So I think this is one of the rare battery builds on YT which actually checks all of Syonyk’s safety and sanity ideals for working with lithium cells.
So, @Syonyk, anything major that he missed? Seems pretty solid to me.
1 Like
Seems reasonable. He mentions “storage mode” settings for battery assembly (I agree, you don’t want to assemble at 4.2V/cell). I don’t think the grid spacers are strictly required (most packs don’t use them and I’ve never seen even a single worn wrapper in packs assembled tight), but I can’t argue, either - having a slight gap between the cells does prevent any sort of shorts.
He (properly) lights into those idiots who solder to cells, and demonstrates spot welding. I like his welder - it’s actually more capable than mine. There’s nothing wrong with his “negative strip across the parallel group and single strips over to positive” either - I tend to do a strip across the positive group as well, but that’s really more for structural reasons on the packs I build, and isn’t an issue with his cell holders.
No arguments with his “This is why you need a BMS” section, and I share his concern about cheap BMS units.
Clean BMS installation too - very nice and well protected.
Not a fan of the mid-video VPN service plug, but… shrug Beats pre-roll ads, maybe? Easy enough to skip.
But, I see nothing of concern there, and he builds a very well thought out, quite safe pack, using the proper techniques - which is, as you note, rather rare from a YouTube video involving lithium. It’s a lot less sketchy than my power toolbox build, for sure (which has been working fine, and I built that way because I had the old BionX 37V pack guts laying around).
Good to know, I’ll have to keep this one in mind for when people ask about it.
1 Like
You mean I shouldn’t be tossing a old SLA deep cycle marine battery into an ammo can?
SLA, do whatever you want.
You might not want to toss a “DIY Powerwall” lithium pack with exposed terminals in an ammo can, though.
2 Likes
Also his use of nylon bolts/nuts is great, it’s just that beyond extra little touch that’s part of that build.
Super fancy BMS though, but that’s not a bad thing for most folks who’d be building from a YT like that.
One thing I really liked was he specifically called out adding some additional fuses for stuff that was already fused on all the electronics/converters/etc. Just gives me warm and fuzzies, although I think he didn’t put a single large fuse on the direct output from the BMS pack. Or at least a breaker. Yeah, in theory the BMS could do the current limiting, but I’d still like a fuse/breaker to avoid that to keep the stress off the BMS. And port covers, especially on the direct battery output. But those are getting to be nitpicks.
And honestly, probably can be scaled up enough for a DIY powerwall type deal, only thing left for discussion is choice of invert and AC charger. Although personally I think he didn’t go quite far enough to say “don’t actually build from scavenged 18650 cells”, but at least he touched on the fact that you need some extra caution/checks.
That’s the job of the BMS, and they tend to be quite reliable at it. They’ll typically trip out past rated current (I’ve had this problem with the PowerWheels packs - those little buggers draw more amps than they have any business drawing on lithium), and the typically cut immediately if they detect a short.
Overcurrent protection is one of the key jobs there, and they just snap the output shut. I’m not sure what you gain with a breaker past them.
You might put a massive fuse on the input to the BMS, but if the BMS has shorted out, you’re having a bad day anyway.
That sort of build wouldn’t concern me if scaled up, though if you were going to “DIY Powerwall” it, I might add some ducting for cooling airflow and some fans. One thing I don’t think he added, which isn’t a big deal for his pack, but would be a problem on a larger pack, is a couple thermistors. The fancy BMS boards support several, typically.
I don’t mind scavenged cells, IF you are very, very conservative. Guidelines I’d have for using scrap cells in a build:
- If a cell is under 2.5V, it’s scrap. You do not charge it. You recycle it.
- Cycle the cells in an automated tester and firesafe location, several times, from 4.2V down to 3.0V. Measure capacity and IR across the range. Rate not by mAh, but by percent of stock capacity. Discard anything below 80% or with an internal resistance more than twice spec.
- Build some sort of automated spot weld eraser that grinds the ends consistently. You should be able to index on the corners of the case and mill the center.
- Mix cells such that you have an equal mAh and equal “percent original capacity” across each parallel group.
- Spot Weld Only! (see grinding the ends, half the reason these muppets solder is because you can’t spot weld properly on top of previous spot weld spots, it just doesn’t work)
That should get rid of most of the high risk cells, and let you build something reasonably safe out of scavenged batteries.
That’s a nice packable emergency power pack to carry around a few kwh. Would be capable of putting a couple miles worth of charge in an EV in case you came up short of a charging station. Pair a 240v inverter and a portable level 2 EVSE and it’d do it rather quickly too.
I was waiting for him to list the total build cost. A quick google comes up with anywhere from $2-9 per 18650 cell varying on brand and capacity, so maybe $700-800 since he was using some nicer components.
Um, I doubt you’d want to use a level 2 charger on this, I wouldn’t want to pull that much power all at once out of a smallish pack like that. With the right cells, sure, probably could manage it safely I suppose.
