Mounting 18650 cell packs on DIN rails

[enki]

I wanted to share a 3d printed mounting bracket design that I made to mount 18650 cell packs on DIN rails. I believe this can be advantageous for diy powerwall builds because of the following factors:

• DIN rails are a standard industry component which can be easily purchased instead of ordering custom-cut sheet metal like you would for the HBPowerwall mount
• they are very versatile with mounting equipment other than the packs themselves with a lot of typical electrical equipment mountable on a DIN rail
• they are easily expandable for multiple sizes or amounts of 18650 cell blocks, for example HBPowerwall mount is designed for a 4x20 cell block and needs extra tricks to fit bigger blocks (like shown on this example with 10x20 blocks)

Just to be clear I am not marketing this as being strictly superior to HBPowerwall's design . It was simply the most common and best designed device for mounting 18650 cell blocks that I have found documented in the community. Because of this it was a natural point of comparison.

The main disadvantage of using DIN rails to mount cell blocks is that the alignment of rails needs to be precise if two rails need to support a single cell block. This will be the case for most deployments and might require some kind of base board instead of using the wall directly. Naturally, using a bent sheet metal holder which already has the proper dimensions takes care of the above for us.

This is not a totally new idea, while looking for existing designs for mounting 18650 cell blocks on DIN rails a number of existing solutions have been documented

• View: https://www.youtube.com/watch?v=aI3y65HWzqs
  
• View: https://www.youtube.com/watch?v=Znbdh_Lq_w8
  
• View: https://www.youtube.com/watch?v=wuQiwXhlYKo
  
• View: https://www.youtube.com/watch?v=yzfTqGsklB8
  
• View: https://www.youtube.com/watch?v=kJQh48c9h5Y
  

After studying these solutions I have found a number of drawbacks with all of them:

• they are designed to hold relatively small cell blocks not near enough to having for example 4x20 block capacity
• many of them use a lot of filament in order to clip the cell holders from two sides and allow the block to "clip into" the holders, I believe this is overengineered
• some of them use the triangle-shaped cutouts in the holders as attachment points which don't allow for the Batteryhookup nickel fuse roll to be used on the blocks

The above findings have prompted me to design my own system of DIN rail mounts for 18650 cell blocks with the following features in mind:

• flexibility - the mounts should allow for either small or large blocks that might need to be supported by more than one parallel DIN rail
• price - the amount of filament used should be as small as possible
• unobstructed both positive and negative sides of the cell pack in order to facilitate all ways of connecting the cells to busbars

The mounting bracket design has been performed in FreeCAD and is based on https://www.thingiverse.com/thing:500504. It can be downloaded from Thingiverse. The basic design can be seen below:

The brackets are designed to mate with the dovetails with popular 20.2 mm pitch 18650 cell spacers.
This part has variants for both male and female dovetails as well as only hook, only clip or both hook+clip variants:

Dovetail	Variant	Bracket image
Male	Hook
Male	Clip
Male	Hook and Clip
Female	Hook
Female	Clip
Female	Hook and Clip

How do the different mounting brackets work together? In order to understand this let's look at an example 5x16 cell block. This will be mounted on two parallel horizontal DIN rails using two sets of brackets. The top set of brackets will be the Hook variant and the bottom set of brackets will use the Clip variant. This arrangement will allow for the entire block to be secured as if it was a normal DIN-rail electrical equipment. Don't worry about the corroded junk cells, these are just to simulate the shape and mass:

When you zoom in into the particular brackets you can see how the male and female dovetails mate with eachother:

A video below demonstrates the brackets being attached to the two parallel DIN rails:

TH-35 (DIN rail) mounts for 18650 cell packs

The demonstration of the process of mounting a 18650 cell block using the DIN rail holders documented on https://pop.fsck.pl/projects/th35-rail-18650-mounts.html

diode.zone

Please note, that all of the above information is also found https://pop.fsck.pl/projects/th35-rail-18650-mounts.html and on hackaday (https://hackaday.io/project/183288-mounting-18650-cell-packs-on-din-rails).

 

[italianuser]

I like reading about different ways of managing cells and holders. I'm sure you can get good results using DIN rails, I already knew about Adam Welch's project, my interest for lithium cells is thanks to him, too (thanks Adam!).

