The video at Philipp Seidel's blog (linked in the description) suggests putting some solder on those long lanes that pass the power across the board so they can handle the additional current from daisy chaining the boards together. I was thinking about adding a wire before I saw his video. He also uses a diode to protect the TP4056 from a reversed battery. The diode would short circuit the battery when it's backwards and blow the fuse.
I just build an 18650 charger for 48 cells - using a 30A 220v to12v dc converter and three 9A voltage regulators - all set to 4.2V. I made the cells in parallel - working just fine. I only need to check from time to time to see if one cell gets warm - that cell I take out. I only run in the daytime when I am around - this way I can charge 48 cells per day.:)
If I recall the Tp-4056 boards come default to charge at 300ma. You have to change a resistor to do 1amp. I am using a 30amp supply for my 36 slot charging unit no problem. Looking good, good luck
I just wanted to suggest you check out a metal conductivity table. I've done so much electrical work that I instantly had regrets about after reading one. One fact that blew my mind was that although brass contains 70% copper, it only conducts 28% as well. Tin is 15% and lead 7%. I definitely prefer leaded solder, and due to that, use silver-bearing lead solder instead. There's also silver-bearing lead-free that also contains copper, and likely conducts the best that I hope to try in the near future
You should defiantly add more solder in the holes of the TP4056 boards and You probably already know this but you should try Solder Flux it works so much better than that Solder Paste you're using. That stuff is more for SMD applications, but it does work theirs better stuff out there that will make soldering so much easier. Flux is the best thing ever created for soldering lol I can live with out it. Try the Pure Rosin ones there my favorite, you can use it to make liquid flux to. Good video tho. Gotta build one someday i'll probably add some header pins on the boards. my 10 Lii-500s aren't enough lol
Why don't you just solder a solid copper wire of the appropriate size that you would need along the traces? That would go a long way in the protection of your boards. Don't you think?
your original charging board was tacky when you built it lol. The negative track is actually the entire front of the board not just where the silver is, that is just to add some solder on it, you can put a solid piece of wire to beef it up! maybe I should redesign that board for the LM2596 CC buck boards I am using!
@@happer2009 he's asking what do you use to capacity test them. The TP4056's brother is the ZL2B3. Not quite as cheap but on Aliexpress or Banggood you can get them for like $4 each. They aren't the most accurate testers but they are fairly consistent between each other. My testing has shown they are rather optimistic giving me results several percent higher than what they should be. The ZL2B3 isn't constant current, its a simple resistor pulling down the cell so it can be hard to compare to other testers
Why don't you just install some protection diodes at each battery charging board so you don't blow out your boards? These diodes are very cheap and not a lot of work soldering them in place. Just asking mate.
protection diodes or using a diode will create a voltage drop, Around 0.6v off the top of my head. Which will cause a problem with the charging circuit unless you somehow created an offset in the chip. Remember the tp4056 is wanting to hit 4.2v CV and then do CC after that, so if the voltage wasn't able to get up to 4.2 held back by the 0.6v voltage drop it will cause the cells to never fully charge, That's why I didn't use diodes as it messes with the tp4056.
Hi Mike. I have tried charging 20 x 18650 all in parallel with a 4.2V 10A current limited power supply and I have had no problems. I use Rui Deng DPS5015 as the power source and stop charge when current drops below 100mA. May pay to put 0.5 ohm in series with each cell to limit surge currents when different voltage cell are paralleled but so far I have had no problems. I was wondering why you have to use 4056's?
Hi Peter, The TP4056 gives each cell 1amp and each is independent of each other, Meaning if some finish first they can be removed and another put in to charge. With the idea of 20x in parallel you will have to wait until all have finished or below 100mA. The problem with the 100mA cut off is if the cells overheat during charge and some never charge because of this the current will not fall below your cut off being 100mA. 8 out of the 20 Sanyo cells not charging should do that. Charging cells that are all around the same start voltage and are known to be good are fine to charge in your 20x parallel but I wouldn't be putting cells pulled straight out of laptops in there, Some will not charge, some might be dead shorts, some might be 1v with other at high 3s, My advice would be always charge unknown cells independently and connected to there own charge circuit for the first time. This also helps with speed because as soon as one finishes another can be put in, as they all have different capactity some will finish much sooner than others and some will never finish and are binned.
Thanks for your reply. Your system is more effective in weeding out the really bad cells. My system is very simple so only good for non leaky cells. I have only tested about 500 cells and generally dump any cell with
if they are heaters then 1 they will be hot and two all other calls would have finished charging but the bad ones will not be so they get binned. Some do get hot. No i dont only charge when i am around. Its runs 24/7 charging cells. Never had a problem after doing 1000s could be getting close to over 10,000 these days
Thank you for your response! Not challenging your methods, this build is awesome, just trying to understand. If they charge overnight, wouldnt some heaters have the chance to cool back down before you got to them in the mornings? And surely all 40 cells aren't beginning at the same voltage so wouldnt varying charging times be normal and not simply a sign of an internal short?
