Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB- Mtbr.com
Page 1 of 2 12 LastLast
Results 1 to 200 of 232
  1. #1
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810

    Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB

    I've already wrote about those in another battery thread. I've got them to test real capacity from Gearbest. They choose to release coupon code to lower the price:

    Name:  1430162954471-P-2572266.jpg
Views: 3997
Size:  14.7 KB
    2600mAh 8.4V 4 x 18650 Portable Rechargeable Water - resistant Battery Pack-12.07 and Free Shipping| GearBest.com
    $10.59 with Coupon: 4Bpack

    Name:  1430162732886-P-2572199.jpg
Views: 3891
Size:  14.4 KB
    8.4V 3900mAh 6 x 18650 Portable Rechargeable Water - resistant Battery Pack-15.95 and Free Shipping| GearBest.com
    $13.69 with Coupon: 6BPack

    Funny I have realized today those are copies of Magishine MJ-6038 and MJ-6036S

    OK, let start. Both are made from plastic and has standard Magischine type connector 5.5x2.1mm. Cable is flat and mid sized. Both packs has rubber straps to be mounted on the bicycle frame. Alli in all the look is good.

    Teardown has been done with screwdriver prying edge between the top and the body. To get to that point you can move ruber band to the bottom. Inside was welded pack with foam on every side to prevent rattling:


    It is good there is zip tye to prevent pulling cable out:


    The cells are unknow (no markings) with small protection circuit:


    Protection circuit:


    After full charge and one day rest I discharged 2S2P pack with iCharger 106B+ at 2A down to 5.6V and I've measured only 2198mAh. Since the pack is very likely unbalanced I would give it bit more, but not much.
    Here is the discharge graph:


    Similar to the previous one the 6 cell pack has only 3273mAh. The voltage falls down so fast that protection trips in before charger realize it reached 5.6V. Discharge graph:


    Discharge graph of 4cell pack showing each pair of cells. You can see one pair is fairly good (ie. it has smooth curve) and the other is having some trouble.



    I've done the same for 6 cell pack..... After charging pack it is noticeably unbalanced with 4.24V and 4.13V for each set of 3cels.
    6cell pack is also having some issues and quite some difference between 2 sets in series.


    Gearbest has corrected their description of the capacity based on my test. Not exactly the same figures, but close enough.
    ----------------------------------------------------------------------------------------


    DIY PART

    Q: why would one buy these quite limited capacity packs?

    A: For anyone who have some soldering skills and can do some modifications those packs might be very usable. Imagine you would have 10000mAh battery pack for a very long rides or for very power hungry lights.
    How much would that cost? See it at Action Led Lights or Magicshine. How about half of that price and having balancing feature included in the battery pack. Sounds to good to be true? I don't think so.

    What parts would you need for self balancing 6cell pack:

    1. Disassembled batery pack from the top of this post
    2. Two pieces of Original Xiaomi Pocket 10000mAh Mobile Power Bank
    3. Protective/Balancing circuit from Hunk Lee: 4A or 7A version
    4. Soldering station and some tools, Kapton tape would be nice too
    5. Optionaly some older 18650 cells to make Xiaomi power banks usable

    How to do it? Luckily I've got 10000mAh Xiaomi power bank today. I've dissassebled it to see if the cells are the ones and YES they ARE Panasonic (Sanyo) NCR18650BF the 3350mAh cells. Dissassembling is described HERE.

    So you have to use two power banks to get two sets of cells and make an 2S3P pack. Since they are already welded you need to do just few additional connections.
    Will post pictures later on...

    Protection/balancing circuit should be properly soldered and nicely put into the shell. You can imagine all those dirty details to do.

    Actualy I haven't done that mod yet, but I've used an old laptop pack and salvaged 6 Panasonic 2900 cells from 3S3P pack where 3 cells are already welded, just like with power bank. I've soldered together two sets and put into the pack without protection. It works nicely and there is enough room for protection circuit. I'm not shure it can fit with smaller 2S2P pack. It might fit verticaly, but it remains to be seen....
    Old Panasonic cells:




    I will have to wait for another power bank and protection circuit to actualy finish this DIY. But idea is doable and you might want to repeat it.

    Will continue with some correction of the pictures later on or at the end ....

    Edit 9.9.2015:
    I would need to add the PCM doesn't have to be neccessary balancing one. Anyway I would go with the higher current ones as they have less resistance and are more apropriate for the BT70 type of lights.
    Hunk Lee has some - HERE. Take a look detailed description of each where you can see Maximal continuous Discharging current. Be aware of Over charge detection voltage, some are for 4.30V cells - like this.
    Last edited by ledoman; 09-10-2015 at 02:43 AM.

  2. #2
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Added DIY chapter at the end of review. You might get some good idea..... Tell me what you think.

  3. #3
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Thanks for the information!

    Now it's time for GearBest enthusiasts (GJHS? tigris99?) advise them to sell these shells as a DIY-kits: not glued and with no garbage inside...

  4. #4
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Archie, yes that could be nice so we can finaly build high quality battery packs. It would be fine to find similar cheap and quality 2 cell power bank with cells welded.

  5. #5
    mtbr member
    Reputation:
    Join Date
    Jul 2015
    Posts
    16
    Hah I've actually received mine just today too! (Ordered two through allbuy.com, took about 35 days >_> to arrive, are original)

  6. #6
    mtbr member
    Reputation:
    Join Date
    Nov 2012
    Posts
    1,043
    Very good idea. So for around $50 you can get make a 6-cell 10Ah panasonic battery. That make is possible to run those 4 amp lightheads (bt70 turbo) for over 2.5 hours.


    My only concern is how waterproof the pack is after you put it back together.

  7. #7
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Do those Hunk Lee circuits actually provide balancing? How can they if they simply connect to positive and negative of the assembled pack?

    Awaiting pics. Tempted to pick up two of those power banks!

    -Garry

  8. #8
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    varider, I think you even not need to glue it back if you don't want. You can use just some duct tape. Or if glued maybe add some sealant and you are done. I would do it also at the bottom if you want perfectly waterproofed pack.
    Last edited by ledoman; 07-21-2015 at 10:54 PM.

  9. #9
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Garry, in specifications says: "Balance current for single cell 70mA" and you can see it is differently constructed if you compare it with any other PCM.

    Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB-s-l1000.jpg

    You won't see the pictures of complete DIY from my side very soon since it takes much more time to get anything from China than you do in US. This PCM has just been sent. So 3 weeks at best until it arrives.
    Last edited by ledoman; 07-21-2015 at 11:00 PM.

  10. #10
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    If Original 16000mAh Xiaomi Mobile Power Bank Aluminium Alloy Material for Samsung S6 HTC ONE M9 Tablet PC Smart Phone etc-23.99 and Free Shipping| GearBest.com would be cheaper then you could get 2S2P + 2S3P cells out of two devices. There are LG 3200 cells installed. So we could make nice 6400 and 9600mAh packs.

  11. #11
    mtbr member
    Reputation: Cat-man-do's Avatar
    Join Date
    May 2004
    Posts
    6,080
    Quote Originally Posted by ledoman View Post
    If Original 16000mAh Xiaomi Mobile Power Bank Aluminium Alloy Material for Samsung S6 HTC ONE M9 Tablet PC Smart Phone etc-23.99 and Free Shipping| GearBest.com would be cheaper then you could get 2S2P + 2S3P cells out of two devices. There are LG 3200 cells installed. So we could make nice 6400 and 9600mAh packs.
    OR...if you wanted you could just buy a Hunk Lee battery pack of your choice and fit it inside the case. Heck, since they're both coming from China why not?

    Anyway, thanks for the review ledo. Didn't surprise me that the cells in these things were crap. ...And speaking of crap; I would of crapped myself if you'd of said these were built with Samsung, Panasonic/Sanyo or BAK cells.

  12. #12
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Just finished discharge test of a 3P cells in the Xiaomi Power bank and got 9690mAh down to 2.8V at 3A load (1A per cell). Since the cells are declared down to 2.5V I can conclude they are true 3350mAh (on average) cells. I think it's hardly you get better price elsewhere for those cells and power bank shell as a bonus.

  13. #13
    mtbr member
    Reputation: Cat-man-do's Avatar
    Join Date
    May 2004
    Posts
    6,080
    Quote Originally Posted by ledoman View Post
    Just finished discharge test of a 3P cells in the Xiaomi Power bank and got 9690mAh down to 2.8V at 3A load (1A per cell). Since the cells are declared down to 2.5V I can conclude they are true 3350mAh (on average) cells. I think it's hardly you get better price elsewhere for those cells and power bank shell as a bonus.
    Okay, I'm a little confused. This power bank, this is something used to charge cell phones right? Usually these kinds of power banks are limited to a certain current output. I've never seen one more than 2A. I looked at the output rating for the Xiaomi Power bank 16000 and supposedly it is rated to 2.1A maximum current output.

    Anyway, LG is a good make of cell. Should make a fine battery pack. Yeah, a Hunk Lee battery might be a little more money but then again it's all wired up with a protection board already in place. All you would have to do is cut and splice the power leads to the connector on the case, put some foam around the cells to keep them from rattling and you should be set. ( not to mention you get to choose the brand cells you want it built with ). I'm not knocking your DIY idea just pointing out an easier option for people who wouldn't want to wire up a board and do all the soldering.

  14. #14
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Cat, of course I've tested bare cells not the power bank circuit since we are interested in the cells, right? Device was dissassembeld to get to the cells. I've connected them with aligator clips directly to my iCharger 106B+ which was set to 3A discharge and Li-lo limit at 2.8V. So again I can confirm in the 10000mAh power bank there are geniue Panasonic NCR18650BF cells.

    LG cells are supposed to be in a 16000mAh power bank.

    I agree Hunk's packs are more or less ready to go, but the main goal here (besides cells itselfs) is the balancing circuit. So no more worries the cells are out of balance. Hunk says he doesn't use balancing type of circuits in his packs. He is using only protection features.

  15. #15
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    EDIT - Sorry, ledoman beat me to it.

    Cat, ledoman has removed the cells from the power bank and tested them on a hobby charger. The power bank circuitry has no effect.

    I still don't get how you're ending up with balancing with only two battery connections. Hopefully it will be clearer once you finish the DIY build posts.

    -Garry
    "My Bike Lights" Thread on BLF teardowns, measurements, and beamshots. Moving my photos, PM or post up if you can't see them.

  16. #16
    mtbr member
    Reputation:
    Join Date
    Jul 2015
    Posts
    9

  17. #17
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Garry, where did you get "only two battery connections"? I see ordinary 5 connection points, 2 external (P+/-) and 3 internal (B+/-, MB).

    Name:  $_14.JPG
Views: 3530
Size:  1.0 KB

    See all the pictures from the link on the picture above.

  18. #18
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    I'm still lost. I don't see the 3 internal B+/B- connections. I'm looking at these two connections:



    -Garry
    "My Bike Lights" Thread on BLF teardowns, measurements, and beamshots. Moving my photos, PM or post up if you can't see them.

  19. #19
    Drinkin' the 29er KoolAid
    Reputation: kwarwick's Avatar
    Join Date
    Jun 2004
    Posts
    1,495
    Typically with those protection boards, B+ and B- is where you hook up the + and - ends of a 2S pack and MB would be the center tap of the pack. P+ and P- is for the load, so that's for the cable that goes off to the light.

  20. #20
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Ok, so having that "MB" connected between the two sets of parallel cells (two sets of parallel cells then connected in series) provides the balancing? So it's not providing balancing to each individual cell, but to each parallel bank of cells? This makes sense to me.

    -Garry
    "My Bike Lights" Thread on BLF teardowns, measurements, and beamshots. Moving my photos, PM or post up if you can't see them.

  21. #21
    Drinkin' the 29er KoolAid
    Reputation: kwarwick's Avatar
    Join Date
    Jun 2004
    Posts
    1,495
    Quote Originally Posted by garrybunk View Post
    Ok, so having that "MB" connected between the two sets of parallel cells (two sets of parallel cells then connected in series) provides the balancing? So it's not providing balancing to each individual cell, but to each parallel bank of cells? This makes sense to me.

    -Garry
    Assuming the board actually does balancing, that would be correct. When cells are wired in parallel (2P, 3P, etc) they are pretty much forced to be in balance so all the board would be able to balance would be the voltage of the two halves of the 2S pack.

