New 14s battery pack, US18650VTC6 vs NCR18650GA

[flippy]

This car seat heating pad is giving 22W at 12 V. Put on the blanket, thermocouple inserted between heating pad and blanket is showing after few minutes 30 – 40 °C depending on the thermocouple position at ambient temp 25 °C. Dimensions of heated part 28 x 44 cm. No problem to bend in radius 2 - 3 cm.
https://www.aliexpress.com/item/32729791102.html?spm=a2g0s.9042311.0.0.44a74c4dJVSX6V
Car seat heating pad.jpg

unless you have a 12v battery that aint gonna work.

you can get those kapton 3M backed heating pads in every size you want and even custom sizes and specs if you are willing to pay a start fee for a akward size/power rating.
i have sereral custom pads in large surface areas (25x25cm) and low power (10W@100V) especially to evenly heat up large packs slowly.
they are also great on pouches as you can shove them between cells and perfecty heat them up from the inside.

 

[Jan-Erik-86]

I allready have two seat heater pads ready to install on my battery. I was planning on using a digital thermostat instead though, and was thinking of inserting them between the battery and the insulation instead of between the battery case and insolation. Haven't decided to use one or two yet though.

Flippy, altohugh not 100% efficient you could still use them with a small buck converter, which also lets you set the required power very easily (and it will be constant, regardless of battery voltage).
Most small and cheap buck converters are for less then 60v though, but for higher voltage you can often just use a normal 100-240V AC/DC adapter (but output current is often a bit limited).

 

[Altair]

If you're using heating pads (silicone or kapton) that are usually made to work on 120V AC, and your battery is around half that, or 60V, the power dissipated by the pad will not be 1/2 of its rating. It will be only 1/4 of its rating.
This is because of Ohm's law P=EI. As the pad has a fixed resistance, halving the voltage will also halve the current into the pad. So half the original voltage, times half the original current into the pad = one quarter of the original wattage.
To counter this, you just buy a pad that has FOUR times the wattage you need.

With good insulation, 5W should be well enough to keep a pack 20-25 degrees above ambient.
I would place the pads on the ends of the cells , rather than on their side, even if this is the least convenient placement.

 

[john61ct]

Buy pads designed for the voltage range you are supplying.

Put them under the cells, with an air gap and integral thermostat on each to cut off @overtemp.

The master thermostat regulating overall temp range, with hysteresis delay, sensor ideally buried middle of the bank.

 

[larsb]

There is one problem with insulating your pack.. Heating

And one connecting a heater directly to it - overheating and risk for heater or switch failure.

It’s not worth the risks in my opinion.

Have you two ever done it? (Flippy&john)
what are your experiences in this case?

 

[john61ct]

No one would do it unless their use case actually required it, not optional.

Better IMO to go modular and bring the packs into the heated work / living space when things get so extremely cold.

 

[eMark]

I see some of you recommend NCR, and some say forget the GA, so i'm a bit confused..

What about NCA or is that too mild-mannered? How close are you to narrowing your choice down to your final selection?

Vruzend no longer relies on "friction fit", as of v2.1
https://youtu.be/rylbFnTgFI8

V2.1 is still a "friction fit" (ask Micah if U don't believe me) although it's an improvement over V-sag with the V1.6. That's if your intention is sup'-dup' speed (heavy pedal to the metal) then Vruzend V2.1 is far from the best fit.

The V2.1 kit (enuf pieces for a 10S3P for beginner build). A viable choice for cell experimenting (small pack), and as a suppport companion pack for keeping the primary pack well above sub-zero temps. Two bodies better than one :thumb:

Being from Minne-sota (and half-Swede) am doing some experimenting with a V1.6 kit as a warming companion for my primary 10S4P pack. One advantage is an etrike with a Plano storage trunk between the rear wheels ... https://www.homedepot.com/p/Plano-56-Qt-Sportsman-Trunk-Black-161900/308744534 ... modified to mount on the rear of my Liberty etrike.

Just wondering if a Kenda etrike ... https://www.addmotor.com/shop/m-330 ... would be more suited/sufficient for Erik's winter ebiking needs? I've mounted that Plano storage trunk on the rear of my Liberty Trike. Having relocated the LT battery pack in the Plano trunk for better cold weather protection next to it's V1.6 10S3P companion hopefully keeping both packs closer to at least 45 F for sub-zero jaunts to grocery store, McD's, HD, etc with 12v heating pad that's still in developmental stage ... why do i always procrastinate

 

[john61ct]

Vruzend no longer relies on "friction fit", as of v2.1

V2.1 is still a "friction fit" (ask Micah if U don't believe me)

I did not say they aren't friction-fit.