No the cells don't cool down, The cell gets hot and this causes the charging circuit to keep charging the cell and never cuts off. The charging circuit will only cut off to that intervidual cell when the cell drops to under 100mA in current draw. Given the cell is using more current than that in the form of heat then it never turns off and the cell stays hot no matter how much time pases. So its very easy to tell. Also if a cell did reach charge and then drop down in voltage to under 4v then it will start charging again this cell might not get hot but when all other cells are charged bar 1 or 2 then its also easly to tell that cell is not holding the charge and needs to be binned.
That link is only to download the PCB files from. The other parts needed are the holders which i got from Aliexpress. www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20190124133604&SearchText=4x+18650+Li-ion+Battery+Storage+Plastic+/
its a old server 460w powersupply so nothing great but is rated 33amps 5volts and 32amps 12volts. The key is the 5v rating. The buck converter steps down the 12volts to 5 volts for 8 cell chargers. This was done so i didnt over load the 5volt rail by adding the extra 8 chargers to it. If you have less 5volt amps on your power supply you could use two buck converts and take more power from the 12v rail etc
@@pitbull1969rr Yes you could but you will need a buck converter able to deliver alot of amps at 5volts or a few 10amps ones. If you use a 10amp one dont run more then 7 amps continues etc
Mikes DIY Tesla Powerwall The server power supply( Dell E2700)I have is 12 volts 108A If I use the buck converter from 12V to 5V I will get calculating 5V 45A 225w continues tha amperage will be ok or should lower it to 40A? Thanks for you advice
Mikes DIY Tesla Powerwall yea definitely not, but it's fun working with them, so far all I've done is make some small 12v packs to run some lights in my shed that has no power. Thanks for the videos
or you could use 2 new computer power sources at $11 each 400 watt, that already have fans (that will blow air on the accumulators if oriented well but in reality 18650 don`t get hot, or even warm except if they`re old and bad), and together they will give you much more than the 40 amps 5 volts you need (more like 80 amps combined) ...your project will look much better
Jebus man, you must have too much money. I recently upgraded my charger from 12 cell to 30 cell, total cost was about £30, Its reasonably tidy, but who cares about looks, only I will see my charger and id bet that only you will see yours. Video of mine coming in next few days, but i don't do powerwall forums or fakebook because i live by my own rules ;)
hahah fair enough, hahah fakebook. Ya thats a good price. Mine was around that cost. 30 is a good upgrade. Your right it doesnt need to be tidy it just needs to work. However given this was done for the video i tried to do the best i could without going over the top.
I just need more cells now, 30 chargers and 16 dischargers gets through them pretty quickly. Thanks for all vids and comments, but please don't be so much a stranger. :)
The video at Philipp Seidel's blog (linked in the description) suggests putting some solder on those long lanes that pass the power across the board so they can handle the additional current from daisy chaining the boards together. I was thinking about adding a wire before I saw his video.
He also uses a diode to protect the TP4056 from a reversed battery. The diode would short circuit the battery when it's backwards and blow the fuse.
PS - I just finished binge watching your whole series. Filled 3 or 4 sheets of paper with notes. Great info here!!
Excellent perfect job very good nicely made thanks god bless you my son
Great vid bud
Think you can spec thicker tracks when ordering the boards. I would have tinned the board before adding the tp4056 boards.
Thanks for the vid 👍
Thanks for the feedback i might do that next time
I just build an 18650 charger for 48 cells - using a 30A 220v to12v dc converter and three 9A voltage regulators - all set to 4.2V. I made the cells in parallel - working just fine. I only need to check from time to time to see if one cell gets warm - that cell I take out. I only run in the daytime when I am around - this way I can charge 48 cells per day.:)
If I recall the Tp-4056 boards come default to charge at 300ma. You have to change a resistor to do 1amp. I am using a 30amp supply for my 36 slot charging unit no problem. Looking good, good luck
Thanks :) hmm these boards i used seem to be set at 1amp. As you can see 34:42min its pulling over 7 amps on 8 cells that i put the meter on.
I just wanted to suggest you check out a metal conductivity table. I've done so much electrical work that I instantly had regrets about after reading one. One fact that blew my mind was that although brass contains 70% copper, it only conducts 28% as well. Tin is 15% and lead 7%. I definitely prefer leaded solder, and due to that, use silver-bearing lead solder instead. There's also silver-bearing lead-free that also contains copper, and likely conducts the best that I hope to try in the near future
i make similar setup but i used a 2 units PC power supply and run very very nice
What current do you have the charger charging the battery at? There is one resistor value that sets the current to charge the battery.