  22. #22
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Gotcha! Thanks! I didn't know what "MB" meant where it was to be connected. Sorry ledoman for my misunderstanding. Now your post makes sense! Duh!

    BTW - Even though Gearbest's price went up, Banggood still has the Xioami power bank for $12.66.

    -Garry
    "My Bike Lights" Thread on BLF teardowns, measurements, and beamshots. Moving my photos, PM or post up if you can't see them.

  23. #23
    Drinkin' the 29er KoolAid
    Reputation: kwarwick's Avatar
    Join Date
    Jun 2004
    Posts
    1,495
    Quote Originally Posted by garrybunk View Post
    BTW - Even though Gearbest's price went up, Banggood still has the Xioami power bank for $12.66.

    -Garry
    Yeah, that's a great price for high quality power banks.

    I've been boycotting GearBest, due to the poor service I've been receiving from them lately, so I ordered two of those from BangGood the other day and got my shipping notification in less than 24 hours!

  24. #24
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Garry, FINALY! I thought you would already know those things since you have tear down your battery packs where the connections are exactly the same (or at least should be in your 3xled example). There is only missing balancing feature at all others PCMs I've seen so far.

    This balancing PCM or any other protection board should works regardles how many cells you have in parallel. It is 2S circuit.

    BTW. Since you are engineer I'm shure you know system of connected pots where the fluid is equaly high regadless of size of the pots. The same is with cells in parallel.
    When having cells in series you need something to measure and regulate the voltage of each set in series. This is where Hunks PCM is taking the place. It gives small "pipe" up to 70mA so the energy flows in any direction from higher to lower. Again system of connected pots but with regulated pipe this time.

    I hope you get it now in detail.

  25. #25
    mtbr member
    Reputation:
    Join Date
    Jul 2015
    Posts
    9
    you think i can use that batteries from the xiaomi power bank charging with a nitecore i4 and put in this? Solarstorm New 2S2P 8.4V 2000mAh Dual Water - resistant 4 x 18650 External Battery Pack for Bicycle Light Digital Device-10.69 and Free Shipping| GearBest.com

  26. #26
    mtbr member
    Reputation: Cat-man-do's Avatar
    Join Date
    May 2004
    Posts
    6,080
    Quote Originally Posted by ledoman View Post
    Cat, of course I've tested bare cells not the power bank circuit since we are interested in the cells, right? Device was dissassembeld to get to the cells. I've connected them with aligator clips directly to my iCharger 106B+ which was set to 3A discharge and Li-lo limit at 2.8V. So again I can confirm in the 10000mAh power bank there are geniue Panasonic NCR18650BF cells.

    LG cells are supposed to be in a 16000mAh power bank.

    I agree Hunk's packs are more or less ready to go, but the main goal here (besides cells itselfs) is the balancing circuit. So no more worries the cells are out of balance. Hunk says he doesn't use balancing type of circuits in his packs. He is using only protection features.
    Okay, sorry about that. Your post made it sound like you were using the power bank, my bad.

    About the Hunk Lee batteries; Yeah I guess the PCM circuit board you are using sounds better. I think most batteries only self balance the parallel sets. What you're using balances the cells separately, right?

    Anyway, I'm assuming the cells in the power bank don't have tabs. How are you attaching the wires to the cells. Do you have one of those spot welders?

  27. #27
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Cat, cells in the power bank are very nicely welded. You should look what I've posted in the OP "Dissassembling is described HERE." You only need to make middle cross connection. I'm using 0.1mm copper sheet stripes and solder them on welded stripes. Very simple, but you need hi watage soldering tool.

    I think Hunks PCM linked in OP is balancing between sets in series all the time. I've asked Hunk, but he didn't give me direct answer. He was talking about benefits of balancing not how it works with his PCM. So I need to wait to get it and check. Basicaly I think it works like I've described to the Gary above with the "linked pots" ie. PCM is 70mA pipe between two sets in series. All cells in parallel (ie each set) are always self balanced.

  28. #28
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    I went ahead and ordered two of the power banks from Banggood (but missed where to add tracking). I have 2 Hunk Lee PCM circuits, but I'm pretty sure mine were not balancing type.

    -Garry

  29. #29
    Drinkin' the 29er KoolAid
    Reputation: kwarwick's Avatar
    Join Date
    Jun 2004
    Posts
    1,495
    Quote Originally Posted by garrybunk View Post
    I went ahead and ordered two of the power banks from Banggood (but missed where to add tracking). I have 2 Hunk Lee PCM circuits, but I'm pretty sure mine were not balancing type.

    -Garry
    If you ordered two, you likely got free tracking as I did... seems the threshold for free tracking is $25.00.

  30. #30
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Garry, it is OK with other PCM too. Since you have hobby charger there is chance to make the balancing leads in the pack. If you are going to use hard case from the OP, there is plenty room to have balancing connector hidden in the case. If you don't glue it back (just using duct tape) you can open it from time to time and rebalance using hobby charger. You can use THIS or better yet THIS from ebay.

    BTW I've ordered 7A balancing PCM so it can pass a lot of power and it has quite low resistance of just 15mOhms. So the voltage drop over this PCM would be very small.

  31. #31
    mtbr member
    Reputation:
    Join Date
    Jul 2015
    Posts
    9

  32. #32
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    I do have some of those connectors but I need help wiring it up. Where does each color wire go?

    -Garry

  33. #33
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Kw - my order status page shows "untracked" unless that's just referring to current tracking status and not the shipping type. I think I saw "Regular Airmail" as the shipping.

    EDIT- my total was $24.35 due to using discouht points (and code "blf" didn't work on this type of item).

    -Garry

  34. #34
    Drinkin' the 29er KoolAid
    Reputation: kwarwick's Avatar
    Join Date
    Jun 2004
    Posts
    1,495
    Quote Originally Posted by garrybunk View Post
    Kw - my order status page shows "untracked" unless that's just referring to current tracking status and not the shipping type. I think I saw "Regular Airmail" as the shipping.

    EDIT- my total was $24.35 due to using discouht points (and code "blf" didn't work on this type of item).

    -Garry
    If it was less than $25 you would have had to pay something like $1.30 extra for tracked shipping. I used some discount points too, but not enough to bring it under $25.

  35. #35
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by garrybunk View Post
    I do have some of those connectors but I need help wiring it up. Where does each color wire go?
    -Garry
    Garry it's simple. Middle goes to the MB, red to the B+ and black to the B-. You don't need to solder it to the PCM directly, but anywhere where those points are connected. You did not specify which connectors exactly you have. I hope the ones with red and black wires only, cause the other one is bit tricky with yellow where black should be.

    Or you might try it yourself to use Google, don't you? https://www.google.si/#q=2s+balancing+wiring

    PS. This picture is showing both the red/black only and tricky yellow wire if you have those type of connectors Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB-lipowiring-1.jpg

    And keep in mind cells in the picture above represents sets of cells in parallel. It can be one to many. In our case we have 3 in parallel.

  36. #36
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    kw - I got shipping notice this morning already! I did get a tracking number too, although I had to hunt around in my account viewing order details to find it (wasn't sent in the email). This is only my second order with Banggood.

    ledoman, thanks for posting those pics! Not sure exactly which wires I have, but yellow sounds familiar (so the yellow is just treated as if it's another black wire?). So I connect the "normal" charging leads (i.e. the 5.5/2.1mm DC Jack where the wall plug would connect to charge) in addition to the balance charging connector? (Well of course I would connect that lead to power a light head, but I'm only looking at charging purposes right now.) And I can charge the pack using only the DC Jack lead if so desired (of course without cell balancing during that type of charging)?

    I have no background with RC stuff, so this is all new to me.

    -Garry
    "My Bike Lights" Thread on BLF teardowns, measurements, and beamshots. Moving my photos, PM or post up if you can't see them.

  37. #37
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Garry, when you charge with hobby charger and want balancing feature you need to connect main pack leads (P+ and P-) to the two 4mm jacks on the hobby charger main port (you would need adapter from 5.5x2.1mm connector to 4mm jacks) AND balancing (JST) connector to the balancing port - so esencialy all 5 connections from the PCM (P+ and P- to main port, B+, B- MB to balancing port). Over the balancing port charger measures voltages and does low current discharging in order to get sets balanced.

    When charging without balancing you can use your ordinary charger or just main ports from hobby charger.

    Just browse around and I'm shure you'll find your answers.

  38. #38
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Thanks. Confirms what I thought is correct. Now I just need to source some shrink wrap for a 2S2P pack to try & keep it pretty. Any ideas?

    -Garry

  39. #39
    mtbr member
    Reputation:
    Join Date
    Jul 2015
    Posts
    16
    You can also just charge the cells in parallel wiht 4.2V each instead of charging the entire pack with 8.4V.

    Am I correct? That way every time you recharge the pack it'll be rebalanced and all cells will be charged to the same voltage.


    Edit: People without a good charger can probably just wire up the parallel cells to the xiaomi powerbank and charge it through that.

  40. #40
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Well first try it unwrapped, just use tape to fix everything in place, then you probably know why they invented duct tape ;-)

  41. #41
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by cor35vet View Post
    You can also just charge the cells in parallel wiht 4.2V each instead of charging the entire pack with 8.4V.

    Am I correct? That way every time you recharge the pack it'll be rebalanced and all cells will be charged to the same voltage.


    Edit: People without a good charger can probably just wire up the parallel cells to the xiaomi powerbank and charge it through that.
    Both is true but impractical with sealed packs and it's time consuming. Balancing wires are also to thin. And when sets of cells became internaly different, they might not charge to the same voltage level.

  42. #42
    mtbr member
    Reputation:
    Join Date
    Jul 2015
    Posts
    16
    I think that most people will be making their own packs out of these xiaomi power banks, so instead of exposing only two connections, just expose the middle connection too and you're done.

    I'll see with what kind of pack I'll be coming up with when my cells arrive but I'm pretty sure that I will be charging them in parallel so I can perfectly charge all of them to the same voltage level (and not spend too much money :P)

  43. #43
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    As it suits you. If it's OK to you I see no big problem. But with one 4.2V charger you'll double charging time and with 10000mAh pack it can be very looooong. Be aware mutibay chargers use common negative pole, so you can't just connect two bays to each set as like they would be independed.

    Anyway using balancing PCM the problems above would go away.

  44. #44
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Other thing is unless something new has come to pass, can't charge series welded cells at their individual cell voltage in parallel. With a proper balance charger you can tap each pair of cells, balance at 8.4v and balancer will do what it can to even the pair out. Its not perfect compared to separating cells and charging in parallel but it works alot better than not at all. And is only option I know of with welded packs. You also have to cut the tabs to separate each series set from the rest. Not worth the hassle imo except to do the first time your building the pack.



    However the power bank is already in parallel so they will self balance if you charge near full (research says around 80% is best) and let sit for a day or so. I usually do it a couple times. Charge around 80%, let sit for 24hrs, charge again at a low current till done. Its automatic properties of the batteries wired in parallel to self balance. Assemble the pack correctly, add a balance lead at the bottom that will read both sides (which should be connected at pack assembly anyway) and balance charge the pack.

    Its cheap packs that run into balance issues because cells are used at random so no where close to matched, then no balance attempt is made prior to building the pack. Basically pack is doomed to be crap before its even put into the plastic case.

  45. #45
    mtbr member
    Reputation:
    Join Date
    Jul 2015
    Posts
    16
    I am not sure how well this would work since the entire pack would be on 4.8V anyways, but since there's a balancing connector in the middle, it shouldn't matter, should it?

    Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB-image.png

  46. #46
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Well your whole lay out is series trying to use dual chargers. Pack is still 8.4v and i dont even see that set up working without frying something. Charges arent meant to be ran in series with each other. Way too much work for what a single balance charger is going to do anyway running at 8.4. Balance chargers are designed and programmed to read the cells and increase or decrease output via main leads and balance leads to balance the pack.



    Anyway think im buying 2 of then to build a 3s2p pack out of lol.

  47. #47
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    I have the balance leads with the yellow wire:



    My PCM circuits also have a "BM" connection, so does that mean these are balance charging pcm's?



    -Garry
    "My Bike Lights" Thread on BLF teardowns, measurements, and beamshots. Moving my photos, PM or post up if you can't see them.

