Eastwood’s first battery build

I pulled this photo off a facebook forum, It shows why p-groups cannot be glued together. They need a separating insulator/ This pack is still insulated, but now that the PVC on the right cell is ripped, a tear on the left side will short circuit the right hand cell,

The top of the left hand cell is the positive end of the group. It's missing o-rings under the nickel. Helps prevent the nickel from contacting the end of the case.



View attachment 332176

Thank for commenting and sharing the pic 👍
I’ve purchased the cell spacers, been working on the battery shape/configuration. I’ll probably end up using hot glue or RTV in conjunction with the "cell spacers", just to prevent and twisting or shifting within the pack due to vibrations from off road use.

Yeah, for sure I’ll be using O-rings on the positive end of the battery 👍.. Maybe double stacked o-rings
 
It's also meant just in case you damaged that thin plastic sleeve either through building the pack or if it happens to get Jarred or rubbed against each other there's a protection from one P to 2p to 3p as they are just 44ah of 3.6 charged to 4.2v. this sentence isn't just for you. Somebody else might be following.
 
Do you mean, cell capacity 4 line tester, like this pic?
View attachment 332189
Nope, the amazon link you posted previously (the YR) is a 4-line tester. You'll see it when you get it: each probe is actually 2 spring-loaded contacts, so when you press each probe end onto the cell contacts, you're using 4 points of connection, rather than 2 like a standard multimeter. Hence 4-line.

For cell capacity tests, cylindrical cells, you'll want the Opus BTC or the Liitokala. Pros and cons to both, and the consensus seems to be that the Opus is a bit more accurate, the Liitokala a bit cheaper and easier to use (with the latter being a matter of opinion, of course). For either, you'll want to only be using the charge test mode: the machine charges the cell all the way up, then discharges it, and displays how many mAh were discharged. After displaying the discharged capacity, it'll charge it all the way up. You can let it do that, of course, no harm, but I like to pull them when I see the volts around 3.7-3.8; there's no need to top them off, and they store better at midrange voltage. And of course, you'll need to rebalance all your cells before you start building your pack anyway, very important.

Before you do capacity tests, you may want to consider checking for self-discharging cells, if you haven't already. Charge all the cells to 4.2v, set them aside for a week, 10 days, 30 days. Check all the voltage again. They should definitely all be above 4.1v, and some picky testers will insist on everything being about 4.15v, that depends on your criteria. This is where an IR test is beneficial: self discharging or somewhat weakened cells will almost always have an IR reading higher than spec. So, if you got a cell that self-discharged to 4.09V, but it has a great IR reading, maybe keep it. If you got a self-discharged cell at 4.11, but the IR reading is suspiciously high, maybe toss it.

Lots of guides online, here and elsewhere, for cell testing procedures. Whatever you do, pick a procedure, maybe write it down, and stick to it for every single cell. Regardless of what you decide as your pass/fail parameters, don't change your sorting procedure or methods halfway through all your cells, because you risk ending up with an unbalanced pack. Ask me how I know.

Both testers, when you place the cell into the springs, will display and IR value. Ignore it. They're not designed to test IR, it's an educated guess at best.
 
Nope, the amazon link you posted previously (the YR) is a 4-line tester. You'll see it when you get it: each probe is actually 2 spring-loaded contacts, so when you press each probe end onto the cell contacts, you're using 4 points of connection, rather than 2 like a standard multimeter. Hence 4-line.

For cell capacity tests, cylindrical cells, you'll want the Opus BTC or the Liitokala. Pros and cons to both, and the consensus seems to be that the Opus is a bit more accurate, the Liitokala a bit cheaper and easier to use (with the latter being a matter of opinion, of course). For either, you'll want to only be using the charge test mode: the machine charges the cell all the way up, then discharges it, and displays how many mAh were discharged. After displaying the discharged capacity, it'll charge it all the way up. You can let it do that, of course, no harm, but I like to pull them when I see the volts around 3.7-3.8; there's no need to top them off, and they store better at midrange voltage. And of course, you'll need to rebalance all your cells before you start building your pack anyway, very important.

Before you do capacity tests, you may want to consider checking for self-discharging cells, if you haven't already. Charge all the cells to 4.2v, set them aside for a week, 10 days, 30 days. Check all the voltage again. They should definitely all be above 4.1v, and some picky testers will insist on everything being about 4.15v, that depends on your criteria. This is where an IR test is beneficial: self discharging or somewhat weakened cells will almost always have an IR reading higher than spec. So, if you got a cell that self-discharged to 4.09V, but it has a great IR reading, maybe keep it. If you got a self-discharged cell at 4.11, but the IR reading is suspiciously high, maybe toss it.

