on having a lot more than 2000 I presume?

Power isn't really your limitation, energy will be.



So... what happens when PVC tubes, in a vibration-heavy environment, through sharp bends, in heat and cold... break?

You're going to hose your batteries with water and antifreeze?

I like your solution but, I'd worry about imitating it. I am also doing 18650s, but I have no cooling solution other than perhaps forced air and a prayer (my are square packed, which does leave a fair amount of airspace for cooling).

26 cells in parallel then 9 parallel sets in a pack for a total of 260 amps continuous at 36 volts. 4 packs in series to make 144vs and then 4 more packs in parallel to make a total of 144v at 540 amps continuous. I'm limiting my controller to 500 amps which should be completely safe for my batteries.

Maybe someday if people stop trolling I'll post my videos, schematics, code, and 3D printing STL files so others can see exactly how I made these packs.

BTW I have a master's in EE so I'm not a newbie at this. I've been driving EVs since 2005.

 

[Duncan]

So you spend a LOT more money to produce a pack with dubious cooling that can supply a fraction of the power

It's your money !

 

[Mustang101]

I made a pack that I controlled completely, that is less weight, more efficiently cooled and will last much longer than any salvaged pack.