The tp4056 is designed to blow itself up in any fault case
You should defiantly add more solder in the holes of the TP4056 boards and You probably already know this but you should try Solder Flux it works so much better than that Solder Paste you're using. That stuff is more for SMD applications, but it does work theirs better stuff out there that will make soldering so much easier. Flux is the best thing ever created for soldering lol I can live with out it. Try the Pure Rosin ones there my favorite, you can use it to make liquid flux to. Good video tho. Gotta build one someday i'll probably add some header pins on the boards. my 10 Lii-500s aren't enough lol
Nice job, but ... what are you using those 18650 cells for when they are full charged?
Very nice and neat, good job done Mick
Thanks man
Why don't you just solder a solid copper wire of the appropriate size that you would need along the traces? That would go a long way in the protection of your boards. Don't you think?
I thought this, i use tinned desoldering ribbon, stays nice and flat, or nickel plate that is used for connecting batteries.
your original charging board was tacky when you built it lol. The negative track is actually the entire front of the board not just where the silver is, that is just to add some solder on it, you can put a solid piece of wire to beef it up! maybe I should redesign that board for the LM2596 CC buck boards I am using!
tacky yes haha Hmm i never tested if that ground track is connected to the entire front but i will have a look.
With you experience using these little tp4056's have you come across a hack or anyway to adjust the voltage that its seeking or sensing?
What do you do with all the single cells when that are charged? What are they used for?
Wouldn't the board be protected if you use a diode instead of a fuse?
Has there been an updated version of this? Looking for a tp4056 pcb design that is within 100mm x 100mm
Wouldn't it been cheaper just to install a small transformer instead of the boost converter?
Great idea
you cant use more modules in series ? and x times the voltage? lets 8 in series with 24v dc?
Great project, what do you do to capacity test the cells?
capacity ? its charger. Not tester.
@@happer2009 he's asking what do you use to capacity test them. The TP4056's brother is the ZL2B3. Not quite as cheap but on Aliexpress or Banggood you can get them for like $4 each. They aren't the most accurate testers but they are fairly consistent between each other. My testing has shown they are rather optimistic giving me results several percent higher than what they should be. The ZL2B3 isn't constant current, its a simple resistor pulling down the cell so it can be hard to compare to other testers
Why don't you just install some protection diodes at each battery charging board so you don't blow out your boards? These diodes are very cheap and not a lot of work soldering them in place. Just asking mate.
protection diodes or using a diode will create a voltage drop, Around 0.6v off the top of my head. Which will cause a problem with the charging circuit unless you somehow created an offset in the chip. Remember the tp4056 is wanting to hit 4.2v CV and then do CC after that, so if the voltage wasn't able to get up to 4.2 held back by the 0.6v voltage drop it will cause the cells to never fully charge, That's why I didn't use diodes as it messes with the tp4056.
@Mike how about the total costs of it ?
I"m wondering if cheaper would turnout to buy 10x Liitokalas Li500..
its alot cheaper but does take time to build. 40x tp4056 are alot cheaper than buying li500s
By the way great video just need something like that....👍
buildup the tracks with solder as aussie dave did
I see you don’t have enough battery’s yet. How many will it take to finish? I haven’t really tried to get more batteries. But would love more.
Never enough, not sure this project will ever finish
Will this charge, discharge and re-charge automatically?
No this will just charge the batteries fully
Hi Mike.
I have tried charging 20 x 18650 all in parallel with a 4.2V 10A current limited power supply and I have had no problems. I use Rui Deng DPS5015 as the power source and stop charge when current drops below 100mA. May pay to put 0.5 ohm in series with each cell to limit surge currents when different voltage cell are paralleled but so far I have had no problems.
I was wondering why you have to use 4056's?
Hi Peter,
The TP4056 gives each cell 1amp and each is independent of each other, Meaning if some finish first they can be removed and another put in to charge. With the idea of 20x in parallel you will have to wait until all have finished or below 100mA. The problem with the 100mA cut off is if the cells overheat during charge and some never charge because of this the current will not fall below your cut off being 100mA. 8 out of the 20 Sanyo cells not charging should do that. Charging cells that are all around the same start voltage and are known to be good are fine to charge in your 20x parallel but I wouldn't be putting cells pulled straight out of laptops in there, Some will not charge, some might be dead shorts, some might be 1v with other at high 3s, My advice would be always charge unknown cells independently and connected to there own charge circuit for the first time. This also helps with speed because as soon as one finishes another can be put in, as they all have different capactity some will finish much sooner than others and some will never finish and are binned.
Thanks for your reply. Your system is more effective in weeding out the really bad cells. My system is very simple so only good for non leaky cells.