  48. #48
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Garry, the "bm" is the center contact for protection (and can solder the balance plug mid wire to it). Thnk like the wire we rig from the bottom of a solarstorm case up the center into the protection circuit. Protection requires a + in, - in, and "bridge" in as I would call it. The bridge is connection between cells that makes them the series circuit. packs that are 2p 3p on up use a single extra bridge between all the series bridges connecting them together into one. Wish I had a draw program that worked lol.

    What you guys are trying to do if you have a balance connector is becoming OVERLY COMPLICATED lol. At least IMO.

    2s3p using the power bank cells prewelded together:

    1. charge each bank to 50% or more, let sit over night so cells are for sure balanced out. IF YOU WANT PERFECTION: use a jumper wire to make each 3 cell parallel set into one big six cell parallel set. THEN EVERY CELL WILL BE PERFECTLY BALANCED. After letting sit to balance, check voltage, may want to charge up a bit more and let sit again for a couple hours. PERFECTLY BALANCED 6 CELLS. But this wont effect discharge curves of each cell. They have to be properly matched or they will fall out of balance. Which none of us are going to do, continue reading:

    2. get some of the solder strips (same thing as what they use to weld these packs)

    3. put the 2 sets of cells on end, one set with the negative side on the table (use a wood or plastic table PLEASE, I know we all know this but posted just in case) other set with positive end down on the table. Wrap a couple wraps of electrical tape around the 2 sets to hold together

    4. take a couple solder strips (or 3 if you want) and bridge the tabs facing up, connecting the 2 sets of cells together positive to negative

    5. another strip where you want it based on where youll be mounting PCB as this will go to the center contact point for the protection circuit to read. Attach CENTER BALANCE LEAD WIRE to any point of this whether its bottom bridges of pack or directly to the tab you just attached to the pcb.

    6. flip pack over, run tabs or wires from each the positive and negative to proper locations on pcb. Solder matching balance plug wires to their locations on the pack tabs or where they are soldered to the pcb

    7. connect battery plug wire your using to output side of pcb, wrap pack however you plan to and charge with appropriate balance charger.

    DONE.

    PS I used to build LIPO packs in my RC days

  49. #49
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Garry,

    I wish you have the picture of the other side of PCM too. Anyway I'm shure you have 2 cheap protection circuits probably working up to 2A (my prediction since I can't see the other side). You might try to compare to those at eBay.
    And no, NO WAY it has balancing feature. There is Hunks Lee PCM only that I'm aware of having balancing feature. NO OTHER has it!

    Your balance leads are the worse one, but usable, no problem. It is just the other ones I've linked are with bit thicker silicone wires. You can change yellow and black if you want, but not necessary, you can use wiring diagram I've posted on the picture above. There you can see yellow is on the negative side (B-) where we would normaly use black color.
    So if you don't change them you would wire them up:
    - red to B+
    - black to BM
    - yellow to B-

    P+ and P- goes to the ordinary 5.5x2.1 connector

    By now you should know everything!!!

  50. #50
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by cor35vet View Post
    I am not sure how well this would work since the entire pack would be on 4.8V anyways, but since there's a balancing connector in the middle, it shouldn't matter, should it?

    Click image for larger version. 

Name:	image.png 
Views:	3915 
Size:	18.0 KB 
ID:	1003962
    I'm not shure either. The middle point is that worries me and how electricity will flow. Also there might be different currents on each charger. @Archie is more skilled on that matter. Shurely charging each set one at a time would definitively work.


    Anyway take two 18650 cells and try it out ;-) And as said it's inpractical for ordinary users.

  51. #51
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Quote Originally Posted by garrybunk View Post
    My PCM circuits also have a "BM" connection, so does that mean these are balance charging pcm's?
    It could, but it is more likely the BM connection is simply used to monitor individual cell bank voltage.

    This feature does not balance while charging, but if your pack does go out of balance, for example if one cell is down to 2.5V while the other is at 3.5V your light only knows the total pack voltage is 6V and will keep going. But because the protection circuit knows one set of cells have reached 2.5V, it will cut off to protect your cells from damage. 2.5V is only an example - protection circuits come with many options for cut-off voltage, could be anywhere from 2.3V to 2.8V+ lower cutoff.

    Works the same while charging, if your pack has come out of balance such that one cell bank reaches for example 4.3V (which is too high for standard Li-Ion cells) while the other is only at 4.0V the standard round-plug charger only knows the total voltage is 8.3V so will happily keep supplying current. But because the protection circuit is set up to monitor individual cell banks it will cut off charging to protect the cells. An interesting effect of this is, when this happens the charger will see an open circuit and indicate a green light as if everything was fine and peechy!

    The hunk-lee protection circuits look to have actual balance charging added on to the above features. While charging, if one cell bank is at a higher voltage than the other, it will drain 70mA of the charging current away from that cell bank until voltages equalize. So those look really interesting and I did actually order one.

    But word of caution - if you seal up the pack you won't be able to confirm the balance charging circuit is actually working. If it fails (or never worked in the first place) and cells become out of balance to the point where the overcharging protection kicks in, your round-plug charger will again simply indicate green. I am planning on making one of the hardcase 6-cell packs (Thanks for the tip Ledoman!!) and sealing/gluing it up watertight, and then breaking it open once every season (or every other season..) to check if cells are still balanced. The rest of my packs in rubber softshells will still be charged regularly on the hobby charger with balance leads. I carved out a cavity in the rubber lid where the balance connector lives - when charging, simply pop off the rubber lid (opposite end of cord) and connect balance lead.

  52. #52
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Just got the second Xiaomi Power Bank and I've already dissassembled it. This time I took battery pack away from the circuit. I've used carperter knive and gently pry the welded conections on the circuit.

    To test of further circuit usabilty I took two old parallel welded cells and solder 5.5x2.1mm connector to it. Same I did on the circuit side so now I can connect other cells to it. At first it didn't work regardless there was more than 4.0V coming from the batteries. Then I remebered protecion circuits on our bicycle packs needs to be reset if you make some hocus pocus. So I've just took USB cable and charge it for the moment. This has reset the circuit and it started to work normaly. So as predicted it is usable. Will solder some older cells to it later and give it someone or use it by myself.

    Now the Panasonic cells are being charged with hobby charger to the storage level. When they came they had 3.66V and storage level is 3.85V. After both sets are being charged I'll conect all 6 cells in parallel for some time just to balance themselves. Unfortunately I will have to wait for a balance PCM for at least two weeks. Until then will set them up without protection. I might a balance connector.

    And to add to the Morten writings I think I can test each balancing PCM with two unevenly charged cells. Just need to connect them in the right way and see if they get balanced after some time. Voltage can be monitored with small device like that:
    Name:  m4btmeJ_32pErbziV8IkYeQ.jpg
Views: 1882
Size:  8.7 KB

  53. #53
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Thanks guys! My PCM circuits are Hunk Lee, but just cheap ones - probably 2A.

    I have those voltage monitor/alarms and wondered about incorporating those as well. Would have to be able to disconnect it or else it would constantly drain the pack. Hmm...

    -Garry

  54. #54
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Some pictures of Panasonic cells from the Xiaomi 10000mAh Power Bank. At current price $15.99 still worth of buying.

    Selfbalancing all 6 cells:



    Both packs with circuits:





    Test setup to check reusabilty of the circuit:


  55. #55
    mtbr member
    Reputation: Cat-man-do's Avatar
    Join Date
    May 2004
    Posts
    6,080
    Just curious, how are getting electrical connect with the metal strips across the cells? Do you have one of those battery spot welding set-ups? ( or are you just using more magnets? ) BTW, where do you get those nice strips with the perfectly spaced cut-outs? Those are nice! Anyway, you're giving me some ideas...

  56. #56
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    What im seeing in those pics is everything is how the cells come inside the power banks. Except his magnet set up to jumper the 2 sets of cells.

    Going to order 2 of these soon, 6 cells for $30 with tabs that I can cut and solder back together to make the 3s2p pack I need is freaking cheap lol.

  57. #57
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    I see Banggood raised their price too! Good thing I bought in time.

    -Garry

  58. #58
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Always happens when we post these things here and blf, prices go up instantly.

  59. #59
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by Cat-man-do View Post
    Just curious, how are getting electrical connect with the metal strips across the cells? Do you have one of those battery spot welding set-ups? ( or are you just using more magnets? ) BTW, where do you get those nice strips with the perfectly spaced cut-outs? Those are nice! Anyway, you're giving me some ideas...
    Cat, the first two pictures showing cells in a temporary selfbalancing setup. Front (+) side is conneted by two aligator clips since one of the banks is not cut away from the circuit. Back side is connected with two magnets and copper strip. Once I've bought 20mm wide 0.1mm thick copper sheet. I'm using it on different places (additional heatsinking at KD2 light). Easily cutting with scissors strips can be used to solder all kind of battery setups. It is advised to use high power soldering tool so the battery cells are exposed to the heat less time. Reasoably priced copper sheet can be found HERE. I'm shure many of you would find it usable.

    PS. Will ask May at Gearbest if she can get us some coupon for those power banks. I think you would buy more, don't you?

    Edit: updated link to copper sheet.
    Last edited by ledoman; 09-26-2017 at 03:49 AM.

  60. #60
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    Damn, I am too late, now the price is 22$ for 10KmAh version at BG while Gearbest sales 16KmAh pack for 24$...
    Waiting for 12$ deal on 10KmAh pack.

    P.S. Guys, stop messing around with yellow and red wires you are going to blow something up
    You can check this link to get some more info about how to make balancing leads...

    P.S.S. ledoman, can you give some comments about the quality of that plastic case and rubber bands on 3P2S pack? Is it at least close to Magicshine's case that coms with MJ880?

  61. #61
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Sirius, I think the plastic case is quite similar to the Magicshine. Can't tell exactly as I don't have MS case. I had one for repair more than year ago so I'm writing from my head. The wires are quite thick inside. Might be even 20AWG thick. I was supprised when I saw that. Unfortunately it has no markings.

    The ruber band is nice too. You can dismount it from the pack. But don't know for the winter. I think in winter you need to work with them more gently. Same is probably with the wires. Below some point in freezing condition they might crack. But this is true for many cheapo wires anyway.

    To bad the cells are so low capacity, otherwise it would be very nice pack. And if the existing PCM is sufficient you only need to replace the cells. But I would advise using more powerful PCM and/or balancing leads if you are going to power more demanding lights.

    With coupon the price is better than using Pannovo or Solarstom case. You can have better PCM, almost no parasitic drain and you can make it really waterproof. And of course you can choose between 4 or 6 cell pack.

    And BTW, I've just noticed there is small gap in the cover so you could add some apropriate thin o-ring for better sealing.

    If they would selling them below $10 without cells this might sell even better. Of course it woud be only for the DIY skilled people then.

  62. #62
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    I just want the plastic case and those rubber bands if they are as on magicshine pack because they are very nice and do not slide on the bikes frame, they hold nicely (I had MJ880, sold it, too expensive for my pocket).
    guts would be discarded and NCR cells will be placed inside together with 7A 2S balancing/protection pcb from hank.

    I also have this "state of charge" indicator that I thought I would integrate also! Yes, it is meant for single cell (so up to 4,2V) but if we assume that balancing pcb will do its job, discharging cells in the pack should also be fairly equal so if I connect this battery level monitor to just one side of the pack (3P) I should get acceptably accurate readings.
    How this sounds to you, good idea?

  63. #63
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Yes with balancing PCM this indicatir should work corectly. It is only how you would integrate it and not to loose waterproofing. Maybe if you make some soft transparent window so you can also press the switch.
    Also does it make some parasitic drain or it only consumes power when you press the switch.

  64. #64
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    For indicator leds I thought I would just drill small (1.5-2 mm) holes in the top part of the case, fill them with hot glue and then just smooth out top and bottom part, that should work (also I hate riding when its raining so waterproofness should not be that big problem), for button I have small rubber boots for maglite flashlights and a set of tact switches with different stem lengths bought long time ago...

  65. #65
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    It would be nice if you can find out if your indicator circuit drains battery on it's own. Ie. it has parasitic drain. It is just we would be aware about it. If you can measure the current it would be perfect.

    Otherwise it would be nice to see final result in a case.

