48V ebike battery pack - 91 cell 14Ah

[Philaphlous]

So here's a pic of my design... I think I wired it right...

My go-to is excel...lol

Explaining my idea. Each color, yellow, green, orange is a parallel series of 7 cells each. So for each bank, you have a total of 21 cells except for the last bank (or first) which is only 7 cells. The cells will be protected inbetween each bank so the connections don't touch and short out the battery. On second thought I may have the last (13) group of cells at the bottom since then I could solder it directly to the next set of cells for 7.4V total, if that makes sense. The number inside each color corresponds with the number of cells in series, for the BMS hookup. I'm not sure if you start the BMS 13 at the lowest total voltage or highest. I think BMS1 should be the 48V and BMS13 should be the first 3.7v bank.

Let me know if you guys see any fatal flaws...literally. ha. Should be getting 27 more cells today or tomorrow. Once I get everything charged up I'm going to discharge test each capacity and group the cells accordingly. I should have my B6 charger tomorrow or Friday.

The lottery is coming!!! I'm super excited to see what cells are in these laptop batteries. lol. Since the battery pack is going to be sooo big, the current draw per cell should be relatively low. This should hopefully keep the explosive power of 91 cells down to a moderate level..lol

Welp... Guess I got what I paid for. The first pack of Chinese cells were TS 18650 cells. They weighed in at 44g and I was unable to find any specs on them online. My best guess is they're 2000mAh cells from China obviously. The next set I found quite easily, 46g cells, FST18650-2000mAh. They were actually printed 2000mAh on them and both are dark blue wrap. Standard discharge current is 2C which would be around 4A which is right at my limit. Hopefully going to get to doing some real world discharge tests this weekend on the capacity. At $1.50/cell it's not a bad deal but its not overly good either with what I got... The second set I actually am not that disappointed about. I can find data sheets on them, they seem to weigh similar to that of a samsung cell and supposedly the capacity isn't bad also.

 

[Philaphlous]

Testing under way!

The cheap Chinese cells are actually ending up right around 1900-2000mAh. Both blue cell types, the TA and FST come in at 1978mAh under a 1A load. The purple Chinese cell is a tad lower around 1890mAh. Max draw on these cells shouldn't be much more than 2A I'd say, maybe 2.5A at times but overall they shouldn't see excessive loads. They're rated upto 4A max draw and 2A continuous load. Now for testing the older brand name LG, Samsung cells!

Apparently the iMax B6 was cutting off at a 2 hour timer. So some LG cells I was testing got cut off right at 2000mAh. They still had a decent voltage left in them when they hit the 2 hour mark. Unfortunately it looks like my Samsung ICR18650-26F cells are bad. Tested 1 and it came out to 234mAh. lol. I'm actually impressed with the Chinese cells. At around $1.50 a pop I tested a few and they get 2050mAh at 1A discharge which would be about 1700-1750 usable mAh when connected to the bike from my calculations.

 

[Philaphlous]

Well it looks like I've got a good mix of cells. Everything seems to be working as normal with the laptop cells. Only "bad" cells I have are some LG cells that top out at around 1500mAh, all others reach at least 1900mAh at 1A draw.

I noticed my fake iMax B6 is 0.2V low on the measurement so I was cutting off my discharge capacity tests at 3.1V instead of 2.9V. From what I can tell, most people say 2.9-3V is a relatively safe discharge voltage to end on.

Is 91 cells a small ebike battery or no? I've seen some threads where people have a stupid amount of cells and I'm not sure now if this is going to be a decent battery for my needs.

 

[Philaphlous]

So the Samsung cells were all bad. They could hold a voltage but when I ran them with the iMax I got anywhere from 100-250mAh before it cut off... fail. I have some Panasonic NCR18650BD cells that take FOREVER to charge but they appear to be in decent condition. I've run through 2 of the 4 cells and I get over 2100mAh on them which is pretty good for their life. I should be getting 12 more cells today which should be the same FST 18650 Chinese cells that give me around 2050mAh and are rated for 2000mAh with a 2C max draw (4A). I also should be receiving my solder wick, flux, and apple green heat shrink for 100 18650 cells. I'm planning on rewrapping every cell so I can double up on the cell protection for my build (no pun here).

