How to make a DIY Tesla Powerwall for $300

[EEVblog]

  

Wow, this is seriously bad.
No battery managment and thermal system, and he says you can just hook it up to a standard 12V/24V/48V storage inverter!
No wonder he didn't actually plug it in and use it.
 

[nctnico]

Being (somewhat) involved in battery packs I can say that this is definitely not the way to construct a giant battery pack both from a safety and a prolonged cell life point of view.

[kripton2035]

there is a protection system : a fuse between each cell and the positive voltage
if a battery fails, most of the times, the fuses is blown.
he explains it perhaps in another video.

[wraper]

there is a protection system : a fuse between each cell and the positive voltage
if a battery fails, most of the times, the fuses is blown.
he explains it perhaps in another video.

Which fuses? he didn't even spot weld the batteries. Soldering them directly is a big no, especially in such quantities

[kripton2035]

he soldered a resistor leg between each battery and the positive pin of the pack
he measured that way the "fuse" gets blown around 5 amps, and shorted batteries draw around 6 amps when they die
in normal use each battery draws around 3 amps.
that's the method used in tesla car batteries (but not with resistors legs !, with real industrial fuses included in the battery holders)

if this works, I find it kinda clever.

[Kilrah]

What's lacking is a balancing circuit (and a charging circuit, but if you pointed that out he'd just say "yeah that's gotta be built in to your inverter"... or I'm too optimistic.
He obviously doesn't give a shit beyond his video view count.

[kripton2035]

if you wire a pack of batteries all in parallel, you don't need a bms.
you can have a dedicated charger per battery pack (which contains 150+ batteries in parallel) if you want
depends on the inverter you will use after that, but yes if you want more than 3.7V you can have a bms if you want.

also I think the output wires he uses are a bit thin for heavy amps like that...

[kripton2035]

fuses showed at https://youtu.be/Bk50IuWXg-c?t=213
the resistor legs explainations are in another video.

[electr_peter]

At least his neighbours can hope for free entertainment with with lights and flames (i. e., fire) in the near future...

Hole project is BS, it misses the point by a mile. Where are the proper electronic control and safety equipment?
The value in Tesla power wall is in the fact that is safe, reliable and usable - not that is 2 cents cheaper and causes fires every other day. Project that was shown on video captures non of that.

[PChi]

Wow a flammable battery holder. I been involved in a building evacuation and the arrival of the fire brigade after an explosion and fire of an abused lithium ion cell.

[station240]

Later video.
He does talk about needing to put a BMS in the box.
On of his chargers does have a built in BMS, as it's a commercial RC unit. I think he used that one to charge the battery.



He does explain the 'fail" in the APC UPS he bought cheap, which doesn't work.

Oh and I do agree with the timber cabinet being a hazard, others are doing the same thing. I dunno why, it's not hard to get a steel cabinet to put this sort of thing in.

[Red Squirrel]

This is an idea I've been toying with if I decide to go solar, as you would get higher energy density.  I would probably water cool it or something and have lot of thermal sensors etc... but seeing the "hover board" issues and what not, I'm not sure if I'd feel comfortable with a DIY lithion ion based battery pack like that in my house.  I guess with all the proper circuitry it should be fine.  Come to think of it, have they figured out why the hoverboards have so many issues?  Is it a design issue or something that people are doing wrong?  Perhaps they use it, the batteries get really hot, and they immediately charge it?  I have a small drone I've been playing with and the battery packs get insane hot, I always wait before I charge them, just to be safe.

[Kilrah]

Come to think of it, have they figured out why the hoverboards have so many issues?

Piss-poor, bottom price design and/or manufacturing on some brands/units, others are perfectly fine. Just a trend that developed so fast and with a short opportunity window that caused a faster than ever race to the bottom.

[mtdoc]

This is an idea I've been toying with if I decide to go solar, as you would get higher energy density.  I would probably water cool it or something and have lot of thermal sensors etc... but seeing the "hover board" issues and what not, I'm not sure if I'd feel comfortable with a DIY lithion ion based battery pack like that in my house.

For home power use, the DIY lithium battery solution of choice right now is building a bank using prismatic LiFePO4.  Lots of people are doing this now for off grid systems.  You don't need the energy density of LiPo for stationary home storage and with LiFePO4 you have very little fire risk - no need for water cooling or thermal sensors.

My next battery bank will very likely be LiFePO4

[Red Squirrel]

Interesting, is LiFePO4 a different form of lithium ion or completely different tech?   I'll have to read up on those. 

I've always pictured using golf cart or similar batteries but they arn't cheap for the power density you get compared to lithium ion and similar tech.  I think lithium ion has a better discharge/recharge life rate too.  I like the idea of powerwall except for the wall part... I'd rather have something rack mount or otherwise stackable for better density and expandability.  I know the powerwall will have this but it will probably be only for businesses, so really hard to get.

[station240]

Today's video
He looks at EV West's DIY 33kWh Tesla powerwall using off the shelf parts.


Total price for everything including solar panels
US$15,670

[KE5FX]

Heard some very compelling arguments for using forklift batteries for this sort of thing.  Apparently there are companies that will come out and install/replace them for you at reasonable rates.

[Red Squirrel]

One thing I've really been pondering on is doing something like this and start it off as basically a dual conversion UPS.  Rectifiers would charge the batteries and inverter would run off batteries.  Idealy you want a very big powerful inverter or some kind of inverter system that can handle multiple modules in sync.  That way you can have it feed the main panel, and not need to have a bunch of separate small panels if you went with a bunch of smaller inverters.    Everything would run off it, even the big loads like stove etc.  Once you have that setup working, then it's just the thing of injecting renewable energy, whether it's solar, wind, etc.  Eventually you could introduce enough power that you don't need to rely on the rectifiers as much.   But even before you add solar and such you could save a lot by simply running off the batteries durring peak billing time then charge them up off peak.  It's like 10c per kwh difference so it adds up.