With the addition of the through-bolts, maintaining connector pressure against the cell ends no longer **relies** on friction alone.

 

[Jan-Erik-86]

I see some of you recommend NCR, and some say forget the GA, so i'm a bit confused..

What about NCA or is that too mild-mannered? How close are you to narrowing your choice down to your final selection?

I'm not sure if i understand what you're asking? My confusion was due to people recommending NCR for long cycle life, but also not recommending GA (NCR) due to less then ideal cycle life.
I already ordered my M36 cells on 23. September, have been using the new battery for a while already, and I'm very happy with it so far.

Just wondering if a Kenda etrike ... https://www.addmotor.com/shop/m-330 ... would be more suited/sufficient for Erik's winter ebiking needs?

Although a trike may be practical, I prefer 2 wheels instead of 3. Besides, I already had a bike i was happy with, i just needed a new battery..

 

[eMark]

Vruzend no longer relies on "friction fit", as of v2.1

V2.1 is still a "friction fit" as was its predecessor the V1.6 kit which has 12 compression barrel bolts. The V2.1 kit only has 8 compression barrel bolts, but each is suitable in the right hands for a particular application.

With the addition of the through-bolts, maintaining connector pressure against the cell ends no longer **relies** on friction alone.

Whether a Nishi kit or Vruzend kit they both rely on a "friction fit" in that the buss bars are not spot-welded to the cell ends. Both the V1.6 and V2.1 rely on the buss bars being held in place with screw terminal nuts (friction fit) as well as the barrel bolt tightening screws. One disadvantage of these Vruzend kits is being restricted to a [larger] rectangle (with space between the cells) as opposed to a smaller pack of staggered spot-welded cells with the option of other shapes (e.g. bottle, triangle, etc.)

You sound like you're promoting the V2.1 (?shilling?) with your "no longer relies on 'friction fit' " giving the wrong impression to newbies. It's not that the V2.1 doesn't have it's place/applicaton as a beginner DIY build by a novice and/or for experimenting by a more advance builder or for use as a variable powerwall. Whether a Nishi kit or a Vruzend kit they definitely have a place and offer an advantage for certain applications. They are a "friction fit" kit, but that is not a bad thing anymore than it is a good thing.

Peace!

 

[eMark]

Vruzend no longer relies on "friction fit", as of v2.1

V2.1 is still a "friction fit" as was its predecessor the V1.6 kit which has 12 compression barrel bolts. The V2.1 kit only has 8 compression barrel bolts, but each is suitable in the right hands for its suited application.

With the addition of the through-bolts, maintaining connector pressure against the cell ends no longer **relies** on friction alone.

Whether a Nishi kit or Vruzend kit they both rely on a "friction fit" in that the buss bars are not spot-welded to the cell ends. Both the V1.6 and V2.1 rely on the buss bars being held in place with screw terminal nuts (friction fit) as well as the barrel bolt tightening screws. One disadvantage of these Vruzend kits is being restricted to a [larger] rectangle (with space between the cells) as opposed to a smaller pack of staggered spot-welded cells with the option of other shapes (e.g. bottle, triangle, etc.)

You sound like you're promoting the V2.1 (?shilling?) with your "no longer relies on 'friction fit' " giving the wrong impression to newbies. It's not that the V2.1 doesn't have it's place/applicaton as a beginner DIY build by a novice and/or for experimenting by a more advance builder or for use as a variable powerwall. Whether a Nishi kit or a Vruzend kit they definitely have a place and offer an advantage for certain applications. They are a "friction fit" kit, but that is not a bad thing anymore than it is a good thing.

Peace!

 

[john61ct]

I would never touch any version of either with a ten foot pole.

How's that for shilling.

You and I are not assigning the same meaning to "friction fit", but I'm not going to bother arguing with you over semantics, as we've clarified enough about the difference that those who may be interested can make a better informed decision.

Please do try to be less obnoxious.

 

[flippy]

I allready have two seat heater pads ready to install on my battery. I was planning on using a digital thermostat instead though, and was thinking of inserting them between the battery and the insulation instead of between the battery case and insolation. Haven't decided to use one or two yet though.

