• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

3D printed battery bracket for EVE 280-320AH prismatic cells

Anx2k

Solar Enthusiast
Joined
Nov 11, 2023
Messages
78
Location
Phoenix, AZ
I've been working on a 3D printed battery bracket for EVE 280-320AH prismatic cells - I had originally been using some of the Chinese ones that I had sourced on Alibaba, but the more I worked with them the less happy I was with them, so decided to give it a shot with my handy Bambu X1C. I originally printed them with PLA, but then switched it up to PETG, although either should work fine so you can pick whichever you're more comfortable with. In terms of material costs, depending on how much you pay, each bracket ends up being about $1 of material - which was about what I was getting the Chinese ABS brackets for (in quantity), but that doesn't include shipping, which makes these considerably cheaper. That $1 price assumes you're paying around $16 for a 1kg spool, and I've been looking to do a larger Alibaba PETG order, so if you were printing a bunch I could easily see that being 1/2 that (keep in mind this is just raw material costs, not time, etc. Each one takes about 1.5 hours to print with my tuned X1C profile, and it takes a total of 16x of these to do a 24v pack or 32x for a 48v pack.

Now, why I did I end up redesigning this? It really came down to some ways I felt like the ABS design could be improved - the first area was how they interlocked. On the ABS one, they had little nubs that clicked together, but I had these break before, which makes the bracket useless. For my design, I replaced this with a 2mm x 16mm stainless steel dowel (100x cost about $7 on Amazon), and it uses one on each corner to lock them together. There's 8mm inserted on each side, so they really hold things together well. It's a loose friction fit, as when it's assembled it's meant to be held together with fiber reinforced strapping tape (as was the original bracket).

Next, the ABS design had one end plate, and a different middle plate - while not a big deal, it was a pain when you ran out of ends but still had middles - so with my design, there's only a single one, and they just connect either back to back or front to front. This also is advantageous on the 3D printing side, as I didn't want to use any support material, and this way everything works and the overhangs (for cutouts) are small enough to just bridge.

While the ABS one did somewhat lock horizontally together, it was kind of poor (in my experience), and also made the overall pack size slightly larger for no real reason. One mine, I have alternating cutouts on all corners, top and bottom, so that each bracket locks in with the one next to it with two contact points - one going into the neighboring part, one going in from the neighboring part. This makes a much stronger side connection than I had before.

The ABS used fiber reinforced strapping tape to hold it together once assembled, and I actually quite like this so I stole that design element. ;) I go around the pack twice so the tape locks to itself, and do this at the top and bottom. It's not really meant to hold the pack together, but more just to ensure it doesn't fall apart. You'd still want to have it either resting on something or mechanically constrained in some manner.

Finally, there was really no rhyme or reason I could figure out for how they put some of their supports or bracket spots, and most of the things I build are using 80/20, so this is designed to work with their 15 series extrusion. This means that it will 'lock' into the channel in the extrusion, both helping to really keep packs together - on both the bottom and the top. On the top I added extra blocking and thicker ones so that it makes extra contact with the enclosure, since this is on the top and between the cell posts, and obviously you wouldn't want the 80/20 to get anywhere near the posts. The bracket works fine without 80/20 though, so don't think you need to use it - it just has features that are nice is you do.

Anyway, here's some photos to get an idea of how it all looks, and I'm just waiting on it being approved on the 3D printing side for the posted STL file, and I'll post that when they do. You should be able to print it with pretty much any 3D printer, as long as you have a print area large enough for it - but most of the normal 3D printers around do (not the mini's).

Bracket3D-1.jpgBracket3D-2.jpgBracket3D-3.jpgBracket3D-4.jpgBracket3D-24v.jpgBracket3D-48v.jpg
 
Don't take it as criticism, but I really don't see the point of these. the plastic is not going help with compression, does this just give a place for the strut to go?

I feel like just wrapping the entire thing in kapcon tape would actually be stronger and provide more compression.