  66. #66
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    with my, 20 years old multimeter there is no way to measure parasitic drain, with display resolution of 0.00A anything under 10mA is displayed as 0 and up to 100mA you can only read in steps of 10mA...

  67. #67
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Then connect it to the good battery for the one month until you get parts to put things together. Difference in voltage would tell you if is there or not

  68. #68
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    OK, here we go again, Xiaomi 10K battery pack US$11.99 (not aff link) needed
    for this project, for 1 tracking is not included (about + $1.70) but if you buy 2 pcs you get free tracking.
    I ordered 2
    Edit: price drop to $11.99 <- best price ever
    Last edited by Sirius9; 07-28-2015 at 10:46 AM.

  69. #69
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    The Gearbest is also at good price again - $12.56

  70. #70
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    I like doing business with GB because they provide so many good deals and its easy to comunicate with then, but today they were too slow :P
    Woke up this morning, powered on my computer and first thing I see in my emails is BG promo: "1st Round of XiaoMi Week!" (I am telling you, they read ths forum, BLF as well ) I missed the last deal so I had to be quick this time, first this I did, before going to toilet, was to buy those 2 banks then breakfast and then off to work... :P

  71. #71
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Yup, at the moment GB has the best price. I would suggest anyone interested to buy now. I don't think the price would be any lower in the future since the Panasonic cells alone would cost you more elsewhere. And there is high possibilty to fake ones will apear in a short time as can be seen with 10500mAh version -> see on Youtube .

  72. #72
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    Quote Originally Posted by ledoman View Post
    Yup, at the moment GB has the best price.
    Well, it's not actually "THE BEST" offer but who cares about $3.25 :P

  73. #73
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Is there anywhere lower than $12.59? Where?

  74. #74
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    I don't believe there is, GB and GB are top players right now, everyone else are few steps behind.

    Edit: And even if there is cheaper place to buy I would be suspicious of fakes...

  75. #75
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    BangGood is down to USD 11.99. Gold and Silver are in stock, Pink on the way.

  76. #76
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Just bought 2 from GB (have a bunch of reward points) so under $20 shipped with tracking.

  77. #77
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Will you make pack from the housing described in the OP? I'm thinking to buy more for my friends.

  78. #78
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    I dont know yet, thought about it but undecided.

  79. #79
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    I've just got the Hunk Lee balancing PCM. For testing purposes I've soldered it to the unbalanced 2S3P pack with cells taken from 6cell case described in the first post. Curently voltages are 3.76V and 3.90V for each set in series. After soldering it needed reset ie. few seconds of power from charger in order to give output to the light.

    Will leave it over the night and see if the cells gets balanced. If not, balancing probably works only during charging, Will see.....

  80. #80
    mtbr member
    Reputation:
    Join Date
    Jul 2015
    Posts
    16
    So I just riped the two powerbanks appart I've got.
    I fully charged both of them overnight.
    Fully charged the pack has a voltage of 4.26V, seems like too much imo. I charged two phones with the packs to get them down to 4.18V.
    Sadly I destroyed one of the PCBs while tearing the pack apart (screwdriver slipped off into some small SMD component and ripped it off)
    Don't try to tear the welded contacts off of the PCB, just cut them >_>

    Lets see what I can find to make a waterproof battery pack of these, ordered this cable to use it with the Nitefighter BT70: Cheap 5.4mm Male to Female Extension Cable for SKU 29489/30864 (100cm)
    Hopefully the wire used in it is thick enough.

  81. #81
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    The wire from DX is about 20AWG and is thick enough even for higher currents.

    4.26V is just on the top of alowed limits, no big problem, except regular charging to that level would shorten litefetime/capacity. On the other hand power bank circuit consumes some power (parasitic drain) so the voltage would get lower anyway.

    To report about Hunk Lee PCM. Over the night cells didn't balance itself. Voltages have been 3.77V and 3.89V. Difference 0.01V to previous reading could be difference in the measurments and/or due pure cells I'm using for this test.
    Now I'm charging the pack to see what happens during that process.

  82. #82
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Received my 2 Xiaomi power banks from Banggood. Checked authenticity code on one and it checked legit. These are very nice power banks and a shame to tear apart and rip the cells out of! Didn't open mine up yet.

    -Garry

  83. #83
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    EDIT: Some more explanations from post #107 onward. I might be wrong at some points, so not all below is correct. At least it's NOT total CRAP.

    I've been trying Hunk Lee balancing protection module (PCM) last few days and can tell it's A JOKE, CRAP, whatever you like in regarding balancing feature. Totaly wrongly designed. Did not test other functions except overcharging kicks in at 4.29V on any set in series.

    I've been discussing with Hunk and he doesn't want to admit balancing feature in this PCM is useless. He is taking some reasons for it's usenes which should never happen. Here is his description of balancing feature:

    "If the voltage of one set of battery is lower (A), the other one is higher (B) during charging, when charging voltage up to 8.5V(4.25V each) around, the circuit over-charging action and stop charging, balance function begin to work, first the battery B discharging , if the whole voltage low than 8.5V, over-charging switch off and begin to charging whole battery again. Re-stop when the voltage up to over-charging voltage, then the battery B re-discharging.
    Need much more cycling to get same voltage, not in one cycle in charging and discharging.
    "

    In the first place proper charger should never charge as high as 8.5V so the balancing function would NEVER happen and indeed it does not. I have tested situation with set A of cells at 4.19V and the other name it B at 4.00V before start charging, so 8.2V for the whole pack. After charging some time, set A reaches 4.29V and PCM cut off the whole charging which never restarts.

    Since Hunk Lee also writes:

    "First, balance function is only apply to small voltage tolerance between two sets of battery. As the balance current is very small, higher tolerance voltage need more cycle charging and discharging to get same voltage."

    I've tried it with small difference. Again with good charger the whole voltage of the pack couldn't reach 8.5V so as expected balancing didn't take the place.


    Then Hunk says:

    "What you said is correct, the professional charger is no more than 8.4V, like the chargers in our stock.
    So not only PCM but also charger protect Lithium battery, more safe.
    But someone may use 9V charger to charge battery as they haven't the knowledge of battery and charger.
    "

    So if one not knowing is using bad charger (ie voltages from 8.5V to 9V) might experience balancing while the others using proper charger can't. WTF.

    Will see what Hunk have to say next as we are all using good chargers.
    Last edited by ledoman; 08-13-2015 at 01:48 AM.

  84. #84
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Could you please post hi-res pictures of both sides of PCB (or at least, the component side)?

  85. #85
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Shure, more you can see when click on the picture and navigate left/right
    Shrani.si

  86. #86
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    In other words Hunk is full of crap and I wasted 15 buck in his store

  87. #87
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Well, protection features should work, but this PCM is to expensive for that only. Archie might help us with some ideas. Maybe resistor mod or something. If we lower the trigger when "balancing" (ie discharging) starts we might get out the working stuff.

    Of course I'll continue discussion with Hunk to see what he can do. I've got feeling he has low attitue on this issue.

  88. #88
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Oh wow! Didn't expect to hear it's crap! At least I only bought one!

    -Garry

  89. #89
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    Generally even if you use balance charger like accucell 6 balancing works kind of funky: first it charges all cells and when it detects overcharging of one side of the pack (let's say it's 2S2P pack and one side of the pack is at 8.2V and the other side is at 8.5V) then it stops charging both sides, start discharging overcharged side until they are both at 8.2V, then starts charging again both sides and so on, very time consuming process because of quite low charging current in balancing regime.

    I think you mentioned this but I was also thinking the same thing/idea for a long time: making each cell as a independent unit. Each cell would be equipped with its own charging/discharging control unit so battery packs would be highly modular and scalable systems, just plug couple more cells to get desired capacity or reconfigure their connection to get desired voltage. Something similar as if you were to use protected cells except kicking off protection on one cell would not power down entire pack..

  90. #90
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by garrybunk View Post
    Oh wow! Didn't expect to hear it's crap! At least I only bought one!
    @Garry, so far it looks only balancing feature is a crap, but this is why we've bought this PCM, isn't it?
    I didn't expect to work like this either. Actulay I was trying to get some info from Hunk before, but couldn't get anything usable until I bought a pair and starts testing and complaining. Then he has answered as I wrote above.

    @sirius. If you look closer you would see balancing feature in all hobby chargers works in the similar way ie. discharging single set of cells after charging all. Hobby chargers charge over the main port and check and discharge over balancing port when using balance charge mode. This is done until all cells (or set of cells) in series reach 4.20V (or whatever you have set). Of course this is not the most efficient way of charging. It would be better to work like in the multibay chargers. But can it be done so once they are assembled into the pack? Multibay chargers use common negative side while hobby chargers can't and so can't this Hunk Lee PCM.

    I'm not EE but Achie might hopefuly explain the situation and possible solutions. The only thing I can think of is using big capacitor and temporary store energy from higher cell in it, switch to another cell and release energy from the capacitor to it. Not shure electricaly this can be done. I'm imaging this like you have two barrels of watter and use tiny coffe cup to equalize their levels.

  91. #91
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Quick general question about these PCM circuits and their current ratings: if a PCM circuit is rated "2A", will that be for total output coming out the DC connector to the lighthead? Or is that 2A per bank of cells meaning you could get up to 4A output on a 2S2P pack?

    I also wonder how out of balance good cells will get if the cells are properly balanced to begin with. So I guess I'm back to the idea of connecting balance leads for hobby charger use (which won't be difficult and I'll be able to monitor if/when they do get out of balance).

    -Garry

  92. #92
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    Quote Originally Posted by garrybunk View Post
    if a PCM circuit is rated "2A", will that be for total output coming out the DC connector to the lighthead? Or is that 2A per bank of cells meaning you could get up to 4A output on a 2S2P pack?

    I also wonder how out of balance good cells will get if the cells are properly balanced to begin with. So I guess I'm back to the idea of connecting balance leads for hobby charger use (which won't be difficult and I'll be able to monitor if/when they do get out of balance).

    -Garry
    I presume that rated curent is for entire pack so 2A in total, thats why I bought 7A rated PCM.

    Even good cells can unbalance quite fast, you would have to select cells that have as close it is possible similar internal resistance and capacity.
    Here is a video that I recently made, showing capacity of new Panasonic Eneloop Pro AA cells, notice the difference between cells, the cell would not work well if assembled in bat.pack


    this video is just an example to show how capacity can differ from cell to cell and this was 4x blister pack so they were paired by manufacturer.

  93. #93
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Garry, 2A would be total output current. As with capacity you calculate current only in parallel not in series. In series only voltage is changing. Take this PCM as a flow limiter - bottle neck if you wish.

    How much cells can get out of balance depends on many things. There is no final answer. First depends on cells quality (even the best cels has some variation in it's structure), then how much you stress them (used current). More you stress them more likely the diffirencies would pop up. Other things are number of cycles, opearting temperature, how you charge them + topping, etc....
    Tipicaly with high quality you would get less difference and/or in longer time. Of couse you can't know what would happen for shure unless you try it. Having balance leads and using hobby charger from time to time is shurely vey good aproach. This way you would also discover your pack behaviour and see how frequently if at all you would need to balance it.

    In this thread I was hoping I can use Hunk Lee PCM in order to set and forget about balancing and to make sealed pack without need to open it. Unfortunately it looks this is no go with this PCM.

  94. #94
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by ledoman View Post
    I'm not EE but Achie might hopefuly explain the situation and possible solutions.
    Well, I've tried to reverse-engineer the schematics based on your pictures. The balancing parts are indeed present on PCB, and work by applying the load (68E resistors) to the each branch of battery. I'm unable to identify the exact ICs and transistors used, but Q3/Q4 are obviously the switches (either FET or bipolar), and U2/U3 are most likely OpAmps or comparators controlling them:

    Name:  WH-3811A27.gif
Views: 1926
Size:  3.8 KB


    So, while technically correct, this PCB is probably just faulty, or some components used are out of specification. Without manufacturer's PDFs for these components and further investigation/measurements, the only solution I'd recommend is, to replace the board.


