Cell mAh capacity testing. Best/cheapest/fastest way?

S

ssamish

New member
Joined
Feb 19, 2017
Messages
9
Hey guys, I'm in to the testing phase of a bunch of cells that I have harvested from laptop batteries.

Out of roughly 1200 cells, Ihave sorted a few hundred to the top of the pile that look like good candidates for my current project.
I'm about a week in to charging them allup to about 95% or sowith a couple of hobby chargers, weeding out the warm cells. etc. The charged pile is starting to build.

I've searched all over the confuser, far and widefora cost effectivemethod to conduct testing with a resistive load type discharger. I'd like to to 10 or more at a time, as it would niceto have my batterybuilt this decade.

This is what my search has come up with so far:

ZB2L3 v2.1 Module

image_zcchpt.jpg


Pros: Seems pretty accurate and reliable from looking at reviews and youtube tests. Cheap! around $6aud per module
Cons:Out of stock on banggood, long postage times from other sources (China). No casing.

Opus BT-C3100(or similar)

image_hxvxco.jpg


Pros: Recharges cells, cutting out the need to go back to the hobby chargers. Compact. Available.
Cons: The conclusion of that test in the link states that the peak currents can be a bit high for charge and discharge, but acceptable for what Iam doing. More expensive at around $60 a piece. I would want at 2 at least.

"Mobile power" tester

image_ytjwdz.jpg



Arduino? Though all projectsIhave seen only do one at a time.

What do you guys use for bulk testing?
Thoughts on a good/cheapset up?
 
Just order yourself a bunch of the ZBL23 modules. Its the only way to do it cheaply. You will have to wait a few weeks to get them, but then you will be processing your batteries a lot faster. I bought 5 and wish I had ordered more. You will also need single cell holders. I found that the resistor supplied with those units does not put much of load on the batteries. I used a small automotive light bulb and a resistor to get a 1 amp draw.
You can also use an IMAX B6 or similar hobby charger. I have one of those as well. The Opus looks like a good option but I would only buy one. Its good to have to verify the results against your ZBL23. Hobby chargers are good because they have other programmable functions so its nice to have at least one along with a bunch of the cheap TP4056 and ZBL23's
 
With the Opus I believe it's a 500ma discharge
I've ordered 4 of the hobby chargers and waiting for them to come in.
But you can have to cells in parallel and use a resistor to draw 1a (so 500ma from each cell).
The opus is probably more reliable in terms of consistency while the hobby chargers can be more or less accurate with each individual one.
I'd say cost wise, go for a few hobby chargers and just cut down cost and be more able to process more cells
One opus is $30ish ($usa) if I remember correctly while some chargers can be $4 so you can have more
 
I'm becoming very wary of these mass produced chargers like the opus, b6 clones. etc.
I've got 3 different ones now, 2 from my local hobby shop, one ebay cheapie. None of them seem very accurate/consistent. But that is a topic for another thread.

Does the TP4056 have a cc/cv charge cycle? Or just a 1A charge up to 4.2v?
 
Even with the tp4056 I have some that stop at 4.15 some 4.12 and others at 4.19
It's really quality control once it's in your hands
 
ssamish said:
I'm becoming very wary of these mass produced chargers like the Opus, b6 clones. etc.
I've got 3 different ones now, 2 from my local hobby shop, one ebay cheapie. None of them seem very accurate/consistent. But that is a topic for another thread.

Does the TP4056 have a cc/cv charge cycle? Or just a 1A charge up to 4.2v?
Click to expand...


Ok so you already have those...buy a bunch of ZB2's and TP4056's and then you can compare. They are so cheap, its really the only way to bulk process.
For my TP4056's, I use 3-cell holders. This way I have a choice to charge 1 cell @ 1amp, 2 @ 0.5 amps, or 3 cells at once, at .333 amps. I dont get why people only put a single cell holder on these. I cant get up in the middle of the night every 2 hours to swap batterys, so with triple holders, I can just wake up in the morning to 30 charged cells.