I have only tested about 500 cells and generally dump any cell with
How do you check if the cells are heaters? Do you only charge when you are around to manually monitor the temps?
if they are heaters then 1 they will be hot and two all other calls would have finished charging but the bad ones will not be so they get binned. Some do get hot. No i dont only charge when i am around. Its runs 24/7 charging cells. Never had a problem after doing 1000s could be getting close to over 10,000 these days
Thank you for your response! Not challenging your methods, this build is awesome, just trying to understand. If they charge overnight, wouldnt some heaters have the chance to cool back down before you got to them in the mornings? And surely all 40 cells aren't beginning at the same voltage so wouldnt varying charging times be normal and not simply a sign of an internal short?
No the cells don't cool down, The cell gets hot and this causes the charging circuit to keep charging the cell and never cuts off. The charging circuit will only cut off to that intervidual cell when the cell drops to under 100mA in current draw. Given the cell is using more current than that in the form of heat then it never turns off and the cell stays hot no matter how much time pases. So its very easy to tell. Also if a cell did reach charge and then drop down in voltage to under 4v then it will start charging again this cell might not get hot but when all other cells are charged bar 1 or 2 then its also easly to tell that cell is not holding the charge and needs to be binned.
Where did you get the parts from? This link you provided seems to be kinda clunky to navigate. Very frustrating. Good vid!
That link is only to download the PCB files from. The other parts needed are the holders which i got from Aliexpress. www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20190124133604&SearchText=4x+18650+Li-ion+Battery+Storage+Plastic+/
what is the Amperage of your power supply? does the Buck converter step down your voltage and amperage to 5V , 8A ? from? what is the input?
its a old server 460w powersupply so nothing great but is rated 33amps 5volts and 32amps 12volts. The key is the 5v rating. The buck converter steps down the 12volts to 5 volts for 8 cell chargers. This was done so i didnt over load the 5volt rail by adding the extra 8 chargers to it. If you have less 5volt amps on your power supply you could use two buck converts and take more power from the 12v rail etc
I have a Dell server power switching supply 12v 108A 12Vsb 4.5A
Can I used with the buck converter to power the whole module?
@@pitbull1969rr Yes you could but you will need a buck converter able to deliver alot of amps at 5volts or a few 10amps ones. If you use a 10amp one dont run more then 7 amps continues etc
How would you tell the capacity of every single battery in mAh ? Is that single battery tell you the capacity?
Mikes DIY Tesla Powerwall
The server power supply( Dell E2700)I have is 12 volts 108A
If I use the buck converter from 12V to 5V I will get calculating
5V 45A 225w continues tha amperage will be ok or should lower it to 40A?
Thanks for you advice
Cool as New Zealander 👍
:)
Hello Mike
Nice PCB board's where can I get them???
And can you tell me about price/PCB...
Just updated video details above :)
Thx for the pcb
I need this
how can you tell the Battery capacity with this charger?
You can't, this is just for charging, capacity testing will need something else, like an opus 3100 or similar.
Can someone just design a pcb where you could solder tue components directly?
Hey mate I'm in new Zealand too, where do you get your batteries from? I source most of mine from the recycling centre but is pretty slow.
oh cool, There are a few of us on here from nz :) Ya i get most of mine from asking around. Nothing is free tho.
Mikes DIY Tesla Powerwall yea definitely not, but it's fun working with them, so far all I've done is make some small 12v packs to run some lights in my shed that has no power. Thanks for the videos
No probs :) Cool hope to see some videos on your setup
or you could use 2 new computer power sources at $11 each 400 watt, that already have fans (that will blow air on the accumulators if oriented well but in reality 18650 don`t get hot, or even warm except if they`re old and bad), and together they will give you much more than the 40 amps 5 volts you need (more like 80 amps combined) ...your project will look much better
Just went to the website that sells these parts. Everything is in German! lmao!
Good translate? You really just need the download link to the files dont worry about reading the whole page, just download and upload to jlbpcb
Cool!
Jebus man, you must have too much money. I recently upgraded my charger from 12 cell to 30 cell, total cost was about £30, Its reasonably tidy, but who cares about looks, only I will see my charger and id bet that only you will see yours.
Video of mine coming in next few days, but i don't do powerwall forums or fakebook because i live by my own rules ;)
hahah fair enough, hahah fakebook. Ya thats a good price. Mine was around that cost. 30 is a good upgrade. Your right it doesnt need to be tidy it just needs to work. However given this was done for the video i tried to do the best i could without going over the top.
I just need more cells now, 30 chargers and 16 dischargers gets through them pretty quickly.
Thanks for all vids and comments, but please don't be so much a stranger. :)
.
Another boring 1A charger, how slow!