Let me be clear I own multiple printers.
 
And the cutout on the face of the cell doesn't provide pressure where it needs to be. At a minimum the face should be solid.
 
These really aren't designed for compression - it's more to be able to mechanically lock cells into a position - and that's what the 80/20 is used for. Here's the the kinds of brackets that we were using before - these aren't the exact ones, but they're all basically the same, and also not designed for compression:
We buy cells directly from EVE, and they didn't have an issues with warranty replacement on uncompressed cells, so that was about as far as I looked into it.

Now, if someone wants me to change the design and make one that would be more suited from compression, I'm happy to do it - I could add more space where the dowel goes, so the two plastic parts don't come into direct contact, and and change the cutouts however people want, and give a little more tolerance on the side locks do compensate for the play that would come from assembly pre and post compression. So if you guys want that, just sketch up roughly how you'd like the cutout to look, and if there's any other tweaks you'd like...
 
For compression, you need outside pieces that has a channel to hold the 80/20. Then rods can pull on the 80/20 for compression.
 
If I were designing to do compression, I would actually just tap the center of the 80/20 and use screws into that - and just make the length of the 80/20 slightly less than the overall length of the battery. Then you could have plates on either side (that would be rigid enough to not deform), and screw them into the extrusion.

I guess my question is, should the cutouts be removed in that case altogether? So just to act as another layer inbetween the batteries so they can't come in direct contact, similar to the FR4/foam they put between batteries in those rack mount cases?
 
If I were designing to do compression, I would actually just tap the center of the 80/20 and use screws into that - and just make the length of the 80/20 slightly less than the overall length of the battery. Then you could have plates on either side (that would be rigid enough to not deform), and screw them into the extrusion.

I guess my question is, should the cutouts be removed in that case altogether? So just to act as another layer inbetween the batteries so they can't come in direct contact, similar to the FR4/foam they put between batteries in those rack mount cases?
Yes, remove all of the cutouts so the cell faces can touch them everywhere.
 
Compress1.jpg
Yes, remove all of the cutouts so the cell faces can touch them everywhere.
How about this? I've removed the cutouts, and also reduced the edge by 2mm, so they won't come in contact before the cell, but still use the dowels to keep everything locked together. The thickness of the back is 2mm right now which I think might be kind of thick in this case, so I could reduce it to 1mm (as it comes in contact back-to-back anyway, and one the ends you'd want something not plastic to give uniform pressure over the side of the battery I would assume). Thoughts?
 
I picked up similar plastic fixtures for a 4s that I’m playing with.

Any reason you’re doing just “end” pieces? The Ali units have middle pieces that hold two cells on either side. It’d cut your part # down.

So for an 8s straight build, 6 middles and 2 ends, vs. your 16 count.
 
View attachment 250063

How about this? I've removed the cutouts, and also reduced the edge by 2mm, so they won't come in contact before the cell, but still use the dowels to keep everything locked together. The thickness of the back is 2mm right now which I think might be kind of thick in this case, so I could reduce it to 1mm (as it comes in contact back-to-back anyway, and one the ends you'd want something not plastic to give uniform pressure over the side of the battery I would assume). Thoughts?
That is much better. I think 2mm is way too thick. The sheet just needs to help provide electrical insulation to ensure that the cell cases cannot touch each other. Additional thickness just adds width to the battery frame. The FR4 fiberglass sheets work great and are thin. My frame is as short as it can be since I wanted to stuff two batteries into an aluminum tool box.

PXL_20230715_201502050.jpg
 
Here is a photo of a 4S battery I built using the same frame type. Notice that the frame is fixed length. One sheet of 1/4" Poron foam per four cells provides compression. For this one the BMS could ride on the side.

PXL_20220621_213623012.jpg
 
I picked up similar plastic fixtures for a 4s that I’m playing with.

Any reason you’re doing just “end” pieces? The Ali units have middle pieces that hold two cells on either side. It’d cut your part # down.

