The ubiquitous Icy Water Challenge has spread across the nation like free money. With the intention of raising awareness towards the ALS disease and promote research donations, many individuals have poured ice cold water on top of themselves, but not until naming three more individuals who must join the cause. At such an exponential rate, it will not be long until the entire planet has participated on this fun event. And with so much people giving it a try, there is no doubt more innovative ways of taking the plunge will show their way into our social media accounts!

As I was nominated to take the challenge I couldn't help but to think on how to make it different and have some fun while at it. After all, dumping 32 degree water while it is 100 out there is just not fun enough ;-)

Since I am addicted to CNC, I thought of making the first CNC controlled ALS Icy Water Challenge. Now, some controversy may arise from my CNC definition. After all, this is not a CNC machine like a mill, lathe, plasma cutter, laser, punch or 3D printer. But then again, I didn't need to make a part. What I needed was to dump water on top of me and since a computer was employed to send numbers to a controller, I think the CNC terminology applies!

It took me the most of two days to build this rig, which was used only once. But darn it, did I have fun while at it! So now is your chance to reproduce this project, if you feel like enjoying from a refreshing watering during the hot summer. Do note the ALS challenge may be over soon, but I am certain new excuses to dump water on people will be quick to follow.

But most importantly, remember to donate!!! Dumping cold water on your warm torso may be a fun way to spend a hot summer afternoon, but it really accomplishes very little!

Step 1: Building the Frame

There are really not too many instructions for this step, because this is as simple as it can get.

I like to pile up in extruded aluminum. For this, I go to auctions as frequent as possible, and when I see some old industrial fixture, I bid heavily on it (by heavily I mean enough to ensure I get it but not more than what the extrusions would be new!).

The beauty of acquiring old assemblies is that you get the extrusions, the brackets and the screws in a single shot! Literally, every single piece I used on this skeleton came from a fixture that I acquired on one bid years ago.

Extruded aluminum is also referred to as Adult's Erector Set. Simply bolt it together, and there you go!

I imagine you can obtain the same results with different medium such as wood, PVC, steel pipes, or tubular steel, but nothing will be as quick as extruded aluminum. It took me less than an hour to put this together.

Anyway, this assembly is basically made of two parts. The first part is the frame holding everything together, and the second part is the bucket swivel arm.

The swivel arm is nothing other than an extruded aluminum frame which is sized as to grab the bucket by its upper lid diameter. The bucket is one of those Home Depot paint buckets you can purchase as a storage bin. I have a bunch of them, so just emptied one and repurposed it through the challenge.

Step 2: Cutting the Bearing Supports

One of my favorite aspects of CNC is that you get to use CNC machines to build CNC machines. It is just mind boggling! Anyway, I knew I needed to fixate my ball bearings to the structure masts and for this I quickly jotted a design which I cut on the CNC plasma cutter.

On the first picture you will see 4 pieces, but only 2 were finally used. Those would be the ones at the left side. The two at the right side were a gross mistake on my behalf. Somehow, I thought of using 4 ball bearings, but this pretty much means the swivel arm could not be driven. I have lost track of how many times I have made this mechanical mistake, but I promise to try and learn this time. That's what I said last time, though...

The sheet metal I used came from another industrial assembly I dismantled, so it was painted. To weld this, it is crucial to remove all the paint. The quickest mean is to use a belt sander, but this is quite dangerous. Since I do not have infinitely long fingers (and neither do you!), I used two magnetic bowls to grab the part while being sanded. This added some protection to my fingers from being sanded off and making them really shiny, but also kept the obnoxious heat away from me.

NOTE: Best thing when using any power tool is assume this thing will hurt you soon. It will make you more conscious while you use the equipment. Whenever I go "Hey I got this!", that's when I get hurt.

The CNC plasma cut part also included some lines I could then use to position the ball bearing assembly. It was MIG welded together.

Step 3: Bearing Coupler (may Be Optional)

The bearing assembly I used came from some surplus acquisition eons ago. I have like 30 of these and they have been taking space for well over a decade. A few weeks ago I thought of trashing them, but I am glad I didn't as they worked quite nicely for this project.