Lots of guides online, here and elsewhere, for cell testing procedures. Whatever you do, pick a procedure, maybe write it down, and stick to it for every single cell. Regardless of what you decide as your pass/fail parameters, don't change your sorting procedure or methods halfway through all your cells, because you risk ending up with an unbalanced pack. Ask me how I know.

Both testers, when you place the cell into the springs, will display and IR value. Ignore it. They're not designed to test IR, it's an educated guess at best.
Very useful information! Thank you sir 🫡
That’s a great point you brought up about charging all cells to 4.2 and check a minimum of 10 days later to see if they have selfed discharge. Great point, makes perfect sense.

It seems like quality battery packs or truly separated from the pre-work with testing all the cell individually. I certainly won’t cut any corners in this area, as I want a solid pack to last me for several+ years.

Edit: Wonder if I should use more than one battery charger since I’m building this pack with 300+ cells. Seems like it would take forever to fully charge all cells with just one charger that has slots of 4 cells?
Thanks!
 
Wonder if I should use more than one battery charger since I’m building this pack with 300+ cells. Seems like it would take forever to fully charge all cells with just one charger that has slots of 4 cells?
That depends on how much you want to invest, in both time and money.

Here's a scenario. You have 300 cells. The capacity tester can fully charge 4 at a time. Assume 3000mah cells, and you only want to deal with this during waking hours (good idea for safety anyway). Test at 1000ma per cell. That means 4 cells get fully charged in 3 hours (maybe less, since you bought them probably at storage, 50% capacity, but call it 3 hours anyway). That means you can charge 16 per 12 hour day. 16 fully charged cells per day, put them in a container, mark it with the date. Next day, same thing, charge 16 cells, place them in a box, date it, store it under the previous days box. See where this is going?

IMG_20220602_092215 (1).jpg
Here's an example of what you could use to sort each day's cells and keep them organized.

19 days later, you'll have all your cells fully charged. Open the box from day 1, check for self discharge. Set aside any cells lower than 4.1v (or whatever you decide your cutoff is). Start a capacity test now for each cell. This will take 2x as long as simple charging: you're starting with a full cell, discharge 100 percent, charge to at least 50%. This is the other reason why I try to catch them when charging after the capacity test is complete, it frees up a slot in the charger for the next cell to test. Sharpie the capacity result on the cell, set aside in groups of 50mAh. This will take longer than 12 hours per 16 cells. When you've finished capacity testing all the cells from day 1, one day will have passed, and you'll be able to check for self-discharging cells from day 2. Continue until the remains of the deceased horse are well and truly beaten.

So perhaps 30 days. Plus the first charge of 19 days, means you can finish testing 300 cells in 50 days on one charge. If you get 2 chargers, you could finish in half of the time. Maybe that's worth another 50 bucks to you. But if you get 3 chargers, keep in mind that even though the first charge for self-discharge test will get done in 3-4 days, you still have to wait 7-30 days anyway. And when you get to capacity testing, when it comes to time investment, you start to run into diminishing returns, as you only have so much time per day to babysit the chargers and switch them out for the next cell, I'm sure you have other things going on in your life. Consider that you're building an e-vehicle, so maybe you can get just 1-2 chargers and use the downtime to work on the other parts of your build.

When do you perform the IR test? Up to to you. I know recyclers who perform an IR test first and foremost, before ever checking voltage. They immediately discard/recycle cells that have high IR, because IR is an excellent test to give you an idea of general cell health before you even do a capacity or voltage test, so they don't want to waste time or energy bothering with a self-discharge check or capacity test on a cell that's unlikely to pass anyway. My opinion is to perform the IR test after a capacity test. If the IR result is high, mark it with IR next to capacity, and set it aside for a powerbank, flashlight, laser pointer, ect., something low draw that doesn't have a high-amp requirement. If it's really bad, and a no-capacity cell, save them for spot welding practice and settings.

I recover and test used and recycles laptop batteries. I have 1 opus and 3 Liitokalas, and sometimes it's great to have 4, if I have hundreds of cells to process, but I could exercise a bit more patience and get by with 2. I also have 3-4 various semi-homemade testers and dischargers, for manual testing or testing of unusual sized cells.
IMG_20220207_202113.jpg
I parallel 40 cells and hook them all up to 1ea TP4056 module, this super super slowly charges low-voltage cells for gentle recovery. I use the same 40 cell holder, attached to resistors, to discharge and balance 40 cells at a time.
 