I double check some of my cells to compare against my B6, and that gives me a range of the cells true capacity...there is no absolute value for the capacity of a cell, there will always be some variation in repeated tests on the same cell, even with the same charger/discharger. Battery capacity is variable and depends on multiple factors, so dont expect too much consistency.
Remember that a variation of 100mah in a discharge test is only a 5% difference for a 2000 mah battery, so if you're getting consistent results with 100mah, that's pretty good.
 
Firefrog: Opus is 1A. I run 10 of them.

Most important is that you stay with same type of tester so you get consistent numbers and that you test all you use so they produce same numbers on same cells...
 
FireFrog said:
Even with the tp4056 I have some that stop at 4.15 some 4.12 and others at 4.19
It's really quality control once it's in your hands
Click to expand...

Yeah, see, this is what I'm wondering. But I'd still get something that's recommended by people here,over the terrible crap that keeps making it's way to my charging station. This is also better than catching a charger at 4.28v still pushing 5A into a group of fresh looking 2015 Samsung:-/

@Daromer,consistency is why I asked the question in the first place. I realised very quickly that every hobby charger is a slightly different beast. Kicking myself that I didn't just go for a couple ofgenuine B6. But my hobby charger budget is spent, lesson learned.
@APD, good call on the cradles, I was thinking that while looking at the spec sheet on the TP4056. Though I've been making my own,out of the springs from 10xAA cradles from jaycar and scrap timber I have laying around. Not pretty, but functional and super cheap.
What's hilarious about that, is that jaycar sells a 2x18650 cradle for $6.50 or the 10xAA for $4.
 
What are ZBL23's? I can't find anything about them. Searching for "ZBL23 charger" returns this thread on google.
 
Korishan said:
What are ZBL23's? I can't find anything about them. Searching for "ZBL23 charger" returns this thread on google.
Click to expand...

There is an eBay link above the photo. ZB2L3 not ZBL23 ;)it's not a charger, just a resistive load discharger
 
Ohhh, ok. I just selected and google searched from the post. "Someone else" misspelled it ;)

I've seen that one used all over the place.
 
Arduino Nano has 6 A/D converter channels. I've done a high precision thermometer with a Nano + temperature sensor and the trick to getting reliable results is to take a lot of A/D readings and average them, that really squelches noise. For something like battery discharging the Nano A/D is fast enough you can take thousands of readings for each average. There is a trick with the Nano where you can change the reference voltage for the A/D to 1.1V (I think it is) rather than the normal 5V by setting a bit in a register and you can increase precision that way at low voltages.

It really depends if you want absolute numbers or relative ones.. I'd find an automotive lamp that discharges a single cell at about the current you want. An Arduino, an eight channel relay board (since I haven't seen six channel ones) and six of your chosen lamps could provide the relative numbers you need. Each of six cells is read by the Arduino as it discharges through its own lamp/relay combination and the Arduino turns off the discharge as the low voltage cutoff is reached and records the time necessary for the discharge. Longer discharge time equals more capacity, you won't know the exact number but you will have a ranking for the cells.

Throw in a current sensor for each channel and a bit more code and you could have a more absolute reading at somewhat higher cost for the unit by measuring the current and summing it over time since it will vary according to the dropping cell voltage. Again, for something like battery discharging you can just use brute force methods and don't have to get fancy with the math since you have what amounts to eternities in microprocessor terms. Of course a resistor as the load will allow you to just calculate the current directly from the battery voltage but that would be too easy.
 
I tested a few spare automotive lamps I have in my car kit, a brand new Sylvania 7506L draws 0.95A @ 4.10V and 0.88A @ 3.5V for an average of about 0.91A. That sounds about the right number for a discharge load. A pair of those lamps at my local Autozone is just over $6.00.. So call it $20 with taxes for six channels worth.

I have a Nano and a four channel relay board, I'll try and put together some code to run a single channel, I've already got a lot of it written for a thermostat app I did. Get that running successfully and expanding it to more channels should be fairly trivial.
 
Sounds interesting. Too bad I have little experience with circuitry and zero with arduino.
I've ordered a dozen of the ZB2L3. But you know what shipping from china is like.