The only problem with these bearings was that their Inner Diameter (ID) was metric; something like 30 mm. The pipe I ended up using as the swivel arm rod was 3/4" Outer Diameter (OD). Hence, I needed a coupler to interface both together. Not a big deal if you have a lathe and some aluminum stock. The pictures above show the interface.

Clearly, without a lathe, I would have been in quite the pickle. If you do not own a lathe, most Maker Spaces have one. A good reason to be a member of your local Maker Space. Otherwise, best thing would be to use a bearing with the ID being the same as the rod you will use.

Anyway, at the end the implementation worked as expected and you can see on the pictures the rod mounted into the bearing holders.

Step 4: Swivel Arm Rod Holders

My first mistake when designing this contraption was to put ball bearings at the bucket swivel frame. As I explained before, this is an horrendous mistake because then the bucket frame will swivel on the swivel! What? Yes, the rod will move as commanded by the motor, but because the bucket frame is mounted on ball bearings, the rod will rotate without engaging the frame.

It is crucial for the rod to be mounted at the bucket frame in a fixated manner. For this I needed to redesign this portion and create a different bracket than the one you saw on the plasma cutting picture. In this case I used the Tormach PCNC mill to create a small bracket out of 1/4" aluminum.

Notice there is a small protrusion on the center hole. This is so that the rod can grab the bracket and not slip. For this, however, I need to add a groove into the rod. This was done in the mill as well.

At the end, the mating was so good I needed to press this in using a HF 12 ton press.

The last picture shows the bracket mounted on the bucket swivel arm. Maximum torque!

Step 5: The Stepper Bracket

The ultimate goal for this application was to have the water bucket being dumped while under computer control. Any motor could have been employed, but I have a fascination for steppers, so this is what I went with.

I have no idea how many steppers I have (at some point in time, before downsizing, I had more than 500!), but the size I have the most is NEMA23. I quickly designed a NEMA23 bracket I could mount into the large bearing bracket and milled this at the Tormach PCNC.

One of the important aspects of this stepper bracket is the ability to swing the stepper in and out so that I can provide belt tensioning. Otherwise I would have had to measure this super precisely and design accordingly. Be assured that if I had gone this path, I would have obtained every possible distance except the one that I needed the most. Hence, I have stopped from deluding myself with the notion that I can design this accurately. Grooves it is!

The timing belts and pulleys, I already had in stock. I have been using them on a gazillion different projects, and always reuse. They are not cheap, so whenever I can get my hand on them from surplus equipment, I am quick to draw my wallet. But, if you must buy them, I would go to SDP-SI. Another plausible supplier would be McMaster, but their stock is not as varied. With either, expect to "brace for impact", because it is going to hurt...

If I ever need more, I may consider milling them out using a 4th axis. That will require some learning curve I have not gone through, though.

Another option would be to make a silicon rubber mold and cast them out. That could work out nicely, but be warned the silicon rubber is not cheap either, so unless you are making a bunch of them, it may not justify the cost.

Step 6: The Controls and Here Comes the Water!!!

For this project I could have employed my "CNC Mother Board" with its respective stepper drivers (something I designed about two years ago and promoted as a failed Kickstarter), but I wanted to make it an HYDRA-X project. I used one of my HYDRA-X10 boards because there was no need to drive big FETs. HYDRA-X10 as a bunch of GPIO's and PWM outputs and those are the resources I needed the most.

In essence I used two GPIO signals to drive the nENABLE and DIR pins and a timer resource to drive the STEP input from one of my Avayan Electronics stepper drivers (also of my own design). The DRV8818 based driver would then commutate the bipolar stepper and get the water deposited into the rightful owner. I had computed that 350 microsteps would get the job done.

To engage the profile, I coded a Visual Studio C# application which was waiting for me to press the letter "a" on my keyboard. The GUI would then send the "Move Steps" command to the HYDRA-X10 board through the COM Port. Once I pressed that letter "a", refreshment was a mere seconds away.

I imagine you have already watched a trillion ALS watering videos, but if at the end you still want to see mine, here it is.