Eastwood the cells that you bought go through quality control but if you like it would cost you a lot of time and money to test 340 cells. I like to have access to my individual parallel cells. With balance wires for 20s you would need 4-5s hobby style cables. Got to go
 
Assume 3000mah cells,

Actually, the cells are 4200mah, Molicel P42A,
So even longer to charge 😅

Ok great, all of this really breaks down the testing procedure. Thank you so much! I’m sure your reply will also help others as well that are reading through this thread 👍

Considering I’m buying brand new cells do you think it’s necessary to do the self discharge test? I asked the seller today some questions, and they said that they test each individual cell and also the cells would all be from the same batch of production.
18650batterystore.com

I most definitely will do the capacity test and IR test but I’m not sure what you or others think about doing the self drain since the cells are new?

Thanks again for all your help, very useful info!
 
Eastwood the cells that you bought go through quality control

Yeah, that’s what I’m thinking too, not sure if the self drain test would need to be done since their new cells that go through quality control. Certainly wouldn’t hurt, but would take up a lot of time 😅
 
Yes do the self-drain test find out the voltage of each cell if they're all the same go to the next step.
. I know I'm a bit of a cowboy but I don't think you would need to. One idiot's opinion. So either leave it but good luck with your build and I'm going to keep reading up on your progress good luck.
 
Yeah, that’s what I’m thinking too, not sure if the self drain test would need to be done since their new cells that go through quality control. Certainly wouldn’t hurt, but would take up a lot of time 😅
What are the chances that the cells you purchased from a reputable supplier are all good with no defects? Low, quite low indeed.

What are the consequences to your build if even 1 cell is a self-discharger? The whole pack loses capacity. Each parallel group is only as strong as it's weakest cell, and the whole series capacity is limited to the weakest series pack. You're also not likely to notice a self-discharger until months into use, as self-discharging cells die fast, and if you want to repair it, its a whole pack rebuild to find and replace the culprit.

The consequences of missing a self discharging cell, however, is not going to be safety, just capacity and performance. Doing an IR and capacity test on every cell will absolutely weed out the possibility of an unlikely dead-shorted, internal shorted, or heater cell from causing physical harm to you or your build, so you're good there. It's just a question of whether you want to risk it for the performance of your build.

Maybe I'm just an untrustworthy bastard who has no faith in someone else's supposed testing procedures, so take my advice with a grain of salt. It's not like you're cobbling together used unbalanced RC packs with hot glue, you're doing a ton of good steps, and its true that it's really unlikely that you have a self-discharger in your new purchases. If it was my build, I would take the time.
 
What are the chances that the cells you purchased from a reputable supplier are all good with no defects? Low, quite low indeed.
Maybe I'm just an untrustworthy bastard who has no faith in someone else's supposed testing procedures, so take my advice with a grain of salt. It's not like you're cobbling together used unbalanced RC packs with hot glue, you're doing a ton of good steps, and its true that it's really unlikely that you have a self-discharger in your new purchases. If it was my build, I would take the time.

OK I will take your advice and do the self discharge test as well 👍
 
Last edited:
These are my thoughts and take them as my opinion
The cells are charged to the same voltage. So when you take them out of the box. They should all be the same voltage. So find out which the lowest voltage and knows maybe the ones you should eyeball to test for later voltage drop and or capacity and or internal resistance. But that is just getting you out of work.
 
These are my thoughts and take them as my opinion
The cells are charged to the same voltage. So when you take them out of the box. They should all be the same voltage. So find out which the lowest voltage and knows maybe the ones you should eyeball to test for later voltage drop and or capacity and or internal resistance. But that is just getting you out of work.
Not wrong. I will add, however, that they were likely shipped to you at storage voltage, 3.7-3.8v. At the middle of the discharge curve like that, if you do have a rare self-discharging cell, they could potentially exhibit self discharging that is hard to see with a voltage check. Because the discharge curve is relatively flat at that mid range, voltage doesn't drop much.

The reason why you go to full voltage for a self discharge check is because at 4.1-4.2 volts, the voltage drops off hard with even a little bit of discharge, so it's much easier to see if the discharge is faulty or normal.
 