I also went through and dremmel'ed the top and bottom of each cell to cut off the nickle weld marks and ruff up the surface in prep for the solder.

At this point I think I'm waiting on maybe 9 or 12 more cells and the BMS which should be here in another week or two. Once I have that, I should be able to start to plan my pack.

 

[Philaphlous]

Few updates:

1. My BMS is in the US now. I believe it will be shipped in the next week or so.
2. I'm still waiting to find out about my job situation, so for now, nothing new is in the works. Till I get a job nothing happens...obviously
3. I bought a knockoff battery from a seller from Austin Texas that apparently made the news for selling counterfit goods. LOL! Fail me. However, good news is I knew I was buying a knockoff. The cells I got are in a red wrap and unmarked. I was surprised to find out they had the best cell capacity of all my cells! 2180mAh and the voltage drop on load was so little. I think it was something like 4.05v @ 1A draw after 100mAh spent... every other cell is around 3.8V or so at 100mAh spent...
4. I believe I have all 91 cells! Woohoo!
5. Once I nail a job I may buy 1 more battery pack and replace a set of cells that only come out to ~1500mAh. Everything else gets 1900+mAh capacity with I'd say the average around 1990mAh
6. I picked up more solder and 12AWG wire for my main leads. 12AWG should be enough I think!
7. I shrink wrapped a few cells to test the fit and wow! The apple green is awesome! The heat shrink really protects the cells and I'm happy I'll have 2 layers of heat shrink on all my cells for added protection from a short!
8. I'm hoping to get my bike here in the next few weeks and I can start measuring things out to make sure my battery idea will fit!

9. I'm thinking I want to add an on/off switch to the battery... is that possible and where should the switch be placed in the wiring? Before/after the BMS? I want it to be a small push button switch, can those handle 1000w?

 

[Philaphlous]

Progress update:

I'm in the process of adding the 2nd layer of heat shrink to the cells. Everything is working well. My lighter ran out so I'm now just waiting to use the hairdryer tomorrow which should be an easy process. Soo many cells! I calculated wrong and I actually have 81 cells right now. I went ahead and ordered another 12 cell battery pack which from my tests appear to be the best of them all. They're 2200mAh cells and currently I'm sitting at 1500mAh on my discharge test and I've got plenty of volts left at 3.5V. Pretty good so far!

Potentially I could have my bike next weekend. I just got my job situation sorted out so I should be moving forward with the bike hub motor in the next few weeks and then I'll be able to put the pack together!

 

[Philaphlous]

Wow! Alot of work today done!

I shrink wrapped all the cells that I have. Unfortunately I only have 79 good cells...so I'm waiting on 12 more to finish up my battery.
I got a clever idea to hookup each of the 13 cells with a deans port to 1, monitor the voltage and 2, do even a quick manual balance...

The glue worked well! Surprised how much I used but the packs are holding together really well. Each pack is 21 cells, 3 sets of parallel cells with 7 in series.

Once I find a proper semi round housing for the cells, I plan to have a hookup for each parallel set of cells and have the deans female connector directly glued to the battery housing so I can monitor all the voltages instead of the BMS balancing...

 

[Philaphlous]

More progress!

The battery is taking shape! I've ordered my 13 deans plugs for battery balancing of each of the 13 banks. That will be an awesome way to keep tabs on each series of cells.
Soldering has been slow but is going good. The soldering iron has worked well with the flux. Takes around 10 seconds for the solder to weld to the cells. The copper solder wick is probably the best choice of material to use for connecting the cells if your soldering....

I'm connecting each parallel connection in series with 4 joins, Basically my idea is I'll draw a max of 4A per cell which each solder wick would carry 1A...but anyway...