So for an 8s straight build, 6 middles and 2 ends, vs. your 16 count.
I suspect they're just the same as the one I linked to above (they all seem to be very similar). The reason I didn't model an end and a middle part is due to the nature of 3D printing. Strength-wise, for the way the battery is held you'll have better performance laying flat than standing upright (as the layer adhesion is a common weak spot), so by lying down the layer lines go around the battery - this means it's more likely to sheer between batteries, but the 8020 protects from that (or if you had it sitting on something). Also if you printed the middle in any orientation, you'll need to use support material for the overhangs, etc - so this ends up just adding to the cost and is just material you toss in the trash. For them it makes sense as it's injection molded, so there's no waste, but as an FDM print, you have to consider supporting it as each layer is added. I should also point out that the price was basically the same between the two - so if you buy the 2 ends and 6 middles, that cost is the same as 16 'ends' that make up my design, assuming you're paying around $16 for a spool of material. And it's also faster doing it this way, but mainly just because it's not spending time doing support - and I also like that I have just a bunch of the same part, so I'll never run out of 'ends', etc...
 
That is much better. I think 2mm is way too thick. The sheet just needs to help provide electrical insulation to ensure that the cell cases cannot touch each other. Additional thickness just adds width to the battery frame. The FR4 fiberglass sheets work great and are thin. My frame is as short as it can be since I wanted to stuff two batteries into an aluminum tool box.

View attachment 250091
Cool, I can easily drop it down to 1mm or even 0.5, I'll see how the print feels at both thicknesses - I'll print a couple today to see how they look...
 
Cool, I can easily drop it down to 1mm or even 0.5, I'll see how the print feels at both thicknesses - I'll print a couple today to see how they look...
.5 on 3d printing is going to be fragile, i generally don't go under 2mm unless using a pa GF CF type filament.
 
I'm not sure I see a value in this for me -- but I do get that it turns a 4-cell pack into an easily managed unit. In my case, I'll have a solid box with close tolerances -- once installed, the cells won't move.
BUT, I have a suggestion. In my case (and in your "real world" assembly), I am terrified of a conductive object falling across terminals at different voltages -- the first cell in your assembly is adjacent to the 8th cell, 24 volts higher (maybe 18V? -- depends on wiring) with bare metal only 3" apart. Consider a wrench, or pliers, or possibly a screwdriver (the screwdriver handle may make that impossible). I'm considering separating terminals at different voltages (ie, between one 4-cell pack to the next in your example) with a sheet of "something" that not only provides insulation between cells, but projects higher than the busbars on the cells. You could still have a short with something with a shape (a socket wrench with a socket installed) but it would be virtually impossible to get a short with a straight object like a box end wrench. If you extended your top corner piece (where the dowels are) up another bit, say an inch, that would eliminate this risk and still be no higher than your 80/20 bar.

You could even do it on the part between adjacent cells. The risk is lower here (7V typically), and it would increase your number of part styles. But consider cells 1 and 2. At one end, the terminals are jumpered (0 volts) but at the other end they are so close and 7V different in potential. Unfortunately, you would then need a L and R version. Or you could make them all with the partition, and "simply" cut off what gets in the way of the bus bar crossover.
 
You need 1 sheet per surface to stay in linear compression range or 5 sheets total. Right now your middle cells will apply force on their terminals.
No. One 1/4" sheet provides the correct compression strength for four cells. Adding more sheets reduces the compression force. I use flexible bus bars, so I doubt I have much force on the terminals.
 
No. One 1/4" sheet provides the correct compression strength for four cells.
Nope that's wrong. Each cell expands 1mm or 4mm for all. 1/4" foam needs to be compressed 15-25% to get initial 11psi force. The window to maintain 11psi is small so more sheets you add in series the better (up to a point of diminishing returns). Pull up a datasheet for your foam and see for yourself. If you gonna do compression at least do it right.
 
