Battery overheating

[GBananas]

Could really use some helpful tips right now. I built a 52v battery and the thing is overheating so bad, to the point that the hot glue actually melted and it was too hot to touch. My first thought is that the amps are maybe too high for the controller, but I'm not sure. I used samsung 30q cells, 2 in parallel and 14 in series. Let me know what you think.

 

[harryS]

How many amps is the controller rating? I have a 52V 14S2P Luna Mini, and they rate that battery for 30A, which is the max rating of two 30Q's. Maybe you need a 3P at least.

 

[GBananas]

It's a 10 amp controller.... Probably should have looked at that before I bought the batteries. lol
Trial and error... Im gonna go frown at my bike for a while

 

[MarkF]

2 in parallel is not enough. Most of our packs are at least 4 in parallel with some going as high as 6P. The controller has nothing to do with your pack getting hot. Your amp draw is too high for only 2P.

 

[GBananas]

Well to get any type of range, no it's not enough, but wouldn't adding more P cells just increase the temperature? 4P would be pulling 60 amps into a 10A controller, seems like things would get heated.

 

[harryS]

WIth twice the cells, you've got more thermal mass to dissipate heat, plus you have half the series resistance, so the self generated heat is cut in half too.

Anyway, the controller design and you limit the current. If it's really a 10A controller, the most it could deliver is 580W (10A x 58 volts), but maybe that 10A is an average rating, and peak is 20A.

Is your pack soldered or spot welded?

 

[GBananas]

It's spot welded. My connections are relatively thin too so that could be the issue. (.15mm nickel)

 

[AguassissiM]

@GBananas
You can make the pack 52v x 50ah if you like, your controller will only take as much as it was designed to just like @harryS said, ah(amp hours) is what gives you distance, volts give you how fast you can go.
How many batteries are there in your pack ?
To connect batteries use nickel strips, to connect the pack to your controller I would recommend using at least 12Ga wire to carry the current safely.
You pack is getting hot because your controller is looking for more current and the pack is trying to supply that demand by dumping all that it can and getting very hot in the process. Just like @MarkF has suggested 2p is not enough to make it work safely.

 

[GBananas]

AguassissiM, there's 28 cells in the pack, they're samsung 30q's which are 15 amps, so in theory the pack is putting out 30 amps. I'm using a 14ga xt60 connector for my discharge wire.

 

[harryS]

You're probably using more. The 15A rating is the manufacturer's spec for maximum draw. A battery designer would size the battery so the typical draw is half of that, like 5-7A per cell. However, the cell is capable of delivering more current than 15A, which will generate heat and shorten its lifetime.

Does your bike's display show power? You might get yourself an inexpensive Tenergy wattmeter, about $10-15 on ebay/amazon. Pay $15 and get it in a week. Pay $10, and you might get it before New Years. That will record the peak amp draw, as well as instantaneous current. Also records AH used, watts, and volts.

 

[GBananas]

Thanks, I'm going to order a display today, I don't have one at all.
Probably going to add 2 more parallel groups as well, it's making more sense to me now lol. That will basically lower the amps drawn from the individual cells, and that will help with the temperature, right?

 

[AguassissiM]

@GBananas

Please do not take this the wrong way but Please do not use that battery as it is, many unpleasant side effects will happen if you push the envelope too far.
You have had the first warning, the pack got hot.
Your well being is much more important than finding out what may happen next.
At this time I would inspect all cells one by one for any kind of damage that the heat may have caused.

You see it's like this, you have a one year old horse and you want him to pull a wagon full of hay, can he do it? Sure he can however he will sweat like crazy. Now get a two year old horse to do it, he will do it with a smile and few spare farts like it was nothing to it.
If this was my project I would not even start it with less than 70 pieces of 18650.
I have been wrong before therefore I do not feel qualified enough to give you proper advise, I am just expressing my humble opinion.
Just my 2cents.

Here is a link to some info that may come handy or this one.

And if you have time watch some of these videos, this guy really knows his stuff.

edited: add info.

 

[MarkF]

My pack is only a 10ah and it has 52 cells. 28 cells is nothing and should not be used to power an ebike.

 

[GBananas]

Thanks for the input guys, I took the whole pack apart today and am assessing the damage. I have a short commute, and I was under the impression that I could build a somewhat small battery pack, as long as the voltage was correct. Not the case apparently lol.

 

[GBananas]

I'm planning another build though, any input is helpful.
I took some measurements, and a 5p/14s pack will fit almost perfectly into a 50cal ammo can. I'm planning on drilling some holes to direct air over the battery, and using the ammo can for a case. With some good wire management and insulation I can definitely fit a controller in there too. Good in theory! Lemme know what you think

 

[AguassissiM]

@GBananas
Have a look here, lots of useful information.
Here is another one.
This guy has played around with batteries as well.

 

[Bruce Arnold]

There's something I'm trying to figure out. The description in the OP is ambiguous. Did you connect two cells in parallel (2p) 14 times, and then put the 14 2p's in series? Or did you connect 14 cells in series, twice, and then connect the two 14s blocks in parallel? Superficially, it shouldn't make a difference. Practically, it does. I don't have sufficient electrical engineering knowledge to explain why. I've never built batteries for ebikes, but do a lot of portable ham radio operation, and I have been taught by my betters to build voltage first, amp hours second.

So, for instance, if we wanted enough juice to run several radios for Field Day, which would be a 12 volt situation, we might employ used 6 V batteries which are cheap and plentiful. Put together two 6V batteries in series to get 12V, then add groups of 2 until there are enough amp hours to operate the whole event.

If that's what you did, great, and all you can do is figure out why there is so much draw. Cells have a discharge rate. If you've got the right voltage (you do), then amp hours only affect how long it can sustain that current. Why does the draw exceed the cells' discharge rate? There's the rub.

 

[GBananas]

I connected them in parallel first. I've read books on E bike battery building and nowhere did it mention that as a concern, in fact that's how it was done in the diagrams. As far as the draw exceeding the discharge rate, I have no idea, the controller is rated for less amps than the battery is capable of producing, so how it's even possible is beyond me. Must be something else going on

 

[bob mutch]

@GBananas

Controller ratings are some times given in Amps Continues and some times in Amps Burst. So if it is a 10A controller then it can do burst 20A typically. If you 14S2P battery pack is still together could you post focused pictures of the top and bottom of the pack please so we can see how you connected the cells. Also are you using a BMS or at least a Battery Protection Board? If so please post a picture of that also.

You don't want to make the same problem again so you want to make sure the problem with the overheating was not some thing other than the fact it was 2P which is 6Ah.


>>>It's spot welded. My connections are relatively thin too so that could be the issue. (.15mm nickel)

It depending on the width of the 0.15mm strips. 0.15mm x 7mm strips are going to over heat at 10A. So if you were running 30A through 1 - 0.15mm x 7mm nickel strips that may be your problem there.

Here are some ratings of nickel strips that M Toll put together.

 

[GBananas]

Thanks for the response, It was 100% my controller drawing too many amps. I added cells to make a 5p x 14s pack and its not even mildly warm after long rides now. Just have to make sure I'm not demanding too much of the cells in the future, and use cell holders because hot glue is not very structurally sound apparently. See my other thread on that.