I am really impressed with the soldering quality on the BMS. It actually looks like SMT production and not manual labor...

One thing I did notice though that needs redone is the thermal paste... The seating of the heat place isn't perfect and there's quite a gap between the mosfets and aluminum... So I decided to put my 1mm thermal pad to good use!

As you see with the thermal pad, the squish is great! Should provide some adequate heat transfer. This is supposedly a 35A 48V BMS...

I'm waiting on 12 more cells to complete my last pack and then I should be able to build my housing. I'm considering molding lexan plastic sheeting into the desired shape of my housing rather than going with some preconstructed plastic tube...I'm still trying to figure that out...
Each set of 21 cells will be connected via a 12AWG wire. I also have to separate out each set of cells which I will divide using plastic and foam (I know not the most fire safe design)...
I may just end up using PVC pipe since its relatively light, is thick enough and wide enough for the batteries, and can easily hold the deans plugs that I'm planning to use for each series of cells... I'll use spray foam to create a vibration free housing and this should finish up my battery build!

 

[999zip999]

Remember the whole case is neg. I would use a gasket on the pos end.

 

[Philaphlous]

Remember the whole case is neg. I would use a gasket on the pos end.

I've double shrink wrapped every cell. The original cell's shrink wrap but then my own apple green shrink...I figured double shrink wrap is plenty... The copper wick becomes really stiff once soldered so I can't foresee there being any shorting out issues...if so..say hello to my little fire...lol

 

[Philaphlous]

Looks like my puppy got in the shot of my battery!

I've finished up at least hot gluing the cells together! I need to solder the last 2 packs and finish making the series connections on the 3rd pack. Now is the time to go figure out the actual housing of the battery.

 

[flat tire]

Beware of the braid stiffening up from wicking solder along its length as this can exacerbate any tendency for joints to break.

Also, 10 seconds is not ideal, that's a lot of heat! A big soldering iron helps do it quickly, but you might as well take that money and put it toward a capacitative discharge pulse welder.

 

[Philaphlous]

Discharged my last set of cells to test their capacity. Looks like things are working better than intended! These Chinese cells ain't that half bad! Still going with plenty of voltage left to get to 4.2V and I'm surpassing 13Ah!

 

[999zip999]

Testing good for knowledge bad for cells more testing more bad cells. Everything has its limits and you're testing limits. Use don't test I mean get some miles out of the money .

 

[999zip999]

Testing good for knowledge bad for cells more testing more bad cells. Everything has its limits and you're testing limits. Use don't test I mean get some miles out of the money .

 

[Philaphlous]

Ended around 13.8Ah but the iMax cut off at 4.16V and not 4.2V. I bet I'd get 14Ah with a full 4.2V charge!

I finished soldering the packs this morning. Now I just need to connect each pack and figure out a housing for them.

 

[Philaphlous]

Question for you guys...

Do you balance down or balance up? Meaning, do I discharge to a set voltage to balance every cell in series or do I charge them up to a certain voltage? It appears the resting voltage of the cells are slightly off. I'm balancing down to 3.7V on my iMax which they will end up resting at ~3.9-3.92v...and my iMax voltage is off so its probably actually balancing down to something like 3.85v...

I just finished putting the back together yesterday. I still need to wire up the BMS which I'm trying to figure out how it works....

the black sense wire is the last negative cell in series? Then from there the rest of the BMS sense wires go to the positive leads of all 13 cells? I did a quick test on my charge and it appears to work great so excited for that! I still need to wire up all the dean connectors on all the cell series. I have 6 done and have 7 more to go. I had a change in heart about the housing.. Because of its strange shape I don't think a normal round housing is going to be so good.. So essentially at this point I'm planning on wrapping the cells in duct/electrical tape and then likely carrying the battery in a cloth bag attached to the bike...

 

[Philaphlous]

Need a little BMS wiring help guys!!!!