I suspect they're just the same as the one I linked to above (they all seem to be very similar). The reason I didn't model an end and a middle part is due to the nature of 3D printing. Strength-wise, for the way the battery is held you'll have better performance laying flat than standing upright (as the layer adhesion is a common weak spot), so by lying down the layer lines go around the battery - this means it's more likely to sheer between batteries, but the 8020 protects from that (or if you had it sitting on something). Also if you printed the middle in any orientation, you'll need to use support material for the overhangs, etc - so this ends up just adding to the cost and is just material you toss in the trash. For them it makes sense as it's injection molded, so there's no waste, but as an FDM print, you have to consider supporting it as each layer is added. I should also point out that the price was basically the same between the two - so if you buy the 2 ends and 6 middles, that cost is the same as 16 'ends' that make up my design, assuming you're paying around $16 for a spool of material. And it's also faster doing it this way, but mainly just because it's not spending time doing support - and I also like that I have just a bunch of the same part, so I'll never run out of 'ends', etc...

Yeah seem to be all about the same.

Either way I was able to get 5 sheets of FR in the “case”. This is going to be just fun battery I’m placing in a Milwaukee box, so I won’t be too concerned about compression.

Ahh the issues of 3d printed stuff, always have a challenge.
 
Last edited:
Nope that's wrong. Each cell expands 1mm or 4mm for all. 1/4" foam needs to be compressed 15-25% to get initial 11psi force. The window to maintain 11psi is small so more sheets you add in series the better (up to a point of diminishing returns). Pull up a datasheet for your foam and see for yourself. If you gonna do compression at least do it right.
The other thing with compliant foam between each cell is you avoid any pressure hotspots from surface irregularities in the aluminium can.
 
I'm not sure I see a value in this for me -- but I do get that it turns a 4-cell pack into an easily managed unit. In my case, I'll have a solid box with close tolerances -- once installed, the cells won't move.
BUT, I have a suggestion. In my case (and in your "real world" assembly), I am terrified of a conductive object falling across terminals at different voltages -- the first cell in your assembly is adjacent to the 8th cell, 24 volts higher (maybe 18V? -- depends on wiring) with bare metal only 3" apart. Consider a wrench, or pliers, or possibly a screwdriver (the screwdriver handle may make that impossible). I'm considering separating terminals at different voltages (ie, between one 4-cell pack to the next in your example) with a sheet of "something" that not only provides insulation between cells, but projects higher than the busbars on the cells. You could still have a short with something with a shape (a socket wrench with a socket installed) but it would be virtually impossible to get a short with a straight object like a box end wrench. If you extended your top corner piece (where the dowels are) up another bit, say an inch, that would eliminate this risk and still be no higher than your 80/20 bar.

You could even do it on the part between adjacent cells. The risk is lower here (7V typically), and it would increase your number of part styles. But consider cells 1 and 2. At one end, the terminals are jumpered (0 volts) but at the other end they are so close and 7V different in potential. Unfortunately, you would then need a L and R version. Or you could make them all with the partition, and "simply" cut off what gets in the way of the bus bar crossover.
My plan was always to have some sort of top protection built into it - although I haven't done it yet, as our bus boards haven't arrived yet, my plan was actually to 3D print a kind of 'shield' that slots into the 80/20, so it covers up the terminals. It would be cool if it were possible with sheet goods, but with the newer terminals plus the bus boards it puts them higher than the extrusion slot, so it's more like an arced piece that slides into place. As I said, I haven't actually done it yet, but that's my plan, and that way it would fully protect the top of the cells from anything falling in.

Another option might be to actually make the top part thicker, so that it pushes up the 80/20 above the terminal connectors - but ideally I'd like it to work with both studs and those new ones, and I don't know if that really is feasible with studs. I'm also not as thrilled with having too much of the support being from the plastic and not the aluminum, right now the plastic really more ensures things stay locked in position and its the aluminum extrusion doing the heavy work.
 

diy solar

diy solar
Back
Top