New DIY battery. Feedback on design needed.

[Luk88]

I'm slowly building up another battery and I thought to post the plan in hope of getting some feedback. Can anyone see potential issues etc.

There will be 3 batteries in parallel, but let's focus on just one. It is a 16s configuration of EVE MB31 cells sitting on top of a thick fr4 like isolating sheet on a shelf.

This battery is using compression as all my batteries are. Cells will be sitting in "normal" position. Fr4 separators will be on all sides. I'm still debating if I should use neoprene in addition to fr4 (my neoprene compresses by about half at 12psi, then resists a lot more).

Pack compression will be provided by 10mm thick steel plates, 6mm threaded rods and disc springs.

Cells will be connected in a "snake" configuration with DIY non flexible bus bars.Here is a drawing from the top.

I've already cut the steel plates and I drilled the holes for the rods. The reason why we have 8 cells two rods then 8 cells again is very limited side to side clearance.

I already have heltec 200A bmses for these.

Any comments? Ideas? WTFs?

 

[740GLE]

Is the steel plate braced at all? Only weakness I see if the inner cell corners furthest from the threaded rod will cause the steel to bend.

But I guess 10mm thick is pretty darn robust.

Also why use non flexible bus bars when there are so many good flexible bus bars these days?

 

[Luk88]

Is the steel plate braced at all? Only weakness I see if the inner cell corners furthest from the threaded rod will cause the steel to bend.

But I guess 10mm thick is pretty darn robust.

Not braced because the set of shelves I planned to use has a protrusion in the back which would interfere. So I used an android app called BeamDesign to find out 10mm x 200mm mild steel plate is more than enough. The max deflection would be 0.6mm if the force was concentrated. With it spread out it will be a tenth of mm or two.

Also why use non flexible bus bars when there are so many good flexible bus bars these days?

I already happen to have copper bar to make them so I can build the battery this weekend and I though it doesn't matter with the "snake" connection method. (Each cell is connected to its neighbor via the short edge - like in my picture. The elipses symbolise bus bars.)

In the end the original shelving unit I wanted to use had so little clearance maintaining these batteries would be a real pain. So I decided to upgrade to bigger shelves, but I already cut the steel so the overall shape is likely to remain.

I could change to two rows of 8 cells instead of 4 rows of 4. Wasting a little steel is not a big deal. But thinking about it, it seems the original idea(4 rows of 4 connected sideways) is beneficial.

I haven't decided if I should use neoprene or not yet. As the cells are connected side to side some slight movement will not cause problems. And some give in addition to springs could be really handy if one or two cells start swelling more than others. My neoprene is far from something like PORON foam, but I measured it at 12psi from 30% to 60% compression. So it seems good enough.

 

[chrisski]

I think your plan will work.

I'll make comments based on my builds.

-Recommend one thing as a spacer. Extra adds up especially if you are hurting for space.
-1 mm Neoprene works as a spacer. I pulled some off after being in a build for three years and it was fine. I had originally planned to replace the neoprene with a plastic flexible cutting board, but the neoprene was like new and reused it.
-I prefer a thin plastic like flexible cutting boards or the fiberglass spacers that come with the cells from China. I think this is your FR4 spacer.

Finally, the last build I did fusing and BMS took up more space than I expected.

 

[Bananaman321]

I'm slowly building up another battery and I thought to post the plan in hope of getting some feedback. Can anyone see potential issues etc.

There will be 3 batteries in parallel, but let's focus on just one. It is a 16s configuration of EVE MB31 cells sitting on top of a thick fr4 like isolating sheet on a shelf.

This battery is using compression as all my batteries are. Cells will be sitting in "normal" position. Fr4 separators will be on all sides. I'm still debating if I should use neoprene in addition to fr4 (my neoprene compresses by about half at 12psi, then resists a lot more).

Pack compression will be provided by 10mm thick steel plates, 6mm threaded rods and disc springs.

Cells will be connected in a "snake" configuration with DIY non flexible bus bars.Here is a drawing from the top.
View attachment 229710
I've already cut the steel plates and I drilled the holes for the rods. The reason why we have 8 cells two rods then 8 cells again is very limited side to side clearance.

I already have heltec 200A bmses for these.

Any comments? Ideas? WTFs?