    The only thing I can think of is using big capacitor and temporary store energy from higher cell in it, switch to another cell and release energy from the capacitor to it. Not shure electricaly this can be done. I'm imaging this like you have two barrels of watter and use tiny coffe cup to equalize their levels.
    You're correct: this method is also used in balancing, ang called "charge pump", where voltages between the branches are compared and energy from higher one is gradually transferred to the lower one until they're equal. But this method is typically slow, and AFAIK not used often. In your particular PCB, more common method (of partial discharge) is implemented.
    Last edited by -Archie-; 08-11-2015 at 01:17 PM.

  95. #95
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Archie thank you for your job.

    As I can see markings on the Q3/Q4 are X1XV (1Y) and U2/U3 are GCLV. You probably couldn't find any documentation about them. What are those 4430 8-pin chips and 6-pin AC0D?

    I can try and ask Hunk if the factory can release some information about this. I don't think PCM is faulty, it could be missdesigned for some bizare situations which should never happen if you use a proper charger.

    What I've discovered late last night was some kind of balancing happens, BUT:
    - it starts only when one branch of cells reaches 4.29V (PCM cuts of charging at that point)
    - you have to have charger that charges over 8.5V
    - difference between branches is relatively small
    - you have to apply many cycles > 10 reconnections - plug/unplug the charger tested with low capacity 2S3P pack - cells from the OP

    So without investigation user would have no idea what's happening. I think I should try it again with brand new cells and 2S1P pack to make it happen faster and more stable (very little voltage drop in the cells after charging). And I can try the second PCM too, but it will takes time....

  96. #96
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by ledoman View Post
    Archie thank you for your job.
    You're welcome!

    As I can see markings on the Q3/Q4 are X1XV (1Y) and U2/U3 are GCLV. You probably couldn't find any documentation about them.
    Sure, most SMD components use abbreviated marking, and it's quite difficult to search for the data by two-four characters.

    What are those 4430 8-pin chips and 6-pin AC0D?
    Power MOSFET transistors cutting the battery when overcharge/overdischarge occurs, and voltage-monitoring IC that controls them. Pretty much standard design...

    So without investigation user would have no idea what's happening. I think I should try it again with brand new cells and 2S1P pack to make it happen faster and more stable (very little voltage drop in the cells after charging).
    According to your description ot PCB behaviour, you'll check the durability of your brand new cells at the same time, too... Better discard it.

  97. #97
    mtbr member
    Reputation: Cat-man-do's Avatar
    Join Date
    May 2004
    Posts
    6,080
    Added thoughts to the conversation:

    When it comes to circuits designed to control and balance a cell pack; I would think ( I'm no expert on this type of thing ) that it would be very important that the cells used in the pack be as closely matched as possible. The internal resistances ( or combined internal resistance when in parallel or in matched P/Series configurations ) should be the same and the capacities should be the same when in match sets. All the cells don't have to be exactly the same but it they are configured and matched probably in the PS set-up they should equal out and provide a decent long lasting battery ( as long as none of the cells develop a "soft short".

    I doubt most self balancing PCM's are really capable of truly balancing a cell arrangement that has cells that are really badly out of balance. More than likely the current used within the battery for doing this is using only very low milliamps or microamps. Whether that would be enough to counter balance a cell with a much higher self-discharge is anyone's guess. That's just my theory though. I don't know for sure because I don't design these things.

    Anyway, If I were to build a battery my major priority would be to match capacity/resistance within the configuration as best as possibly. Since I don't have sophisticated equipment for doing that I would have to do multiple discharge tests and that would be quite time consuming to say the least..

  98. #98
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Cat-man-do, I have necessary equipment and carefully match the cells when building my battery packs. They're indeed work fine without going out of balance for quite long time.

    BTW, it's better to measure not DC resistance of cell, but its AC impedance: it's industry standard for such things...

    As for self-discharge current - it's essentially negligible for LiIon cells.

  99. #99
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Cat, yes I know and probably most of us that using good quality brand cells is the way to go. My point (maybe I didn't emphize it, but I hoped it was obvious) is to build sealed pack "set and forget" similar to the computer laptop packs. As you can see in this thread I was pointing out for brand new Panasonic(Sanyo) cells from Xiaomi power banks. As add-on to that I was naively hoping Hunk Lees PCM is working as self contained laptop pack would. (well, without thermal probe). I've explained to the Garry I didn't get any usable information prior to buy and test it.

    Hunk Lee PCM has specified 70mA balancing current. This is much more than any self discharging should happen. In the ideal world if balancing would work all the time both branches would have same voltage all the time. Of course this can't be true under workload, but could slowly happen when user finishes his work. Same is obvious when you charge pack with the currents over 70mA.

    All in all I was trying anyone could built high quality battery pack you can't buy anywhere. I'm not aware about any that would have balancing feature integrated.

    Still we might get better PCM if Hunk would be willing to cooperate and Archie might help with electrical design. Until then regular checking and balancing with hobby charger (or other method) is the way to go. It is just you can't have sealed pack "forever".

  100. #100
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by -Archie
    BTW, it's better to measure not DC resistance of cell, but its AC impedance: it's industry standard for such things...
    Can you explain little bit more about this and how to do it if it's possible with multimeters. If MM is not enough then, well, we won't do it ;-/.

  101. #101
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    No, it's impossible to do with ordinary multimeter. You'll need special device: impedance meter with 4-wire Kelvin probes, applying 1000 Hz sine wave AC signal to the cell being tested.

  102. #102
    mtbr member
    Reputation: Cat-man-do's Avatar
    Join Date
    May 2004
    Posts
    6,080
    Quote Originally Posted by ledoman View Post
    Can you explain little bit more about this and how to do it if it's possible with multimeters. If MM is not enough then, well, we won't do it ;-/.
    Battery University had a good explanation on battery resistance but didn't quite explain how to make the calculations. I think for our purposes we could forget about the capacitive reactance factor even though if we could add it to the total resistance it would yield a more accurate reading.

    This youtube video shows a guy measuring internal resistance. Looked pretty easy. You need that watt meter though so you can get both the voltage and current reading at the same time. His method look pretty easy. Problem is once you energize a load the resistance of the load can change. There in is the problem that he didn't talk so much about. Still, if you do the test a number of times you should be able to get a reading that is somewhat accurate.

  103. #103
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Cat, my hobby charger iCharger 106B+ and Opus BT-3100 (4bay charger) can measure internal resistance, so no problem for me. It can be good to watch for others, though.
    I was more qurious about AC impedance as the other method to check quality. But anyway it is good to point this out.

  104. #104
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    If anyone still interested in Xiaomi Power banks - currently only $11.99 at BG -> Original XIAOMI 5.1V 2.1A 10000mAh Power Bank For Smartphone Sale-Banggood.com

  105. #105
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Well looks like gb free shipping worked well this time, should be here in a couple days, less than 2 weeks for my pair of the power banks. Though im just stealing cells from them hehe.

  106. #106
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Ledoman, thanks for the updates. Archie, thanks for the analysis and schematic.

    I guess if the balance circuit would work at 4.20V it would be another story. But then, with device-to-device variation you's probably be looking at a range extending below 4.2V and the folks who got one of those would never be able to fully charge their battery with a standard charger.

    Let's say, for example, that your PCM, due to expected tolerances, had the balance drain activate at 4.18V per cell, and your charger was charging to 8.46V (the lowest of my magicshines has this open circuit voltage). Let's further suppose this charger indicates red light when voltage is below 8.38V or 8.42V (memory fails me, it think it is thereabouts). With that setup, since charging currents are very small in the last part of the CV phase, the balance drain (70mA per cell) would sink all the charging current, and the end result would be the charger indicates red - charging forever.

    So, from a conceptual point of view, a balance PCM of this nature would need to be either extremely well matched with the charger, or the charge termination indication would need to be moved from the charger to the PCM. For example, once both Q3 and Q4 are on, charge complete indicator should light up. This is doable, but a caveat is the charger then needs to supply a voltage that is higher than the balance circuit's working voltage, but not so high that current would be >70mA at this voltage point. That's a very big caveat! Essentially, this again means charger has to be closely matched to the balance PCM.

    For comparison, hobby chargers charge at the same voltage they are trying to balance at. So much simpler. They also have a CPU which can compare voltages and kick in the balance drain before they reach CV phase (not sure if they do this, though - I think they only do it in CV phase).

    If an analog circuit could be fashioned to compare cell voltages rather than compare each to a threshold, then a balance PCM of this kind might be made to actually work by performing the balancing during the entire charge cycle. Problem is, Cell 2 voltage is actually the sum of Cell1+Cell2, so in order to compare the two you would need to subtract cell1 from the cell1+cell2 measurement. It's been too long since I designed with op-amps, but I think it might be possible. Not trivial, and would require more components, but possible. Another problem is you probably wouldn't want it to balance during discharging, so balance circuit would need a way to detect direction of current flow. Again, more complexity and more components required.

  107. #107
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    I was bit seduced by Hunk explanation about 8.5V limit when balancing should start working. That high voltage is not necessary. And mostly I was testing to big difference. I have to stand somewhat corrected and will edit my posts acordingly later on. It looks it's not completly worthless.

    Obviously PCM is designed to work only with very small difference between branches (cells in series). The difference should be probably lower than 0.1V. I will test that situation again later.

    What I've discovered so far with bigger differences PCM let it charge up to 4.29V for higher branch and then cuts off charging. PCM is then slowly discharging higher branch down to about 4.20V (don't know exact figures atm) and this is it unless you unplug and plug charger again. It stays as is (except some voltage drop when cells rest a bit).

    So for example: we have 8.43V charger, branch A with 4.06V and branch B with 4.18V at 2S3P battery pack.
    When charging the pack for some reason I don't know voltage of branch B raises faster than the the branch A. When B reaches 4.29V the A reaches merly 4.09V. Since B reached 4.29 PCM is cutting off charging. You need to reconnect the charger. If you do it soon B is still near 4.29 so PCM cuts off charging again. Only solution to that is to leave it for a longer periods so the PCM discharges branch B. With 3 cells in parallel this can be quite some time. Then you repeat that process until branch B charges below 4.29V because voltage of branch A comes higher and sum is 8.43V (defined by charger you use). So A should be: 8.43V - 4.29V > 4.14V.

    Then it depends on charging current as mentioned Morten. It may slowly get balanced, but it might not, at least not in reasonable time. Probably the whole thing is also dependant on cells quality and their number in parallel. This is why (well, I've got that feeling) I don't get consistent results with different chargers and at different voltage stages. Anyway very time consuming process.

    I think using PCM along with quality cells and small voltage difference should show better results. Again it takes quite some time to test different situations. Will let you know.

    BTW, Morten thanks for your point of view.

  108. #108
    mtbr member
    Reputation: garrybunk's Avatar
    Join Date
    Feb 2014
    Posts
    1,745
    Thanks ledoman for your continued testing on this. I don't like the idea of over-charging a bank of cells that much, but like you said using good quality cells that matched well originally you shouldn't be charging a bank that high.

    -Garry
    "My Bike Lights" Thread on BLF teardowns, measurements, and beamshots. Moving my photos, PM or post up if you can't see them.

  109. #109
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Garry, yes I agree. I wanted to discover all potentional situations. Shurely will try describe more normal situations that should happen with quality cells.

  110. #110
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Quote Originally Posted by ledoman View Post
    So for example: we have 8.43V charger, branch A with 4.06V and branch B with 4.18V at 2S3P battery pack.
    When charging the pack for some reason I don't know voltage of branch B raises faster than the the branch A.
    I think the reason is the same reason battery discharge (voltage vs mAh) curves are steep near fully-charged and fully-discharged, but shallow/almost-flat in the middle. I too can't quote a chemistry reason, but I think it is normal and expected.

    Your thoughts sound right on, you may be able to balance that back out by repeatedly charging it to where the overcharge protection kicks in. But I think in doing so you will cause wear on your bank2 cells and by proceeding you are creating a greater need for balancing than you may have had to begin with.

    When you find out the exact cell voltage it discharges down to, let us know. At that point I would recommend you stop your testing and balance the cells properly - assuming you care about these cells. I think you are right, the balance feature may work proactively a lot better than it will in your reactive test case.

  111. #111
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Right now I have situation with A charged to 4.20V and B charged to 4.25V. The pack is still on the charger and it doesn't look it will balance at all. Will disconect the charger and see what happens next. It should balance itself and big resistors should get warm/hot, right?