In the meantime, I do have a box of arduino gear laying around that a mate gave to me. He was intending on doing some diy quadrotor stuff, but was put off by the steep learning curve.
I have a bunch of nanos, UNOs, dc/dc, motor controllers, mosfets, resistorss, displays etc. etc.
I know for a fact that I have all the bits there to make at least one capacity measuring device.

I've sat down a couple of times in the past few days to try and get one running, but this is a deep rabbit hole..
The project that I'm working on has a lot of elements. The battery is only part of that. So I can't justify learning yet another skill right now for the simple task of testing my cells.

But on the other hand, this might be a good chance to get a glimpse in to the world of "DIY" electronics.
 
If you have a Mega 2560 that one has sixteen A/D inputs and you could do sixteen at once with the same effort as six..

Once the software is written putting together a system is almost like building Legos, you just plug stuff together, maybe solder some leads on the lamps, a few wires into screw terminals.

It would be one simple circuit with a battery, a lamp (or resistor) and a relay repeated sixteen times.

I cleaned up some old code I'd written for something else to the point I can start making it do what I want to do for a discharger without getting distracted by code that doesn't need to be there. I'll probably work on that some tomorrow, tonight I'm sacrificing brain cells to Bacchus. Arduino programming is pretty dense, I've programmed a bunch of different stuff all the way back to the 6800 and then the 6809 but I don't like C+ or whatever it is that the 'duino runs and the IDE barfs up some bizarrely inscrutable error messages. But they are cheap and pretty powerful once you can wrap your head around the language and get past the worst of the dumb mistakes you can make with it.
 
I think its great that you guys are looking at ways to save money and do it faster etc.

I ended up pre-charging unknown cells with an array of thetp4056's

I bought 4 0pus's for ~37USD each from amazon. Which I use exclusively for capacity testing.
I have retested cells with them a few times and they are all pretty accurate.
Even if the opus's are say off 10% that puts you in the ballpark to classify your cells. I read early on, not to mix and match types. So you get consistentresults.
I charge/discharge mine all at 1 amp.

for POST 0pus recharging I have used a variety of things, the tp4056's and I built a couple of variable parallel chargers.

I am not made of money, but having said that My TIME is worth money too. In fact one or 2 more 0pus chargers might have spedmy process along at some points. It is definitely the bottleneck.Your mileage may vary, I work from home office, so I HAVE the time to swap them batteries ever hour or 2.

remember that old adage:
Fast good cheap, pick any two. :)


My work has gone this way so far:
Step 1 initial charging done by tp's
Step2 Into the opus for discharge capacity testing.
Step3 When the 0pus displays the mAh I mark the cell, pull itand put in another cell to be capacity checked.
Step4 Bulk recharging via this homebrew:




image_hrldiu.jpg
 
This is the type of thing that I'm talking about:

DIY Arduino 18650 Capacity tester

Adam's 18650 checker

Adam's checker update

If someone here is keen on helping me modify the code to fit the parts Ihave, Iwould be super grateful!


image_uonbbr.jpg


This is what I have, though I can only find a usbcable for the Uno, so that rules out the nano. This is where Iam hitting the wall.
I have the other parts to make Adam's version, though only the 4 pin oled. The one resistor is the same, another is different. I have the same mosfet too.
It seems like a good little project for a first timer (getting something to work with minimal knowledge)

Elwood, you wanna see a dirty ghetto chargingstation? Check this out:


image_vxjwvc.jpgimage_duncsg.jpg


I Cannot recommend any of these chargers btw!

Especially the "690AC". It looks like a charger, but that's it. It will cost you more in the long run with its insistence on overchargingcells!
It charges ok initially. But you gotta keep your eye on it! It will push the cells close to 4.3v, kicking in the cv phase way too late. With no way (that i know of)to adjust. BUYER BEWARE!
But hey, get one if you enjoy hovering over it with a multimeter for days on end....
 
Korishan said:
What are ZBL23's? I can't find anything about them. Searching for "ZBL23 charger" returns this thread on google.
Click to expand...

Typing error I think...see OP , should read ...ZB2L3
 
You must log in or register to reply here.
Back
Top