I recommend downloading the product specification sheet for your cells lot of useful information there.

Also your design does not apply equal compression these cells will expand and contract throughout their SOC cycling However you may be able to litigate this with extra foam at the ends of the cells.

Also recommend eva-foam or a similar substance in between cells as they do not expand equally throughout their surface area. This also litigates the distance between the cells as to not put pressure on the terminal post when using solid bus bars.

This is my opinion but it is also a point that has been touched on by Andy's off_grid_garage YouTube channel in review of many DIY battery boxes.

Anyways hope this helps look forward to seeing the finished product. I will be starting mine soon waiting on cells from China.

 

[Luk88]

I recommend downloading the product specification sheet for your cells lot of useful information there.

It has been my frequent reading for last number of weeks where do you think I got the steel plate idea from...

Thank you for replying

Also your design does not apply equal compression

Can you elaborate a bit on this? What do you mean it doesn't apply equal compression?

The only thing that may affect the compression strength for individual cells I can think of is maybe friction, but that is absolutely tiny at the very small movement amounts and relatively high forces.

Perhaps what you mean is that ,lets say cells on the left swell more than cells on the right and therefore through the centre "pivot" point they apply too much pressure on the right... I'm not sure there is any design that mitigates this fully (other than using some constant pressure elastomer between cells). Springs help no doubt. Imagine you have a single stack of any number of cells. There is always a possibility of one in a middle swelling more and applying more pressure to its neighbors.

these cells will expand and contract throughout their SOC cycling However you may be able to litigate this with extra foam at the ends of the cells.

That's what (disc)springs are for in my current design. I may also add some neoprene.

Also recommend eva-foam or a similar substance in between cells as they do not expand equally throughout their surface area.

This is what neoprene is for in the design, but i agree eva foam is definitely better if one can get the right type. However, I'm of two minds on the neoprene question. I think I'll just add some and see how it feels in place. I'll make a decision then.

I'm not sure it matters in a stationary application that 4 years down the line cells will touch bulge-to-bulge vs entire surface to entire surface. One might say, but this way areas surrounding the bulge will not be under pressure, but I don't think so. Eve gives some ridiculously high number as the bulging pressure (two or 3 orders of magnitude more than recommended). So it becomes very important not to immobilise the pack and have these forces act on neighboring cells. Springs take care of it and the bulge itself transfers remaining pressure to the rest of the cell.

This also litigates the distance between the cells as to not put pressure on the terminal post when using solid bus bars.

Perhaps it was missed or not obvious, but the snake configuration (cells connected short edge to short edge) mitigate the solid busbar issue during cell expansion. It translates stretching/compressing motion to very slight rotation of the busbar on the post. This tiny rotation is well within the elasticity of the post.

This is my opinion but it is also a point that has been touched on by Andy's off_grid_garage YouTube channel in review of many DIY battery boxes.

I too watch Andy's videos, but he hasn't cared much for compression until very recently. Once he finally got truly flat (unused) cells I guess

I'm not a big fan of any of these DIY boxes that attempt to apply correct pressure. My, perhaps slightly controversial opinion is the only thing that prevents the cells within from bursting from overpressure as they inevitably bulge during years of use is the flimsy sheet steel that itself acts as a spring of sorts.

Some of these box designs give up on correct pressure and just immobilise the cells by including a (rather thick) piece of foam on the ends. These probably have a higher chance of lasting a long time than ones that attempt to apply the required 300kg of force. But then, we don't get proper compression.

Anyways hope this helps look forward to seeing the finished product. I will be starting mine soon waiting on cells from China.

Thanks.

I decided to see how foamed PVC board would work as a force spreader between the steel pressure plates and cell separator so I bought some locally and I think it is a good idea. It is very firm, but slightly compliant material. Like polystyrene, but 10x harder. I got 3mm sheet.

I started "dry fitting" the battery and I noticed the steel shelf that is supposed to be fine until 120kg (entire shelving unit up to 400kg) sags badly with just 4 cells so they touch at the top, but there is visible gap on the bottom...

So tomorrow I'll plasma cut a rectangular "spreader" made of 8mm mild steel plate that will go between the cells and the shelf. Why 8mm? Because that's the plate I have big pieces of and it will take tons of weight without bending.