Flippy, altohugh not 100% efficient you could still use them with a small buck converter, which also lets you set the required power very easily (and it will be constant, regardless of battery voltage).
Most small and cheap buck converters are for less then 60v though, but for higher voltage you can often just use a normal 100-240V AC/DC adapter (but output current is often a bit limited).

or just buy the right voltage rating heating pad.

 

[Jan-Erik-86]

I suppose you could, but they do not appear as common and easy to get for a 48-52V system as for a 12V system.
Is there anything i'm not seeing that would make my option not work just as well with the only two disadvantages being a bit lower efficiency and an additional small box (the one i found is 70 x 38 x 31mm)?

 

[flippy]

you can buy a double voltage pad (so a 100v pad and run it at 50v) and just compensate with more power.

 

[eMark]

if you have the capacity you can drop the top voltage down to 4V for example and get thousands of charges.

Charging to 80% and discharging to 50% has become somewhat of an ebiking 'NCR' norm by those striving for at least 1000 useable cycles from a good sized pack (8P14S) made up of cells having at least 3000mAh remaining capacity (90%) at an average cell rating around 10A under moderate ebiking conditions (preferrably just 3 seasons NOT during winter).

What doesn't seem to be discussed a whole lot is discharging to 3.52V (25% capacity of GA or M36) instead of 3.7V (50% capacity of GA & M36) ... https://endless-sphere.com/forums/download/file.php?id=261831 ... Logic would seem to suggest that the 80% to 50% goal is leaving too much unused capacity ... well before the downhill graph slope begins ... https://electricbike.com/forum/filedata/fetch?id=52417&d=1522096223 ... whether it's the Sanyo GA, LG M36, or another NCR cell having a [realistic] nominal capacity around 3400mAh.

Anyone know of a comparative real-life cycle life test closer to 90% to 25% (depleting 2325mAh of the remaining 3060mAh of a 3400mAh cell) versus somewhat of an ebiking norm of 90% to 50% (depleting only 1550mAh of the remaining 3060mAh of a 3400mAh cell) ? Any real-life road experiences (or bench tests) as to how much a 90% or 80% to 25% is going to effectively shorten one's goal of 1000 cycles? Ask this because everyone talks about the advantage of only charging the pack to 4.0 or 4.07 volts (80%). What's the adverse difference in cycle life between discharging a NCR pack to 3.53 volts as compared to discharging it to only 3.7 volts ?

Picking up on flippy's post ("thousands of charges") is it still reasonable for an 90% to 25% discharge of a 3100mAh NCR pack to still provide 1000 useable cycles from the GA 10A & M36 5.1A (used to be rated at 10A) sizeable 8P14S pack if the same pack under the same ebiking daily use is able to "get thousands" of useable pack charges if with an 80% (4.00V charge) or 90% charge when discharged to 25% of remaining capacity (3.52V) ?

 

[flippy]

my testing (and from others here and tesla's battery wizards) concludes that top voltage is way more a contributer to wear then low voltage limit. if you keep a 3v lower limit and a 4.05~4.1V upper limit and keep the discarge within spec you will basically never wear out the battery. this does have a HUGE astrix next to it as this mostly applies to cells lower then 3000mAh. above that you will always see much more wear thne you ever will with the standard 2900mAh cells. the infamous GA's for example wear out considerably faster even while keeping under the 3~4.05V range. higher capacity cells wear out even faster.

so if you want a pack that lasts for a decade or more you need to build it with 2900mAh cells and keep under 1C discharge peaks. if you do that you will basically never wear it out. fun fact: 2900mAh is also considerably cheaper then GA's.

 

[eMark]

so if you want a pack that lasts for a decade or more you need to build it with 2900mAh cells and keep under 1C discharge peaks. if you do that you will basically never wear it out. fun fact: 2900mAh is also considerably cheaper then GA's.

Probably should have mentioned that my use fo 'NCR' is practically synonymous with NCA (LiNiCoAlO2) or is 'NCR' somewhere betwixt and between NMC and NCA, but closer to NCA (as if anyone really cares).

Ok, so now you've pulled what may be an interestng switcher roo (of sorts) that's got me kinda confused. Don't believe you were kidding ... "lasts for a decade".

Sooo, maybe it's time to ask if Samsung 30q is betwixt and between being an even 3000mAh (15A rating), thus not making it into the 2900mAh class ... "you will basically never wear it out" ... if treated right? Could one reason be why you lowered it to 2900mAh is to exclude 30q ? Don't want to derail this thread ... just curious what particular chemistry may be the line of demarcation (2900mAh) ?