As an emergency charge up for another 5-10 miles? Perhaps, perhaps. Seems more likely you’ll be too far from anything (e.g. highway) and so you might as well just call AAA/etc, or you’re in urban type area where if you have an EV, I’d assume you’d have assured yourself of sufficient charging stations around.
He wasn’t using 18650s. Those are larger, IIRC. It’s not a cheap build, for sure.
An extension cord and EVSE in the car is going to be way more useful than that, though. And most EV guessometers are really good down at the bottom end, so there’s no real excuse to run things dry. If you do, that sort of bank won’t get you much.
A silly thing to do. You know your charge before you leave and should be responsible managing it. Drive more power friendly when you’re going to be near the edge.
Anxious and want to carry a boost? (a) Why’d you buy an EV? (b) Carry a small Honda generator good for a wall outlet 120V 12A charge. It’s about the same size.
Or you could just buy that wacky bmw i3 that has the ridingmower class generator built in to double its “we cheaped out on battery” original-leaf-like range. (Please tell me they stopped making those!)
I believe it was a “Scooter-class motor” from their motorcycle division.
However, I really don’t have a problem with the PHEV class of vehicles (the i3 was stupid because of California regulations, once you flashed it back to the European firmware it behaved a whole lot better, if still a peanut tank short range wonder).
The RAV4 PHEV is promising, and I’m really excited for the Mazda PHEVs that will likely have a rotary engine range extender. Loves me some Wankels!
We regularly run the Volt to and past the end of the battery range and end up using a quarter gallon or half a gallon or something - and still averaging 200mpg for the trip.
The level 2 part was sort of tongue in cheek, but a small, 3kw level 2 charger would only be 60 amps total. With his 12S7P battery that’d only be ~9A/cell and the first 20700 cell I looked up was rated 4ah even at 15A discharge. Assuming a full 4.2V/cell charge on the pack and that the spec sheet is to be believed, they should be able to run 9A for 26 minutes before they’re down to 2.5V/cell, if my math is correct. Of course it’d be pretty awful on the cells, but I’ve seen people do worse things, like fry an egg. I meant it’d kill the pack rather quickly, not that it’d charge the car rather quickly, although I’m surprised the math comes out 26 minutes, I was guessing it’d be less than 5 minutes.
I was only thinking 1 or 2 miles actually, but based on my math above (3kw for 26 minutes = 1.3kwh) and that you didn’t have to drive more than 25mph you could go 5-10 miles. You’re still probably right about just calling a tow truck to the nearest charging station though, you’d have to avoid at least 20 tows to pay for a battery pack like that.
Not a 24kwh Leaf, they’re notoriously optimistic when new, after 9 years and 70k miles the battery has degraded to about 20kwh (which is actually pretty good for Leaf batteries apparently) and the guessometer still says you can go 20 miles on that last 22%, but it definitely won’t. In fact it stops trying to guess at all below 6 or 7%. I’m pretty sure the guessometer is based simply on voltage and the immediate consumption, but doesn’t account for faster voltage drop on older cells, at all.
Agreed, both about trip planning and driving more efficiently, but plans change sometimes and some other people that drive my car sometimes aren’t willing or able to trip plan or drive the least bit efficiently, despite persistent prodding.
Usually no, but on the odd day outside my normal commute, sometimes. (a) for the lack of maintenance, lack of noise, instant torque, and the cheap fuel that’s available in my garage. Also, my commute was 11 miles round trip when I bought it, then we moved and my commute is now 42 miles round trip and I’m making it work instead of buying another car. (b) That’s a good idea, but I’d have to convert it to propane since it’d probably only be used a couple times a year and I don’t want my EV to smell like gasoline every time the temperature increases and the tank vents the vapor.
I actually rather like the i3s and considered getting one, but they’re just not enough better then the Leaf I already have to justify switching to me.
My problem with PHEVs is only philosophical. Why carry around a whole engine/fuel system/exhaust system/emissions system to use for 2% of the miles? You still have to deal with all the moving parts and maintenance of an ice, but you do get a lot of the benefits of an EV. At least the serial hybrids, like the Volt and Rav4 Prime can still be driven to shop on the battery if the ICE won’t start.
Do you have any problems with the gas getting old? One of my coworkers had a Volt and would go out and just drive around a couple times a year since he normally didn’t need any gas.
Why hasn’t anyone ever made a electric/diesel or electric/propane hybrid? The 15lb bottles of propane that you can exchange everywhere should be good for 70+ miles and are obviously easy enough for people to swap. The more shelf stable fuel along with the less restrictive emissions standards for propane make it seem like a reasonable combination.
“Most.” As in, “A competently designed EV pack.” Or, as a general guideline, “Anything but the Leaf.” 
Nissan checked every single box they could of "How to screw up a battery pack design by believing the lying sacks of crap datasheets and the manufacturer’s sales reps.
Most of the recent studies I’ve seen put lifetime emissions and TCO of a PHEV squarely within the error brackets around the BEV results. Operational emissions are somewhat higher, manufacturing emissions are rather substantially lower (the battery pack dominates embodied energy on a BEV), and the utility is far greater than a pure BEV. In terms of practical considerations, if we take a long trip, we can just stuff everyone in the Volt and go - and while we use some gas for the trip, if we want to take back roads that don’t go past fast charging stations, we can.