  112. #112
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Ok I have to ask, what pack set up are you testing this on, a 2s or a 2s?p ???

    I only ask because there is a lost point in this that I forgot about till now (from days of dealing with Lipo Packs and my pushing 2 decades as an automotive technician involving vehicles that use multiple batteries)

    A balance circuit to work in packs with parallel series sets MUST be able to read the cells individually.

    a protection circuit can read the voltages in each bank (not sure the algorithm used for that) but a balance charge circuit has to be able to affect each cell individually. Paralleled connections make this impossible. Those connections have to be removed to balance cells in each series set.

    ps: this is why balance leads are not used and are pointless in welded packs that are NOT only series connected cells. any added sets in parallel render that type of balancing pointless and ineffective.

  113. #113
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    No, it isn't correct. If the cells are connected in parallel, there's no any possibility for them to be of different voltage in first place...

  114. #114
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    I'm using 2S3P setup (I mentioned that before few times) as can be seen in the OP in the beginning of this thread.

    And NO, no need to check individual cells since all 3P (bank as you call) are giving out same voltage reading and BTW they are welded. There is no other way electricaly speaking. 3P bank (or any other number of cells in parallel) would always balance by itself - the system of connected pots if we taking about liquids wich is essentinaly the same as electric. Each bank acts as one big cell in this situation.
    Li-Pos are all the same, but since they are not welded you can can probably disconect them somehow (don't think so either).

    PS. Idealy would be as you think so the each cell in the bank would be topped to the same voltage instead balancing itself and put out an average voltage of all in the bank. But in reality this would cost way to much and would involve switches to each cell in the bank + some more sophisticated electronic.

  115. #115
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    well if they are all in parallel, but multiple sets of series pairs. so 2 sets of 2 cells in series connected, each series pair will match voltage with each other, but each pair in series wont self balance. Must be something Im missing but as I recall it you can have:

    series set 1: cell @ 3.4v and cell @ 3.8v = total voltage 7.2v
    series set 2: cell @ 3.6v and cell @ 3.6v = total voltage 7.2v

    connect sets into parallel there will be no change in voltage in any cell as both pairs are the same voltage.

    Wheres the detail Im missing with these that will balance the cells in series set 1 so all cells read 3.6V?

    I know if series set 1 and series set 2 had different total voltage THOSE would self balance to match each other. All "cells in parallel" will self balance to match voltage. I beleive i posted something about it where i Had tried the other methods and found putting all 4 in parallel was the best way to get all 4 cells balanced (since matching is a whole different ball game) before installing in cases.

  116. #116
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    If above you are talking about tipical 2S2P setup we are using you are missing cross conection between each set in the middle point.

  117. #117
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by ledoman View Post
    It may slowly get balanced, but it might not, at least not in reasonable time. Probably the whole thing is also dependant on cells quality and their number in parallel. This is why (well, I've got that feeling) I don't get consistent results with different chargers and at different voltage stages. Anyway very time consuming process.

    I think using PCM along with quality cells and small voltage difference should show better results.
    In other words, this particular PCB is actually not suitable for our needs.

  118. #118
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Quote Originally Posted by ledoman View Post
    Right now I have situation with A charged to 4.20V and B charged to 4.25V. The pack is still on the charger and it doesn't look it will balance at all. Will disconect the charger and see what happens next. It should balance itself and big resistors should get warm/hot, right?
    Not necessarily. You said your charger was 8.43V, right? Or if your individual cell readings are correct, it must be 8.45V. If this is the case, the pack has been charged to the end of the CV phase, meaning charge current was nil and it won't matter that you disconnect the charger.

    Sounds like the balance drain opens at 4.25V, and now that your cell A came up in voltage, the total pack voltage was high enough that charge current supplied to the cell was less than 70mA. This would have allowed cell A to charge while cell B's current was diverted to the drain resistor, thus not causing cell B to overcharge and hit the overcharge protection. This would have gone on until Cell A brought total pack voltage up to the charger's end voltage and current would have gradually diminished to 0A in this process.

    Anyway, that's the theory - in practice, please report if the 4.25V cell stays at 4.25V after being disconnected for a while.

    Assuming Archie's schematic is correct, you should think of this not as a balance feature, but an individual-cell excess charge drain. Which will provide some balancing, but with a greater margin of error/tolerance than what you would see in a hobby charger.

    I don't think it is fair to say this PCM is useless - it may be a case of managing expectations. With cells that start to drift out of balance, it would slow that process considerably.

  119. #119
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Ok here's my question, if somehow the center connections in 2s2p allow all 4 cells to balance to equal voltages (which I still dont see how that's possible gotta read up more) then why is there any concern of a balance circuit??? 2s2p/2s3p packs perfectly balance themselves based on what your saying, there is no need for a balance circuit on the PCB or balance charger then.

  120. #120
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by Morten Schmidt View Post
    Anyway, that's the theory - in practice, please report if the 4.25V cell stays at 4.25V after being disconnected for a while.
    It stayed there for a while. Both A and B sag by 0.01V which is normal being cells are not top quality.
    Then recharging again raised A to 4.21 and B to 4.22. There is still 0.01V difference.

  121. #121
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Tig. Here is brief sketch of 2S2P setup without balancing circuit.
    Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB-2s2p-sketch.jpg
    Think as being A and B one big internaly self balanced cell. Connections in blue are welded. Now think about all routes electics can flow through this setup.How many do you count?

  122. #122
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Quote Originally Posted by ledoman View Post
    It stayed there for a while. Both A and B sag by 0.01V which is normal being cells are not top quality.
    Then recharging again raised A to 4.21 and B to 4.22. There is still 0.01V difference.
    I thought you said cell B was already at 4.25V?? If it sagged by 0.01V, it would be at 4.24V. Please clarify.


    Quote Originally Posted by ledoman View Post
    Click image for larger version. 

Name:	2s2p-sketch.jpg 
Views:	2031 
Size:	63.1 KB 
ID:	1008510
    Now think about all routes electics can flow through this setup.How many do you count?
    More importantly, how many places could you measure voltage and get a different result?

  123. #123
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Ya that lay out I know. A and b will balance on their own between each other, but u have 3 contact points. So 3 voltage readings. Balance reads voltage at those 3 points. + - and center.


    Ok nvm I see what's going on here now, apparently brain not working right atm. The center bridge between series pairs......thnks for the picture I knew the layout in my head but I was thinking of it as 2 series pairs, NOT 2 parallel sets in series. I had it backwards. Without the "H" in the center it doesn't work, the connection between making the series sets connected is what makes the difference.

  124. #124
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by Morten Schmidt View Post
    I thought you said cell B was already at 4.25V?? If it sagged by 0.01V, it would be at 4.24V. Please clarify.
    Yes after few hours it saged by 0.01V. I've reported 4.25V just after disconnecting the charger. So it ended at 4.24V.


    Quote Originally Posted by Morten Schmidt View Post
    More importantly, how many places could you measure voltage and get a different result?
    I measure it at JST connector with 1S-6S voltage checker (see post #52). I've found it very accurate. It repeatedly displays total voltage and voltage of each bank.

    Tig, yes there are three reading possible as described above.

  125. #125
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    Any news about this issue?
    I got the same balance board as ledoman, several days ago actually, but didn't have time to quickly test some cells and build 2S2P pack for balance board testing. Also results that ledoman got are a bit discouraging...

    Just and idea, maybe instead of using balancing board we could use only protection boards and find a way to charge cells as if they are all in parallel and discharge them as they are connected so 2S. Something like a connector that will disconnect serial connection when plugged in and allow using 4.2V ~3A charger!!!

  126. #126
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Quote Originally Posted by Sirius9 View Post
    Just and idea, maybe instead of using balancing board we could use only protection boards and find a way to charge cells as if they are all in parallel and discharge them as they are connected so 2S. Something like a connector that will disconnect serial connection when plugged in and allow using 4.2V ~3A charger!!!
    You would need to hook up that switching circuitry before the protection PCB, and the risk of shorts can't be ignored.. Kaboom... Dropping the protection PCB and routing a 3-lead cable out of the battery pack without any form of protection sounds like a better idea (but I will stick with protection and maybe open it up every off season to check if cells remain balanced).

  127. #127
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Just for your info. I've reassembled 2 Xiaomi power banks from OP with used Sanyo 2800 cells taken from an old 3S3P laptop pack. The cells still holds about 2200mAh at 4.20V charge. Since those Sanyos are 4.30V cells the real capacity of new ones at 4.20V is 2300 mAh (source HKJ tests). The remaining capacity is more than good enough. I'm loosing only 100mAh per cell.

    All in all I've got two working 6600mAh power banks and 6 completly new 3150mAh cells for less than $30. Nice deal which did not involve that much work. Adding some $15 for 2S3P battery case from OP I'm getting also huge capacity 10000mAh battery pack for bicycle lights. Now I would need BT70 light to suck all that energy.

  128. #128
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Could you please clarify, how difficult it is to open the case? As I see on your pictures, top & bottom covers are attached with superglue or the like. Is the plastic brittle, and what are chances to break it during disassembling?

  129. #129
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    You need to open just front side. There is white plastic "sheet" you need to gently pry with knife or similar. It is glued with some kind of tape glue. Keep the glue intact as you can so it will stick later when you reassemble it back.
    Then there are four screws you need to uncsrew. Again you need to pry everything the rest out of the shell or gently bounce so the batteries and circuit will pop out.
    Try it and you'll see, with your skills it shoudn't be difficult
    Report it back so we would know you didn't brake anything.

  130. #130
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Ok, thanks! Maybe I'll order this 6-cell case then. Currently I have plenty of unused loose cells: perhaps, yet another DIY-battery will fit the gap between my low-capacity and high-capacity ones...

  131. #131
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    I got my 6 and 4-cell packs from gearbest. The cord on the 4-cell pack was defective, there seems to be a disconnect somewhere along the cable. If I bend it one way, it works but if I bend it the other way no connection. The 6-cell seems 'ok'.. I used it once without modifying as I didn't have time, and it lasted almost exactly 2 hours with the Ynding light on high (so about 2600mAh). Those are truly crappy cells! But I like the case and the mounting straps seem decent and are quicker to attach than the (old) standard magicshine cloth pouches (especially when wanting to route the straps underneath cables on the top tube).

    Anyways, check your cables folks. Be a shame to do this great mod only to find you have an intermittent connection in the cable. Run a finger along the cable bending it in a tight radius each direction while a light is connected, to check the cable. Gearbest have offered me an acceptable solution btw.

  132. #132
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by Morten Schmidt View Post
    But I like the case and the mounting straps seem decent and are quicker to attach than the (old) standard magicshine cloth pouches (especially when wanting to route the straps underneath cables on the top tube).
    I'd prefer two Velcro bands (or wide single one), personally. In original MagicShine batteries, the buckles on rubber straps are too fragile, and plastic "claw" breaks soon. Probably, the quality of cheap Chinese copy is not better...

  133. #133
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Well, you can still use Velcro after original straps breaks . I can imagine this is more problem in the cold conditions. Anyway it's hard to get good and cheap solution that's why we are making so many mods.

    Morten, what was solution from Gearbest? It would be good to know for other buyers if the same thing happens to them. At some time in the future I might try to change the cable for rounded 20AWG from DX. I'm wondering if it can pass stock gland (or whatever is correct word in English) to keep the pack sealed.

  134. #134
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by ledoman View Post
    I can imagine this is more problem in the cold conditions.
    That's correct; we're riding at winter here, too.

    I'm wondering if it can pass stock gland (or whatever is correct word in English) to keep the pack sealed.
    Gland typically means separate device with compression nut, while in these boxes strain relief is used. In my practice, it's possible to re-use them for new cable, as long as you're careful removing the old one...

  135. #135
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    They offered to send a new one or refund to my wallet, which I guess means store credit. The existing flat cable is a piece of cake to remove from the strain relief btw., and the round 20AWG cable is a perfect fit.

  136. #136
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    Finally yesterday I received my 2 x 10K mAh packs and now, looking them side by side (one gray and other red) I am a bit sad to have to chop them a part, especially since that balancing board was so disappointing that we actually do not have good all in one balancing solution (except using balance charger like accucel 6 which is not all in one thing).
    Edit: changed 10 KmAh to 10K mAh to avoid further nitpicking...
    Last edited by Sirius9; 08-27-2015 at 09:18 PM.