My concerns, in general, about big groups of cells are the following (I'll use the word "plastic" talking about the connectors you printout, although I'm sure there's different kinds/qualities of filament).

1) Weight. I really wouldn't want my cells to drop off the wall. That surely can't happen if they are on a solid shelve. Hanging on a wall should be quite safe because each parallel block is connected to the rail and also to another block, and all together they make a big solid block. But we must consider that cell blocks could require maintenance every now and then, so the plastic connectors will be "pulled" every time. It should be OK, but it depends much on the quality (and quantity) of plastic used for connectors.

I do feel real safe about my other stuff connected on DIN rails, but I will nearly ever touch them, plus they don't weight anything.

2) Plastic hardening in time. Depending on the quality of plastic used, plus exposure to high or low temperatures, I would expect some amount of "hardening" of the material over time. That could happen in 5 years or in 10, but it could happen, and that means risking a plastic connector failure. And I don't know how much cells heat impacts with this aspect. Where serial connections are made there could be more heat, and that happens on the top cells where DIN rail connectors are, that will be top part of the block.

3) Price.
>> price - the amount of filament used should be as small as possible
I'm not sure about this. The quantity shouldn't be "as small as possible" but, instead, enough to make a solid connections and guarantee a long life for the connectors. Also I'm sure there's cheaper and more expensive plastic. To hang a big quantity of cells on a DIN rail I would avoid cheaper/lower quality plastic.

4) About holders. Weight and plastic hardening must be considered also for the battery holders. What quality are they? Again, I'm sure my AliExpress 4x5 cell holders are good for being put on a flat surface. But hung up with top part of the holder under bigger stress and exposed to more heat... I mean... can I be sure the holder is a good quality plastic, or is it like chinese iron? Uhm... I'm not sure about quality of the holders, especially for a heavy 280+ cell pack made with 4x5 holders.

I'm sure you can get good results, but I'd prefer a flat solid shelve.

 

[OffGridInTheCity]

Really nice work and I enjoyed reading your share. I also agree with @itialianuser comments. There's a scale where this will work but also an upper limit. Have you established weight ranges? For example, the 1s100p individual packs in my powerwall weigh about 10lbs each and I currently have 84 of them in operation. That's ~ 840lbs total.

 

[enki]

I like reading about different ways of managing cells and holders. I'm sure you can get good results using DIN rails, I already knew about Adam Welch's project, my interest for lithium cells is thanks to him, too (thanks Adam!).

My concerns, in general, about big groups of cells are the following (I'll use the word "plastic" talking about the connectors you printout, although I'm sure there's different kinds/qualities of filament).

1) Weight. I really wouldn't want my cells to drop off the wall. That surely can't happen if they are on a solid shelve. Hanging on a wall should be quite safe because each parallel block is connected to the rail and also to another block, and all together they make a big solid block. But we must consider that cell blocks could require maintenance every now and then, so the plastic connectors will be "pulled" every time. It should be OK, but it depends much on the quality (and quantity) of plastic used for connectors.

It is true that DIN rails are not designed for carrying very heavy loads. The quality of fastening of DIN rails to the base (ie. wall) will be critical in my opinion. Here you have two things to consider - first not all DIN rails are made equal, there exist flimsy and "heavy duty" DIN rails. The heavy duty rails are made from thicker material and should be able to hold more weight. For example the rails you see in my video are the the thicker ones. There are also G-type rails (see https://en.wikipedia.org/wiki/DIN_rail) which have a shape better for carrying heavier components. The second point is the amount and quality of fasteners. Some DIN rails have mounting holes only at the ends but you can have DIN rails with multiple mounting holes along the rail and these would be preferred.

As for the rugedness of the brackets against occasional maintenance (unmounting and mounting) of blocks I would say that the greatest risk here is for the bottom "clip" holder as the springy plastic bit can break of during unmounting. I have made some simple experiments with unmounting/mounting the block a few times and it's good enough in my opinion but it can break yes. I do not see this as a significant problem though as replacement brackets can be easily printed and replaced if needed. As for the risk of failure of the top bracket (the "hook" variant) I believe it to be minimal as the hook carries static loads and is much thicker.