OSB or plywood would work too,so why use expensive thick steel? This is in a cellar and it might get damp occasionally. I try not to have even a gram of wood in there. Otherwise it will get moldy and it will stink in a year or two.

 

[Bananaman321]

It has been my frequent reading for last number of weeks where do you think I got the steel plate idea from...

Thank you for replying


Can you elaborate a bit on this? What do you mean it doesn't apply equal compression?

The only thing that may affect the compression strength for individual cells I can think of is maybe friction, but that is absolutely tiny at the very small movement amounts and relatively high forces.

Perhaps what you mean is that ,lets say cells on the left swell more than cells on the right and therefore through the centre "pivot" point they apply too much pressure on the right... I'm not sure there is any design that mitigates this fully (other than using some constant pressure elastomer between cells). Springs help no doubt. Imagine you have a single stack of any number of cells. There is always a possibility of one in a middle swelling more and applying more pressure to its neighbors.


That's what (disc)springs are for in my current design. I may also add some neoprene.

This is what neoprene is for in the design, but i agree eva foam is definitely better if one can get the right type. However, I'm of two minds on the neoprene question. I think I'll just add some and see how it feels in place. I'll make a decision then.

I'm not sure it matters in a stationary application that 4 years down the line cells will touch bulge-to-bulge vs entire surface to entire surface. One might say, but this way areas surrounding the bulge will not be under pressure, but I don't think so. Eve gives some ridiculously high number as the bulging pressure (two or 3 orders of magnitude more than recommended). So it becomes very important not to immobilise the pack and have these forces act on neighboring cells. Springs take care of it and the bulge itself transfers remaining pressure to the rest of the cell.



Perhaps it was missed or not obvious, but the snake configuration (cells connected short edge to short edge) mitigate the solid busbar issue during cell expansion. It translates stretching/compressing motion to very slight rotation of the busbar on the post. This tiny rotation is well within the elasticity of the post.

I too watch Andy's videos, but he hasn't cared much for compression until very recently. Once he finally got truly flat (unused) cells I guess

I'm not a big fan of any of these DIY boxes that attempt to apply correct pressure. My, perhaps slightly controversial opinion is the only thing that prevents the cells within from bursting from overpressure as they inevitably bulge during years of use is the flimsy sheet steel that itself acts as a spring of sorts.

Some of these box designs give up on correct pressure and just immobilise the cells by including a (rather thick) piece of foam on the ends. These probably have a higher chance of lasting a long time than ones that attempt to apply the required 300kg of force. But then, we don't get proper compression.


Thanks.

I decided to see how foamed PVC board would work as a force spreader between the steel pressure plates and cell separator so I bought some locally and I think it is a good idea. It is very firm, but slightly compliant material. Like polystyrene, but 10x harder. I got 3mm sheet.

I started "dry fitting" the battery and I noticed the steel shelf that is supposed to be fine until 120kg (entire shelving unit up to 400kg) sags badly with just 4 cells so they touch at the top, but there is visible gap on the bottom...

So tomorrow I'll plasma cut a rectangular "spreader" made of 8mm mild steel plate that will go between the cells and the shelf. Why 8mm? Because that's the plate I have big pieces of and it will take tons of weight without bending.

OSB or plywood would work too,so why use expensive thick steel? This is in a cellar and it might get damp occasionally. I try not to have even a gram of wood in there. Otherwise it will get moldy and it will stink in a year or two.

I apologize I'm not going to go into too much detail for I am on my mobile phone right now

The only complications I see with your design is the distance between your restraints and the Spring pressure this allows the cells to move apart from one another which can put added pressure to the terminals that's where the eva foam comes in to play I'd rather see the modules secured with clamping force and let their expansion/contraction be litigated by dense close cell foam or similar substance. Or use flexible bus bars.

I only mention this because I have been doing research on the same subject for my own battery build.

I am currently awaiting 116 EVE LF280K cells from China. I purchased 4 extras to build a 12v battery out of the weakest cells. I also ordered 7 DIY boxes it just seemed more appropriate for my needs. Apexium

The 12V battery will have the same BMS as the rest of the batteries I consider it functional spare parts.

I wish you the best in your build I'm sure to be fine these are just concerns that come to mind.