  137. #137
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Morten, nice to hear 20 AWG would fit.

    Sirius, sorry about that PCM. Hunk didn't want to give it to me for testing (and he regreat it afterwards) so I was naive and seduced the others. Well it still has protection features, just it's not that cheap.
    With those cells from Xiaomi you wouldn't have much need to balance them, maybe once or twice a year. My goal was to make pack "set and forget", but now I would still make it so I can open it from time to time. Don't be sorry for good looking Xiaomi pack, the cells inside looks even better . If you have spare cells from computer pack you won't regret to replace them. I did and it works like new (except capacity).

  138. #138
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by Sirius9 View Post
    I received my 2 x 10 KmAh packs
    KmAh? KilomilliAmpere-hour? I like this new measurement unit!

  139. #139
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    10K as 10.000!

  140. #140
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    If you divide some number by 1000 (this is what "milli" prefix does) and then multiply the result by 1000 ("kilo" prefix), in the end of calculations you'll get the original value back... There's no such thing as "kilomilli".

  141. #141
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Probably the Sirius was just to lazy and shortened 10.000 into 10K. No need to argue on that. Let stick with the topic.

    Sirius and Morten, have you opened your Xiaomi power banks already?

  142. #142
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Think he gave Archie a headache with his new way of labeling batteries. But I think simply writing 10Ah would have been the ultimate laziness

    Made me think and laugh at how things are named poorly. mAh really could have stopped when battery capacity hit 1Ah. 3800mah is just 3.8Ah. So 10000mah is just 10Ah.

    It would be interesting how many people would loose their minds if companies would just start labeling batteries in Ah. But its just like people buying lights for the number of lumens (whether true or false) not caring if they all actually be able to see anything.

  143. #143
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by tigris99 View Post
    It would be interesting how many people would loose their minds if companies would just start labeling batteries in Ah.
    I have better business idea: label them in microAmpere-hours! I'm pretty sure, for many customers 3.000.000 Ah cell is clearly better, than miserable 3 Ah...

  144. #144
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    ^^smartass^^^

  145. #145
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829

  146. #146
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    Quote Originally Posted by ledoman View Post
    Sirius and Morten, have you opened your Xiaomi power banks already?
    Sorry, just now noticed the question :/

    I did open one of them just to peek inside but I still don't have battery box from first post and I am still thinking what to do about charging this cells as if they are independent "banks"3P + 3P

    Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB-top.jpg
    Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB-cells.jpg

  147. #147
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    The only way I can think to do it but its not worth doing (each bank self balances anyway so no point) is using the center point and only one main lead at a time. So one bank use the positive from the pack as the positive for the charger, then the center connection as the negative to charge that bank, then reverse using negative battery lead as negative to charger and positive is the center connection. But you'll have to use 2 separate plugs for the pack, one for each bank so you dont accidently try to reverse polarity charge the wrong side.

    Using a balance lead off the center and a proper balance charger is so much easier, safer, and will actually get a proper balance.

  148. #148
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Or he would need high current switch to disconnect/select/rearange banks. Could be done, but it's not worth it when you can buy 2S-3S balancer for just $5-$10 at eBay. But it's his choice

  149. #149
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    ledoman, do you know if there is an updated code? The one above is expired accorging to GB...

  150. #150
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Don't know. It has worked a week or two ago. I can ask GB to extend the coupon.

    EDIT: Arranged, both coupons 4BPack and 6BPack should be valid till Oct. 31th.
    Last edited by ledoman; 09-01-2015 at 05:53 AM.

  151. #151
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Thanks!

    Unfortunately, I'm unable to put yet another "+Rep" to you, but hope other members will use the opportunity...

  152. #152
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Yeah, I can't put to you more either We both have two marks, but I don't care to much about that. I'm glad could I helped you.

    BTW. I might have one question for you about repairing SS X5 driver. It seems to me the zenner diode has blown up, but might be wrong. We should discuss it on private. I'm not in hurry as KD sent me spare one so I have working light.

  153. #153
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    I assembled my 2S3P pack using same balance PCM. As assembled, first bank was at 4.16V and bank 2 at 4.19V. After charging with 8.43V charger, bank 1 ended at 4.21V and bank 2 at 4.22V, so it seems the balance PCM does work. As explored above, it is no good at balancing out a severe imbalance, but will proactively prevent such severe imbalance from occurring in the first place, which is what you really want. I will be ordering one or two more.

    I also have to report not all the Xiaomi banks use the Panasonic/Sanyo cells. I have a total of 7 so far, 3 silver and 2 pink all came with Panasonic/Sanyo cells, but 2 Gold ones came with LG F1L cells. I can try to post a photo if anyone needs to see it, but the cells are basically purple and say "LGABF1L1865" on them. A bit of googling confirms this is a LG 3350mAh cell. I don't know if it is just coincidence my two gold colored units came with these, but thought I'd post a heads up and tentatively suggest people don't plan to mix and match gold with silver/pink unit cells.

  154. #154
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Morten, it would be nice to visualize this about LG cells. It's easier to remember. And it would be good to know what was the source you bought from. Maybe not important but there might be different batches around.

  155. #155
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    In fact, it's not a problem to mix different cells as long as they're designed for the same working voltage.

    As for particular brands used in Xiaomi banks, I think I've seen manufacturers note somewhere, with no exact cell supplier noted, but with promise that only brand-name cells are used. So, everything seems to be correct!

  156. #156
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by ledoman View Post
    BTW. I might have one question for you about repairing SS X5 driver. It seems to me the zenner diode has blown up, but might be wrong. We should discuss it on private. I'm not in hurry as KD sent me spare one so I have working light.
    Sure, drop me a line anytime.

  157. #157
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Review: 4 (2S2P) and 6 cell (2S3P) hardshell waterproof bicycle battery packs from GB-img_20150902_015340.jpg


    Well, I won't be mixing cells. To each his own.

    Edit : The gold colored units are from BangGood. I am not saying they are bad cells, I just will prefer to not mix and match. Although they might be good cells they may not last exactly as long as the Panasonic/Sanyo cells (could age faster, or could turn out to be more resillient and age slower) - and in that case you would get imbalances, possibly more than PCM can handle. Best to keep these for a separate 6-cell battery pack IMHO.

  158. #158
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Thanks Morten. Don't have exact curves of LG F1L cells on the HKJ pages yet, but if I compare LG MH1 (3200mAh) to Sanyo NCR18650BL they are pretty close match. I can suspect LG F1L is on paar with that Sanyo. Of course there would always be some differencies.

  159. #159
    mtbr member
    Reputation:
    Join Date
    Nov 2013
    Posts
    40
    We have detailed connection diagram in the description of the item. Pls. see it in our ebay store.
    Or you can contact us for a reference picture.
    FMA Battery
    No Battery, No living.

  160. #160
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Hunk would you provide the link or copy of picture to connection diagram picture here so the users would easier find it. No need to contact you each time. Thank you.

  161. #161
    mtbr member
    Reputation: ljracer's Avatar
    Join Date
    Nov 2009
    Posts
    293
    Would a modded 6 cell version of this pack drive 2 yinding light heads with a y cable or is the voltage drop too high?


    thanks in advance!!

  162. #162
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    6 cell version should run 2 yindings w/o problems. You just could see some voltage drop because of Y adapter and not because of battery pack. But still the voltage of the pack should stay high so the voltage drop would not be noticeable. Even 4A PCM should work, but I would suggest higher current PCM so you can use same pack for more powered lights like BT70.

  163. #163
    mtbr member
    Reputation: ljracer's Avatar
    Join Date
    Nov 2009
    Posts
    293
    Edit 9.9.2015:
    I would need to add the PCM doesn't have to be neccessary balancing one. Anyway I would go with the higher current ones as they have less resistance and are more apropriate for the BT70 type of lights.
    Hunk Lee has some - HERE. Take a look detailed description of each where you can see Maximal continuous Discharging current. Be aware of Over charge detection voltage, some are for 4.30V cells - like this.

    Question regarding the PCM: Is the PCM that comes with the pack total junk? If I was not going to add a balancing PCM, could I just reuse the existing one and add better cells?

    Also what's a good source / name for the flat metal strips that are used to connect cells together?

    thanks!

  164. #164
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Quote Originally Posted by ljracer View Post
    Question regarding the PCM: Is the PCM that comes with the pack total junk? If I was not going to add a balancing PCM, could I just reuse the existing one and add better cells?
    Yes (it seems to be usable).


    Quote Originally Posted by ljracer View Post
    Also what's a good source / name for the flat metal strips that are used to connect cells together?
    Search ebay for "Nickel plated steel strip".

  165. #165
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Can search solder strips too, I found some on amazon.

    But this isn't something just anyone can do, can't just solder liion batteries like you can anything else. You need a high powered soldering iron, have to scratch the battery plates, tin them, tin the strips, then solder very quickly.

  166. #166
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    ljracer, yes you can use existing PCM, but I would advice to get higher current one if you plan to power stronger (more powerful) lights in the future.
    Maybe Archie can tell what current the stock PCM can stand - see the picture of it.

    Instead of "Nickel plated steel strip" I'm using 0.1mm copper sheet like one you can see on my pictures on page 3.

    Tigris, if one goes Xiaomi power bank route does not need hight powered solder iron cause the cells are already welded. All you need is to make bottom cross connection and wire them correctly. Of course you still need to warm up existing strips very well, but 40-50w iron could be sufficient if you use substantial amount of solder and maybe thicker tip.

  167. #167
    mtbr member
    Reputation: ljracer's Avatar
    Join Date
    Nov 2009
    Posts
    293
    Quote Originally Posted by tigris99 View Post
    Can search solder strips too, I found some on amazon.

    But this isn't something just anyone can do, can't just solder liion batteries like you can anything else. You need a high powered soldering iron, have to scratch the battery plates, tin them, tin the strips, then solder very quickly.
    Found 1M on ebay for 1.19 shipped. Thanks!

    Also I'm going to try with 180W gun....should be quick enough... easier than building spot welder..

    Tigris99 have you had good success doing it the way you describe? thanks!

  168. #168
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    180W iron should do it very well and the procedure described by Tigris is correct. And don't forget to use good soldering paste. It makes it much nicer.

    Please post link for the strip you have found on ebay.

  169. #169
    mtbr member
    Reputation: ljracer's Avatar
    Join Date
    Nov 2009
    Posts
    293
    1M 5mmx0 15mm Pure Ni Plate Nickel Strip Sheettape for Battery Pack Welding DIY | eBay


    Now I just need to figure out which is the best PCB for price to run 6Cell pack for 2 light heads.

    Thanks!

  170. #170
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Thanks, I've found also 8mm wide one: 1M 8mmx0 1mm Pure Ni Plate Nickel Strip Sheettape for Battery Pack Welding DIY | eBay
    It might be better for higher currents.

  171. #171
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Hunk lee has auctions for 2-pack of both the 4A and 7A balance PCM's. If you might need an extra one for a 2.nd battery pack, this lowers price a little bit. This is for two 7A PCM's:
    2pcs 7 4V 2S Li ion Li Po Battery Protection Balance bms PCM Current 7A 2S7AW006 | eBay

    If you are putting in good cells, the balance PCM may extend their lifetime and be worth it. If you want to use two light heads at full power with the stock PCM, I think it would make sense to test it before your ride (use a fan to cool the lights) and/or bring a backup battery with you.

  172. #172
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Morten, he wants to run two Yindings who runs at less than 2A each if I recall it corectly. So 4A PCM should be sufficient, but personaly I would go with 7A, just in case.

  173. #173
    mtbr member
    Reputation: ljracer's Avatar
    Join Date
    Nov 2009
    Posts
    293
    Thanks for tips. I'll try the 7A. I am planning on using some 2600 samsungs that i have 6 left of. 2 got killed by that solarstorm box with usb. I dont feel like modding those boxes since i have two original panovo boxes i run protected panasonic 3400s in. I always have at least 3 lights when going for night rides. I typicall only turn on my second yinding for dh sections.

    I'm going to practice my soldering on the dead ones.

  174. #174
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Here is how I would try the soldering if I needed to. Tin one side of the steel strip. Hold tinned side against battery cell and heat from opposite side. Use high powered iron. Remove as soon as strip moves closer to terminal and/or solder is pressed out the sides.