The second failure mode is the point of mating between the dovetails on the bracket and the cell holder. Here everything depends on the amount of dovetails (the length) of the bracket and the quality of the fit between the female/male dovetail parts. I have chosen the amount of dovetails on the bracket empirically and it might be a good idea to make the bracket longer for bigger blocks. As for the fit between dovetail parts this is unfortunately something that needs to be tuned based on the 3d printer you are using and it's settings. The dovetails need to snuggly fit into eachother and every 0.1 mm counts. My FreeCAD design has a parameter sheet for this where I can easily adjust the dimensions of this geometry and iterate towards the best performance. You can also apply glue to the mating surface loosing the ability to replace the brackets but gaining some strength.

I do feel real safe about my other stuff connected on DIN rails, but I will nearly ever touch them, plus they don't weight anything.

2) Plastic hardening in time. Depending on the quality of plastic used, plus exposure to high or low temperatures, I would expect some amount of "hardening" of the material over time. That could happen in 5 years or in 10, but it could happen, and that means risking a plastic connector failure. And I don't know how much cells heat impacts with this aspect. Where serial connections are made there could be more heat, and that happens on the top cells where DIN rail connectors are, that will be top part of the block.

With the plastic hardening over time the problem would appear on the bottom "clip" which can indeed break if you attempt to unhook it and the plastic has hardened overtime. I do not see this as a big problem though as the most load is on the top bracket where the hook portion is much thicker and has a static load during mounting and unmounting (aka nothing bends).

3) Price.
>> price - the amount of filament used should be as small as possible
I'm not sure about this. The quantity shouldn't be "as small as possible" but, instead, enough to make a solid connections and guarantee a long life for the connectors. Also I'm sure there's cheaper and more expensive plastic. To hang a big quantity of cells on a DIN rail I would avoid cheaper/lower quality plastic.

What I was talking about was more the way that the holder mates with the brackets. For example, consider the bracket from Adam Welch:

View: https://www.youtube.com/watch?v=aI3y65HWzqs
this bracket is as long as the block itself (one 5x4 in his case) and would be impossible or unpractical to 3d print for longer blocks. This was the reason for my design, by saying "as small as possible" I was keeping the mechanics in mind.

4) About holders. Weight and plastic hardening must be considered also for the battery holders. What quality are they? Again, I'm sure my AliExpress 4x5 cell holders are good for being put on a flat surface. But hung up with top part of the holder under bigger stress and exposed to more heat... I mean... can I be sure the holder is a good quality plastic, or is it like chinese iron? Uhm... I'm not sure about quality of the holders, especially for a heavy 280+ cell pack made with 4x5 holders.

These are all valid considerations and everything will vary depending on the quality of cell holders you buy. I would myself at least check if the holders are ABS as they are advertised.

The glass transition temperature for ABS is around 100 deg C and it's commonly used in electrical equipment. If your powerwall will heat up to 100 deg C under operating conditions that in itself is a problem. High temperature will come into play in case of a major failure and your points here are very valid. I will have to take those into consideration. My current brackets are printed out of PLA which has a significantly worse temperature resistance, I will research printing them out of ABS or a different more heat resistant filament.

Additionally for mounting of very large blocks (280+ like you mentioned) I would not use the layout that I currently have. My current layout has the cells parallel to DIN rails and the wall. For big blocks I would rather use a perpendicular layout where the cells are perpendicular to the mounting surface. This would require a different bracket design which I have not yet attempted.

I'm sure you can get good results, but I'd prefer a flat solid shelve.

I have found your thoughts on this very informative.

 

[enki]

Really nice work and I enjoyed reading your share. I also agree with @itialianuser comments. There's a scale where this will work but also an upper limit. Have you established weight ranges? For example, the 1s100p individual packs in my powerwall weigh about 10lbs each and I currently have 84 of them in operation. That's ~ 840lbs total.

I have not yet performed the experiments to establish an upper weight limit.

 

[J_Mack58]

I have a 100 bags of these my company was throwing away. They came with some CT's but we just let the CT's hang on the wire. They have pretty good DIN rail bite but I need to test them to see what weight they can handle. Cool ideal for a small pack being used on my test bench. I know I'm late but thanks for the share Enki.

 

[Redpacket]

Other types of rail might also be cheap, widely available but stronger, eg unistrut...