EDIT Update: Doing a quickee check at IMR the one and only 2900mAh (among the 4 popular brands) is the Panasonic NCR18650PF 2900mAh 10A Battery. And in the smaller quantity i'll need only $3.75. Hopefully, the weather won't be to frigid when shipped from Houston to Minnepolis. Thank you so very much as i've been toiling over what cell to get. Was favoring the 30q, but at my moderate to conservative easy going speed the Pan. PF (10A rating) should be very happy . Will do some more followup on the PF, and unless i hear differently i'm looking forward to making some improvements to my DIY Vruzend build to minimize any brief periods of uphill voltage sag. Thanks again :thumb:

 

[999zip999]

1C would be hard to live with.

 

[eMark]

1C would be hard to live with.

When you get to be 75 years young you'll find that it will be easy to live with or maybe not until you're 90

Interesting comparison test on the Panasonic PF and LG MG1 ... https://electricbike.com/forum/forum/batteries/18650/71533-panasonic-ncr18650pf-vs-lg-mg1-2900mah-10a-cells-discharge-capacity-test

EDIT Update: Perhaps too hasty with PF decision, but Panasonic has such a great following. Don't want to be too quick to dismiss Samsung's 30q as it's a great cell.

 

[999zip999]

I'm 61 and my street bike goes 47mph on flats and 36mph on most hills. 7,000 watts. Funny I don't feel old yet.

 

[john61ct]

This "basically never wear it out" idea is just relative to the usually woefully short lifespans for most propulsion packs.

Personally, decades' calendar life and 6000-10000 cycles is my "normal use case" expectation, and what I plan for in care / usage parameters.

So a piddly 2-3000 cycles is no great shakes, and down around 300- 500 just ridiculously short.

So, again for me, getting from the "usual ebiking" latter expectation toward the former range there is not just a hand-wavy "do this and that and done", but requires spelling the various contributing factors out in detail.

However, there are too many such factors, inter-related or even at cross purposes,

so much so there's no possibility of hard research numbers for the various different example regimes,

best each of us can do, is just try to optimize all those factors where you can, and compromise as little as possible where your use case requires that.

 

[john61ct]

NCR is not really a specific chemistry as such, as with IMR, older "spinel" vs INR, more like model# related labels describing tweaking transitions, have not yet cross-correlated to the actual chemistry names, e.g.

NCA is, LiNiCoAlO2
NMC (NCM) LiNiMnCoO2,
e.g INR18650-29E
also CMN, CNM, MNC and MCN all very close
LCO, LiCoO2
LMO, LiMn2O4

I don't usually reco Battery University, but their intro page here is not a bad summary

https://batteryuniversity.com/learn/article/types_of_lithium_ion

 

[john61ct]

Top SoC is indeed critical to longevity.

There are an infinite # of "recipes" for getting to a specific given SoC

a termination voltage setpoint is nothing definitive

If CC only, "charge to and stop", will greatly vary depending on from-Bulk current C-rate

If holding Absorb time, CV to a trailing C-rate stop-charge setpoint ("endAmps"), then much more objective a measure.

Tossing around "80% SoC" is really not useful, unless for **that** pack of **those** cells, you've calibrated with precisely timed CC load testing, and then taken SoH% into account as the pack wears.

The difference between a

highly stressful stop- charge recipe, close to the vendor-spec'd "do not approach except for benchmark testing" maximum on the one hand, and a

nice gentle protocol conducive to maximising longevity

is really not sacrificing more than 6-8% of actual usable mAh capacity.

The **really** critical point is not just a gentle definition of a "working Full" stop-charge point, but that the pack **should not sit there** for any length of time. Only go there just before the load starts pulling SoC back down.

While not actively cycling LI batteries, keep well below full, say 3.4-3.7V.

The percentage of their life spent down there as opposed to anywhere above 3.9V, will have a great impact on longevity.

Whether your termination voltage is 3.05V or 3.15V is much less of an impact.

 

[john61ct]

The other factor flippy correctly emphasizes, is that energy density is counter-indicative of longevity.

If two cells are basically the same chemistry family, but one has been tweaked so that its Ah's are 25% lighter, that very likely will mean a much shorter cycle lifespan.

That's why, if longevity is important to you, stick to the lower mAh capacity versions.

This is famously true for lead as well, if a pair of Deka GCs at 220Ah weighs 127lbs, while the Wally world "deep cycle marine" same capacity labeled pair is only 105lbs, pretty much proves that the former will IRL last at least 3-4 times as long everything else is held equal.