  175. #175
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Morten, I drill a tiny hole where I solder the strip to the cell, then tin. The hole gets solder in it, so when you put the iron too it, its solder to solder contact, seems to go a bit quicker as the solder strip plays heat sink first before melting solder in the other side. The solder tabs that used to be used (well still are somewhat but...) for rc had several holes for this reason, less thermal dissipation through the strip and to neighboring cells before the solder melts and attaches to the cell your working on.



    Ive never done it with liion cells (not yet, about to) and the only major difference is using a bigger iron so you attach each cell faster than the old NiMH cells. Even those had limitations on how hot you wanted to get them, and I used to be able to solder an entire pack, put connector on and shrink wrap in about 5mins once cells were laid in the holder.

    Ps: I use a snap-on 220w gun . Over kill as I can solder a car battery cable with it no problem. So 180 if you do the hole drilling method the cell wont even warm up before your on to the next cell.

  176. #176
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Just a side note. If you would drill the strip then make holes only where you want to solder. Otherwise you make more unnecesary resistance. It's tiny difference, but every bit counts.

  177. #177
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    IMHO the method with tinning opposite side of a strip will result in less heat transfer to the cell. With the hole method, you need to either rely on applying all the needed solder to the soldering iron (which makes controlling the amount difficult and is not optimum for solderability) or you need a two-step process of first heating and then applying additional solder (which will heat the battery for a longer time).

  178. #178
    RAKC
    Reputation: tigris99's Avatar
    Join Date
    Aug 2012
    Posts
    7,147
    Morten, sorry I should have clarified:

    Drill the holes, tin the side that contacts the cell terminal as you normally would. I dont do anything different than you guys would just add the holes. So when you put the iron on it, it has some direct tip to solder contact, no just heating through the strip till it melts the solder on the cell side.

    Hope that makes sense.

  179. #179
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by Morten Schmidt View Post
    Click image for larger version. 

Name:	IMG_20150902_015340.jpg 
Views:	754 
Size:	126.1 KB 
ID:	1012654
    Well, I won't be mixing cells. To each his own.

    Edit : The gold colored units are from BangGood. I am not saying they are bad cells, I just will prefer to not mix and match. Although they might be good cells they may not last exactly as long as the Panasonic/Sanyo cells (could age faster, or could turn out to be more resillient and age slower) - and in that case you would get imbalances, possibly more than PCM can handle. Best to keep these for a separate 6-cell battery pack IMHO.
    Morten, got one golden colored from Banggood and there are red NCR18650BL inside. I've got the same cells in 2 silver, 1 red and now 1 golden power bank so far. So no rule what you would get. :/

  180. #180
    mtbr member
    Reputation: ljracer's Avatar
    Join Date
    Nov 2009
    Posts
    293
    Quote Originally Posted by ledoman View Post
    Morten, got one golden colored from Banggood and there are red NCR18650BL inside. I've got the same cells in 2 silver, 1 red and now 1 golden power bank so far. So no rule what you would get. :/

    I've ordered a bunch of the silver power banks. Only opened one up so far. Has the red NCR18650BL inside. The power banks are really nice even with what appears to be thermal protection on the cell. Seems like such a shame to rip them apart but you can't even buy the loose cells for this price.

    I've gotten good at building battery packs now. I have all the parts except the case from Gearbest. Sigh...

    I went with the 7A balancing PCB from Hunk_lee...

    I wish GB would take a cue from Banggood for shipping...

  181. #181
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    What are you using for charging that pack?
    P.S. my last 2 orders from GB shipped in a day

  182. #182
    mtbr member
    Reputation: ljracer's Avatar
    Join Date
    Nov 2009
    Posts
    293
    Quote Originally Posted by Sirius9 View Post
    What are you using for charging that pack?
    P.S. my last 2 orders from GB shipped in a day

    Planning to use my standard MS style chargers. I'm going to reseal it back up. I've decided I can always rip packs apart and rebuild once they no longer perform to satisfaction.

    Mine all shipped right away(at least the email says so)...it's the part after they leave that takes a looooooong time.

  183. #183
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    According to ledomans tests I think you will need charger that gives a slightly more V than 8.4V for that balancing board to work!

  184. #184
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Quote Originally Posted by Sirius9 View Post
    According to ledomans tests I think you will need charger that gives a slightly more V than 8.4V for that balancing board to work!
    How did you arrive at that conclusion?

    Worked well with my magic shine charger which gives 8.43v (measured). Just don't start out with unbalanced cells. See my comments above.

  185. #185
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by ljracer View Post
    I've ordered a bunch of the silver power banks. Only opened one up so far. Has the red NCR18650BL inside. The power banks are really nice even with what appears to be thermal protection on the cell. Seems like such a shame to rip them apart
    It depends... I've ordered it from GB to be used as power bank (because of good reviews for XiaoMi), but received item was faulty right out of the box: it doesn't charge anything unless the charger is connected to it. So, despite I already have plenty of spare cells, now I have 3 more!

    BTW, mine (silver one) was equipped with LG Chem's F1L cells.

  186. #186
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Archie, to bad. Have you opened ticket for it? Just having it for spare cells is not the best. Well for the price it's OK, but I would exchange cells for some salvaged from laptop packs and still have usable powerbank. I did it for two already and they works fine. I've used old Sanyo 2800 which still holds 2200mAh per cell getting 6600mAh bank capable to fully charge average phone battery twice. Good trade off, I think

  187. #187
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    I'm too lazy to bother with tickets etc. for such cheap stuffs, so I just considered I've bought 3 cells in good protective envelope...

    I already have cheap Chinese powerbank (bought as DIY-kit without cells) which is working perfectly, but it lacks pretty things like XiaoMi "authenticity verification", scratch-off label etc...

  188. #188
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    Finally got to disassemble my 2S3P pack, pulling those cells out was A task, charged them with charger from BT40S.

    When charger indicated green I measured 8.33V (after about a hour of resting) but I also measured individual parallel cells at 4.13V and 4.19V so, quite unbalanced at first charge. Will have to discharge separate parallel "banks" individually with my Accucell 6 and then charge whole pack again...
    Name:  tumblr_m8i1nlyQDb1rdpa5go1_500.png
Views: 1019
Size:  173.9 KB
    Not impressed.

  189. #189
    mtbr member
    Reputation:
    Join Date
    Apr 2014
    Posts
    42
    Quote Originally Posted by Sirius9 View Post
    Finally got to disassemble my 2S3P pack, pulling those cells out was A task, charged them with charger from BT40S.

    When charger indicated green I measured 8.33V (after about a hour of resting) but I also measured individual parallel cells at 4.13V and 4.19V so, quite unbalanced at first charge. Will have to discharge separate parallel "banks" individually with my Accucell 6 and then charge whole pack again...
    Name:  tumblr_m8i1nlyQDb1rdpa5go1_500.png
Views: 1019
Size:  173.9 KB
    Not impressed.
    Yeah, the stock cells are Crap. Just so you know, your charger will probably continue charging up to its open circuit voltage. It most likely just indicates green when current is below a threshold.

  190. #190
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by Morten Schmidt View Post
    your charger will probably continue charging up to its open circuit voltage. It most likely just indicates green when current is below a threshold.
    Exactly.

  191. #191
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by Sirius9 View Post
    Will have to discharge separate parallel "banks" individually with my Accucell 6 and then charge whole pack again...
    I believe, you can safely put these cells in the recycle bin without wasting your time. In my case, measured capacity of 2S3P "battery" was ~3300 mAh (i.e., 1000mAh cells used), and in addition it has enormous self-discharge rate: after full charge & 3-hour rest, charge current of 0.3A was registered!

  192. #192
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Archie, that is exactly the same I've measured and published in first post. Same ~1000mAh cells in both 2S2P and 2S3P packs.

    So I agree, it's waste of time to work with them except if you keep them for some limited testing purposes. It's better to use cells from the Xiaomi banks right away. I've made a mistake using those bad cells to test balancig PCM. I should do it with some quality cells.

  193. #193
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    I have a shed without electricity in my backyard, so I planned to change this pack from 2S3P to 6P + PIR + step up to power 12V COB diodes (2-3 pcs) 2A draw will be sufficient

  194. #194
    mtbr member
    Reputation: ljracer's Avatar
    Join Date
    Nov 2009
    Posts
    293
    I finally received my packages from gearbest. Ordered 9/14 delivered 10/5. Slow but at least arriving.

  195. #195
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Yesterday I've converged one pack for my friend just like described in the OP. I've only used stock PCM as he said it might suits him. I've also changed the cable with more soft and flexible one from DX. It was really nice job. I love those welded tripple battey sets from Xiaomi banks. They are beautifuly maded. Replaced one set with old but still very good 2600 Sanyos getting out with 6500 maAh power bank. Keeping second for myself

    I'm really proud this went out very usable and relatively cheap. If you are handy with soldering and have some spare laptop packs you can get out with 10Ah battery pack and 2 working power banks for just as low as $45 (based on current prices) or for $50 with better PCM.

    BTW, all latest Xiaomi power banks from Banggood comes with good LG cells. Still perfecty usable for what we need. Both banks I've used above were exactly the same voltage so no balancing was needed. Just in case I've put them together in paralell for an hour prior to made 2S3P setup.

  196. #196
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    So, was weekend, I had time to discharge and charge again cells that were in 2S3P pack.
    I charged them about a week ago and then left them unused, yesterday I discharged them using Opus BT-C3100 @ 1A

    #1 3 X parallel cells = discharget at 3486mAh, charget again at 1A = 3491mAh, after ~24h resting voltage is 4.14V (charging time 3:57h)

    #2 3 X parallel cells = discharget at 3571mAh, charget again at 1A = 3683mAh, after ~24h resting voltage is 4.15V(charging time 4:14h)

    Entire pack would then be roughly 3500mAh, with xiaomi cells this jumps to about 6700-6750mAh...

  197. #197
    mtbr member
    Reputation:
    Join Date
    Apr 2012
    Posts
    1,810
    Quote Originally Posted by Sirius9 View Post
    Entire pack would then be roughly 3500mAh, with xiaomi cells this jumps to about 6700-6750mAh...
    Yes those cells are about 1000mAh as you can see in the first post. BT-C3100 shows bit high nubers.

    Don't know where you've got with Xiaomi cells would be 6700? They are 3350mAh cells so in 2S3P pack you should get up to 10.000mAh

  198. #198
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by Sirius9 View Post
    Entire pack would then be roughly 3500mAh, with xiaomi cells this jumps to about 6700-6750mAh...
    How is that? With F1L cells, even in "standard" discharge to 3.0 volt the 2S3P capacity would be 9000 mah, while allowed minimum of 2.5 V will give 9900 mAh...

  199. #199
    mtbr member
    Reputation: Sirius9's Avatar
    Join Date
    Apr 2013
    Posts
    166
    I goofed it up, was thinking on making it 2P2S and using saved space for some extra electronic circuits.

  200. #200
    mtbr member
    Reputation:
    Join Date
    Aug 2013
    Posts
    829
    Quote Originally Posted by Sirius9 View Post
    using saved space for some extra electronic circuits.
    Hmm, that's even more interesting! What namely circuit are you planning to add there?

Page 1 of 2 12 LastLast

Similar Threads

  1. Replies: 409
    Last Post: 04-16-2018, 04:44 AM
  2. C&B Seen upgradable 2 cell Helmet battery Case Review
    By Skyraider59 in forum Lights and Night Riding
    Replies: 39
    Last Post: 11-09-2017, 03:04 AM
  3. NEW: 2s2p battery case available from KD
    By manbeer in forum Lights and Night Riding
    Replies: 13
    Last Post: 05-03-2015, 07:27 AM
  4. user review: Xeccon 4 and 6 cell battery test and teardown
    By mattthemuppet in forum Lights and Night Riding
    Replies: 0
    Last Post: 03-22-2013, 09:14 AM
  5. HBFlex/3-up XM-L/2s2p battery - Will it work?
    By natac in forum Lights DIY - Do It Yourself
    Replies: 2
    Last Post: 09-13-2012, 07:26 AM

Members who have read this thread: 4

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

THE SITE

ABOUT MTBR

VISIT US AT

© Copyright 2019 VerticalScope Inc. All rights reserved.