Looks like all I can find on my BMS is a B-, C-, and B+ then the 13 sense wires.... I take it the B- wires directly to the negative end of the battery, the C- is charging negative, the B+ is battery + or charging +?

 

[Philaphlous]

Wow! It's been what seems quite awhile!

The battery is completed except for a little more hot glue, deans connector for the power out for the bike, and some more duct/electrical tape to seal up the battery! I have the battery fully charged and each series is around 4.19V. According to my multimeter my output voltage is sitting at 54.0V. The battery weighs just under 10lbs and it should fit really nice on top or straddle underneath the frame of my bike once I get it. I confirmed after charging that the BMS does in-fact work! I really think this is going to be one awesome battery! I should be easily getting at least 12Ah out of the battery if not more like 14Ah.

Next step is to actually get my bike and then the hub motor kit. That may take awhile since I'm battling some stuff with medical insurance and I'm also starting a new job here after memorial day. I got the sorta OK from the wifey for biking to work on occasion once we find a safe route!

 

[Philaphlous]

So quick update! The battery is holding up well! I have over 50 miles on the bike and the battery easily pulls around 1400w no problem. I'm running into a charging issue...I believe it has to do with my charger but I'm not sure. When I'm charging the bms will cut off early... Meaning after it turns off I'll end up with a charge voltage of around 54.0-54.2v. I checked each of the 13 series and none of them have topped off at or over 4.2v so I'm not sure why it's cutting off the charger. One theory I have is that I'm charging the batteries too fast, I've got a 2a charger and it stays 2 amps till cutoff so the voltage while charging will easily hit 54.6 then after it cuts off settles around 54v. Any possible fixes?

 

[Philaphlous]

I'm wondering if it's possible to manually regulate the input current going to the batteries from the charger... That would be ideal for me.. The charger charges around 2a and once it hits say 52v I'd like to step it down to 1a then.5a as it gets closer and closer to 54.6v...basically this isn't an intelligent charger at all...


Can I use something like a 48v speed controller that uses pwm to change the current?
https://www.ebay.com/itm/263019014310

I could wire it in line with the charger and dial back the current as I get closer to 54.6v...the bms will still cutoff at 54.6 but at least I'd be regulating the current so I can reach that voltage...

 

[999zip999]

Your one group of cells that hit 4.2v first has lower capacity and is filling up faster. You can put a car brake light and drain it the high group. That should reset the bms and restart the charger to fill up the other higher capacity lower resistance cells. Don't try and find the bottom of your pack just use 12ah and be happy. Testing will unbalance your pack making hard on the bms to balance. Have fun ride bike. Don't test. Or test for cycles not ah

 

[Philaphlous]

Your one group of cells that hit 4.2v first has lower capacity and is filling up faster. You can put a car brake light and drain it the high group. That should reset the bms and restart the charger to fill up the other higher capacity lower resistance cells. Don't try and find the bottom of your pack just use 12ah and be happy. Testing will unbalance your pack making hard on the bms to balance. Have fun ride bike. Don't test. Or test for cycles not ah

So strange thing. I don't think the highest pack is tripping the bms. I found the series that is tripping the bms and I used my iMax b6 to load it to 1a which sags the voltage and I was able to get the charger working again for a little bit. I'm thinking I might need to spend some time manually charging each series to 4.2v... Thatd take forever..

 

[999zip999]

Each parallel group. Charge each group to 4.2v. The bms is not for balancing the pack. But to keep a balance pack balance. You could have a weak cell or two in a parallel group.

 

[Philaphlous]

I think I may have figured out a cheaper way of regulating the current...maybe some people here will know. I'm considering buying another cheap 48v controller and using that to essentially control the current of the charger. My understanding is a brushless controller keeps a constant output voltage but regulates down the amount of current being supplied to the motor...this is the control function of the controller. I'm hoping to wire this up in line with the charger into the battery with a "throttle" for the charger...so as it gets closer to the full 54.6v I can trickle charge via the controller.