First time battery builder critique

DIY Builds Battery Builds
#19

Thanks for the feedback Common. Please read other replies re: soldering.

For the fuse, swapping to a slow blow may also be a good idea. I just don’t know if there will be a real difference since while riding you will not be able to tell if your slow blow is, well, blowing until you lose power. I am also OK with destroying electronics to save life and limb. My main goal with the 150A main fuse was to be so far over what I am planning to pull from the pack that during normal course of operation there should be no way to approach that current limit. I wanted the 150A to protect the battery (mainly for storage/charging) so that if there is a short downstream that starts pulling very high current from the pack then it will blow that fuse to prevent the pack catching on fire in my house.

1 Like
#20

Unfortunately, there’s no way to know. All we can really do is guess. @evilrman will have to decide on his own.

2 Likes
#21

I appreciate the tags, thanks! I do not believe that the heating is an issue with the speed that the soldering takes place. I have a fairly hot iron, so even those huge joints are <10s of heat. I found it also helps Under those nickel strips is a layer of Kapton and fish paper to aid as well. I touched the nickel strip on the positive battery terminals immediately after the joint solidified and was not burned, so I am fairly happy in that regard. Also helpful is that the bent tabs do not actually lay directly on top of the taped p packs, there is probably a 1mm air gap that is difficult to get the metal to bend down into. I would feel comfortable re-soldering that joint if consensus is reached. Thank you!

1 Like
#22

Looks pretty nice, good job

For me if the joint still passes pull test, isn’t lifting, and isn’t too crunchy, I’d just leave it

2 Likes
#23

Thanks Mooch, please also see reply to xsynatic re:heat. I believe the soldering to be relatively unobtrusive to the cells due to amount of insulation and speed of soldering gauged by touching metal parts close to the solder pool immediately after. I would also recommend the cold plier technique to everyone, it helped speed up the soldering of the joint since I had to do less messing around with the iron itself and cooled the pool extremely fast due to heat soak into the cold tool.

1 Like
#24

What @BillGordon and @Battery_Mooch said. I always use rings on the positive side. Keeps the plastic ones that come with the cells from melting into the nickel when you weld. I don’t have pics, but it will do this. Its not surpsing that plastic might melt with really hot metal, even if the heat is as incredibly brief as welding.

2 Likes
#25

Sorry for being a little late to the party, for a first battery this is spectacular! The steps you have listed sound right for the most part, the only things I noticed were missing were cell terminal insulation rings, and end caps on all the parallel groups. From a visual stand point it looks like cold welds with a dirty surface or too little flux, it should be wicked out over a much larger area of nickel to be considered high current rated (also worried about the fuse and XT solder joints too). If you had any pictures of the welds themselves that would be important as you can have a lot of welds but none of them stick if your settings and technique aren’t great. From a material stand point you did good research, it’s a big higher current rated then what a lot of people do

2 Likes
#26

Looks way better than my first battery.

I’ll add some stuff that I’ve come to notice as time goes on.

Overkill of insulation and current capabilities is good. It’s usually not great in an eskate to try to skirt the line of adequacy so closely.

That being said, it is possible to overkill your overkill, as it were.

Using larger gauge wires than your soldering iron or skills can handle will be a detriment to a build, sometimes more so than using a smaller gauge.

Also, you have to always be cognizant of space constraints and what impact they have on the battery. Thick gauge wire can become problematic if they sit too proud and you end up crushing the pack when you tighten the bolts on your enclosure.

The battery looks well built, aside from the soldering commentary already made.

I’d just say that it’s important to take the context of the entire build into account.

Just as well, I don’t think it’s realistic to expect to do that kind of current draw from a 4P of those cells.

They’re great cells, but they’re not magic. They will sag near their rated max. And unless you’re constantly racing up significant inclines, I don’t think you’ll actually need to accommodate that kind of draw.

4 Likes
#27

If your concern is charging/storage, which I think is common and valid, then fuse your charge port. It’s unlikely that a short would happen downstream when your board is sitting idle in your house. This is why many people choose not to fuse their main terminals. Many, including myself feel that the risks outweigh the benefits.

2 Likes
#28

Sorry for the delay everyone, got locked out for 12 hours due to too many posts.

Appreciate the reply Zach. I did test welds with the welder and increased the dwell time until I was not able to remove a nickel strip without destroying the nickel. When removed part of the nickel stayed on the cell. Please see below for a picture of welds.

Weld -meta
Weld -meta3409×2228 1.2 MB

2 Likes
#29

Much appreciated for the analysis Mario. I tried to overspec everything for my intended draw. I will likely be limiting to 90A draw (45A per motor) which is 1/2 of the rated continuous draw of the 4P P42As. The enclosure that I am using from eboards peru will have more than enough height to accommodate the pack with proud wires, however the next pack I build I will likely be more cognizant of vertical space.