I have many years with lifepo4 A123 20ah 9yrs 1,500 charge cycles as 24s with no BMS. As it is 20ah cell but really 18 ah to 19.5 new..Then over discharge and cost me 4cells.
But different chemistry different behavior. Those could be charged to 3.6-3.7v but would sit at 3.5v
Now my Samsung 40t 20s 6p should be 24ah. But only gotten 18.5 ah. I think counterfeit cells . As getting less ah now. Paid $1,100 with Bluetooth BMS.
Also have spaim08 20s 1p advertised as 8ah no BMS. But get an ez 8ah maybe more.
So chemistry matters how old they are matters and are they real or counterfeit.
It's only a price of a wrapper.
So source yourselves from unknown manufacturer.
The guy who built my pack in USA is bicycle motor works. He has a nice wife but he won't talk to me about it.
I I give you more than just good luck.
A battery it's just a lightning bolt in a box that you want to take out bit by bit.
That's why next time a little buy a spot welder and try my hand.
Two my batteries are just bolted together and the other is the N.E.S.E kit 14s8p just sitting BAK 18650 cell 20amp.
I had used Makita Samsung cells made garage fire.
Never never used cells.
 
Last edited:
The reason why you go to full voltage for a self discharge check is because at 4.1-4.2 volts, the voltage drops off hard with even a little bit of discharge, so it's much easier to see if the discharge is faulty or normal.

That’s interesting! I’ve always noticed that my battery drains much faster when it’s fully charged say compared when it’s at 50%. So basically when I got my 72 V battery fully charged at 84 and I go full throttle a couple of times it seems like the battery voltage drops much quicker compared to going full throttle say at 72-ish volts.
 
Yes let's use this as a learning experience you have a 72 volt pack how many ampere hours and what sells if you don't know what sells please give us the company and where you bought it or did you build it DIY.
And yes lifepo4 has lower voltage and more size and weight. But has a very constant discharge and higher final discharge I mean you can feel in your backside 101865 back is going so if you have a 20 amp your back or bigger you just throw bounce wires on it but my new favorite is that little elephant BMS. But at 65 I still don't know how to use it.
 
Harrison where do you source your cells and what are you using them for what is the highest amp per parallel group and voltage that you use them in. As I use my 200 amp packs at 80 or 60 amp.
 
I haven't tried a potentiometer based throttle yet, but was wondering, and maybe someone who uses one knows, if they potentiometer is a simple linear type or if they are nonlinear (log, inverse log). I'd like to see a logarithmic pot throttle to make the lower range output more controllable. I can use my CA to introduce delays or emulate the different throttle control types, etc., but I'd rather have direct control.

potentiometer_taper.png
Back in the day, like 1980's, the only way to get a motor controller to not smash into a wall, I made my own potboxes using a audio taper (logarithmic, like decibels) throttle pot, driven by a cable or directly with a linear travel (slide, like EQ in a studio) audio taper pot. We didn't have microprocessors or ramp control algorithms so the audio taper was awesome for low end control and a zippy mid to full throttle 'ramp'.

Now, most controllers have decent ramp tuning. But I still, on a off road motorcycle, prefer a 1/2 turn, slow M/C trials throttle (Domino) that cable drives a 5k resistive rotary throttle position sensor (TPS) off say a Honda Civic. Electrics can be pretty snappy compared to gas and this 'detuned' control is easier to manage with a hyped out motor/controllers. Less looping! But maybe I'm just getting old....
 
Last edited:
That’s interesting! I’ve always noticed that my battery drains much faster when it’s fully charged say compared when it’s at 50%. So basically when I got my 72 V battery fully charged at 84 and I go full throttle a couple of times it seems like the battery voltage drops much quicker compared to going full throttle say at 72-ish volts.
Indeed, look up pictures of graphs of discharge curves for li-ion or LFP to see it more clearly
 
Yes let's use this as a learning experience you have a 72 volt pack how many ampere hours and what sells if you don't know what sells please give us the company and where you bought it or did you build it DIY.
my current 72 V battery is a UPP battery. Actually, it’s two batteries in parallel of the same exact packs. So two 72v 20ah that gives me 40ah for my Enduro clone build. I typically charge these batteries all the way to 84 V because I didn’t plan on these batteries for lasting several plus years, it was sort of a temporary set up but end up using these for a while now. I will be retiring those two batteries in the next year because I don’t wanna be messing around with cheap UPP pack that’s been pushed to the max.

This battery build coming up, I won’t be fully charging except when I need to balance the cells. Probably 4.1v per cell, 82v for most charges.
 
Last edited:
If you have a programmable BMS, it may be able to balance at lower voltages, to allow for the specific situation you want to use. There will then be setpoints for this in the setup software.

Then you never have to fully charge the cells (and you can set the BMS cell-level LVC higher so they are not discharged as far, either).


Also...if you use well-matched cells to start with, where all the cells have identical properties (internal resistance, capacity, etc), you won't need to balance them until they age enough to no longer be identical.
 
Back
Top