In terms of weight, you end up about the same as well - a PHEV and a BEV of the same class end up weighing about the same. Again, the battery pack is quite heavy, and you don’t need nearly as much of it.
We don’t - we use enough that it’s not been an issue. Last year, when everything shut down, we did have a few “engine maintenance cycles” fire off, but normally we use a couple tanks a year on the Volt and so there’s nothing special.
If the engine hasn’t turned in 6 weeks, it will light it up for a few minutes (5-10?) to splash oil around, clear out the injectors, etc. If the average age of the gas in the (sealed, pressurized, you have to push a button to depressurize it to refuel the car) gas tank exceeds a year, it will start burning gas until you put enough fresh gas in to bring the average down. Again, this isn’t an issue for us.
We ran 69 gallons in 2019, mostly because we didn’t have 240V charging in place yet and 120V on a short range pack isn’t great if you’re doing multiple trips a day. We put 29 gallons through in 2020, and are 10 gallons in for 2021 so far (based on date of fillup, I’m not tracking actual fuel burn by date). That number will go up because we’ve got a couple longer trips planned later this year, so I expect around 80-90 gallons for the year.
Most of the benefits of a hybrid you already get with a diesel (very low fuel burn at idle).
Go draw one of those bottles down in an hour and see how it works. I look forward to the report on Project Icemaker.
Fundamentally, you can’t pull that much propane, that fast, out of one of those in gas phase - and they’re gas outlet only. You’d need a liquid tap to run an engine of any significant power for any length of time, and now you’re up in a hundred pound tank range.
Agreed, but now that I know what it does and why it’s manageable, it’s just a little too sketchy for others, thus the comment about the jump pack.
Agreed, that’s why I labeled it philosophical. Practically, they’re very good, but I’d still prefer the serials.
I did forget about that whole ideal gas law. Solvable with a heated and insulated jacket around the bottle powered by a portion of the electricity generated by burning it, but that starts to get a little wacky.
Are you referring to “serial” as in “literally no physical connection possible from the engine to the wheels,” or… I’m a bit confused, you seem to previously think the Volt a serial hybrid, but the 1st gen isn’t really, and the second gen definitely isn’t (there’s a direct drive linkage between the motor and wheels available for efficient highway cruising on the gas engine). They behave in that way, mostly, but there are ways that mechanical torque from the engine is making it to the wheels in both gen Volts.
It ends up being quite silly, yes.
Having talked with my FIL who ran propane on most of his stuff for a few decades: The traditional way of solving it for propane carburetors on large engines is to take the liquid feed off the tank and run it into the regulator. The regulator converts it from liquid to gas, and is on the core engine coolant loop - it uses heat from the engine coolant to gasify the propane in the quantities needed. The output from the regulator is the gaseous propane, and flows through about a 1" diameter hose to the carburetor where it’s mixed in. On a lot of the systems at the time, the propane carb was bolted on top of the gas carb, and you had a switch in the cab to toggle which system was running at any given time.
However, this made sense when propane was cheap compared to gas (25% or 30% the cost for 90% the energy). Anymore, with them being mostly at parity, it’s just complexity for no real gains.
If you never need the gas engine, a PHEV isn’t a good option. As long as you use it at least occasionally, which most people will, it’s the best option for most people these days.
Also much of the ICE maintenance is quoted in miles for normal cars but it’s really “hours” - this is explicit in planes, tractors, and bulldozers but only implicit on cars.
So a bonus ICE that you rarely use will rarely need maintenance anyway.
Yes, I meant serial hybrid as in the electric motor is responsible for all the wheel torque and the electricity comes from either the battery or the generator, like the BMW i3 and diesel/electric locomotives. Basically, this seems like a much more simple approach.
I was referring to the Volts as the above, but hadn’t researched them much since I haven’t considering buying one. After looking at its drivetrain for a minute, I’d definitely agree that Volts are more of what I’d consider a parallel hybrid, like Toyota’s Synergy drive, or Honda’s IMA.
Everyone thinks that, but conversion losses suck. Going explosions > motion is fundamentally better than explosions > electricity > motion.
As for powertrain simplicity … the Prius (and other hybrids) tend to be the most reliable cars out there. And with the eAWD style PIHs, you still get the whole rear powertrain being purely electric (about 100hp worth in the R4P) along with the (pretty darned reliable) front hybrid system.
I just said more simple, I can see how parallel would be more efficient for non plug in hybrids or even plug in hybrids that are frequently driven further than their EV mode. It is hard to argue with Toyota’s reliability and the design of the primes, being able to go 20+ miles at highway speeds in EV mode, makes a lot more sense than lower power hybrids. They’re all more efficient than driving an ICE only.
Yes, though at cruise, the Gen 2 Volt actually has a direct drive mode for efficiency.
Every company who has had the option to do a full series hybrid or a parallel system has ended up with the parallel system for efficiency reasons.