I also tried to take into account the difficulty of soldering large gauge wire by using two smaller jumpers between each p pack VS one large wire. In a next build I may even decrease to 3 even smaller gauge wires.

#30

Thanks common. I planned to fuse the charge port also, but with all the analysis around the larger fuse, I am considering removing it to directly wire to the VESC. Any thoughts on a XT90 key to disconnect the battery during storage and charging?

2 Likes
#31

Thank you Longhair! It did not make sense to me at the time to add additional insulation rings, however as that seemed to be the common knowledge I did it anyway and am now very happy with the move given the feedback received.

2 Likes
#32

They appear to be a little hot, you might be able to reduce the pulse time a little bit (note that the lower the voltage of the welder the longer you need the pulse if you’re using a battery operated setup). But as long as they pass the pull test they are okay, if you want to have slightly better robustness try to stagger the welds because the positive electrode is weaker

For /positive/ side of cell
x o
o x
x o

For /negative/ side of cell (avoid the center or it might damage the cell)
x o
o x
x o

1 Like
#33

Thanks Shutter, all the contacts are certainly stronger than I can pull even with pliers. I made sure to rough the contact surface with sandpaper prior to all soldering and fluxed all pads and wires beforehand. I think my detriment in soldering is trying to minimize iron contact time to decrease the amount of heat going into the cells below the nickel tabs.

1 Like
#34

XT90s loop keys are both the most reliable and cheapest way to safely disconnect a battery, realistically you do not need a fuse

2 Likes
#35

+1 on the loopkey. They’re cheap, reliable amd easy to make. I put mine in the circuit where your 150 amp fuse currently is. It’s also a good way to break the circuit if you’re on a ride and something goes sideways.

1 Like
#36

Oh I thought it was 3mm from the center of the negative side? Not very certain of that though

Molicel’s stated life of ~500 cycles is given at 4.2A charging current per cell, and the data sheet also allows for up to 8.4A. So while 4A could be perfectly fine for speed and probably slightly better for cell life I’d personally go a bit higher to get as much range as possible in an hour or two of charging

2 Likes
#37

This is correct.

This is not really correct. Mooch has performed testing on these cells and rated them at more like 25 - 30A continuous for best performance. You can read more about the nuances of p42a here → Molicel P42A cell discussion

I find with cell ratings its always better to keep your maximum possible continuous current draw significantly lower than your worst case battery ability.

By this approach, you could say your battery is only capable of 100A and running at 90A is quite close to that threshold. Do not be surprised if you see noticeable sag occurring while running at that power level.

All that said, nice build! Far better than my first!

5 Likes
#38

They appear to be a little hot, you might be able to reduce the pulse time a little bit (note that the lower the voltage of the welder the longer you need the pulse if you’re using a battery operated setup). But as long as they pass the pull test they are okay, if you want to have slightly better robustness try to stagger the welds because the positive electrode is weaker

Zach, great pearls of wisdom here. I was using a 12V 640CCA battery, so higher voltage than many of the spot welding setups that are used. Alternating electrode placement for welds is a great suggestion I had not thought of. I was changing electrode placement from left to right hands, but purely to keep the wear on the welding tips somewhat consistent. I did not alternate them every weld which I will do in the future.

XT90s loop keys are both the most reliable and cheapest way to safely disconnect a battery, realistically you do not need a fuse

+1 on the loopkey. They’re cheap, reliable amd easy to make. I put mine in the circuit where your 150 amp fuse currently is.

Thank you Zach and Common, I think this is the route I am going to go, remove the fuse and replace with an XT90 loop key. It will accomplish my battery isolation goal while it is unattended and remove the failure point during use.

Molicel’s stated life of ~500 cycles is given at 4.2A charging current per cell, and the data sheet also allows for up to 8.4A. So while 4A could be perfectly fine for speed and probably slightly better for cell life I’d personally go a bit higher to get as much range as possible in an hour or two of charging

Mr. Shiteside, This is something I am considering as well as I have not yet ordered my charger. The longer charge times do not bother me much as I typically only charge once a day when I get home, so I have a solid 6 hours before bed where it can be charged while watched. The utility of a faster charger is certainly there, just not sure if I really need it, and if I can get a little bit better pack life, then all the better.

This is not really correct. Mooch has performed testing on these cells and rated them at more like 25 - 30A continuous for best performance. You can read more about the nuances of p42a here → Molicel P42A cell discussion

Hi Glyphiks, This is great info, thanks. I am not really that much of a speed demon on boards, my main goal for this board is to be able to go up a 20% incline at 20mph and have a long range which I think is doable far under 90A of draw. I can go up the same inclines at 13mph with 950W of hub motor power, so even 40A total draw is double that. I will definitely be keeping an eye on voltage sag to determine if I should decrease the current draw.

3 Likes