Transcription
Guys, I think I'm onto something. A unique way to keep your filament dry that I have never seen done before on hobbyist or DIY dry boxes. It's affordable, easy to build, energy efficient, and from what I can tell so far, extremely effective.
Bad news. Not the best news, I'm afraid. I'm not sure what I was expecting here, but it definitely wasn't this. Oh my gosh, guys. Oh, and look at this. So in this video, I'm going to explain what this unique drying method actually is, and then we are going to put it to the test at the end of the video to see if this really is the goldilocks solution that I think it is. Let's get after it.
At FormNext last November, I stumbled upon a manufacturer of these really impressive high-end dry cabinets for industrial filaments. You know the ones that cost like 10 to 15,000 Euros, but can achieve sub-10% humidity. And when I inquired about how these units actually work, the explanation was, at least for me, nothing short of a revelation. Basically, they use blower fans to cycle the air in the chamber through silica gel, causing it to come into contact with more air more frequently and soak up the moisture. It's kind of similar to how a chamber filter cycles air through activated charcoal, but instead of activated charcoal filtering out VOCs, it's silica gel filtering out moisture.
If it sounds incredibly simple, it's because it is. And it turns out this isn't a new idea. I'm just an idiot with no background in engineering. But if you're wondering why they charge up to 15,000 Euros for a glorified refrigerator with some fans and desiccant, it's because their solution also detects when the dessicant is saturated and then automatically dehydrates it in a way that doesn't then return that moisture to the chamber. But we hobbyists can do that in a much simpler and more affordable way, assuming we can get this solution to actually work.
But why go through the effort? I mean, we already have a wide variety of chamber heater and filament dryers out there. Heck, this 3D printed dry box that I built in a previous video already has an option to add some Polymaker PolyDryers in the back. So why not just heat the air up to extract the moisture? Well, as I see it, there are a few benefits to doing it this way instead of that way. The big one is gonna be power consumption. Both peltier modules and AC heating elements are super power hungry, and the ladder just isn't safe to run 24/7 unattended. This approach, at least in the initial moisture-wicking stage, uses extremely low-powered blower fans. Sure this is only half of the drying process. You do still need to extract the moisture from the desiccant somehow, but we can do that later in a much more efficient way, such as in the microwave or alongside filament that already needs drying anyways, rather than heating up the entire cabinet and everything in it.
Which leads me to the second major advantage. The thermal one. I'm no PhD in polymer science, but I still don't think that it's a great idea to constantly thermally cycle your filament every single time moisture creeps into its enclosure, even if those temperatures are relatively low. In the best case scenario, it's probably gonna put strain on the tightly-wound spools of filament as it expands and contracts, but in the worst case, it might even slowly degrade the polymers over time. Especially since the different filament types in that enclosure have different glass transition temperatures.
With that said, there are also some disadvantages to this method. First, because we're not heating the filament at all, not even to 45 or 50 degrees, we're unlikely to extract much moisture from the filament itself. Now, that effectively means that this solution is, hopefully, ideal for maintaining dry filament, but it still will require that you have another solution for actually drying said filament. Though as I mentioned before, if we double up our drying and shove our spent desiccant into a filament dryer, alongside the filament that needs drying, I'm hoping that we get a kind of thermodynamic two-fer. The second disadvantage comes down to maintenance. This system will only work so long as the desiccant is not fully saturated. After that, we will either need to notice it in time to replace it manually or rely on some kind of complicated system to detect the increasing humidity and then dispose of that absorb moisture properly. Now, I don't know about all of you, but I'm willing to do a whole lot of manual swapping and drying to save 15,000 Euros. But hey, maybe we can parlay some of those savings into some kind of HomeAssistant connected humidity sensor or something like that later in the video to alert us when we need to do the swap.
For now, though, we need to first test just how effectively this drying method actually is. But before we can do that, we need a well-sealed dry cabinet to actually test it on. Now as you might recall, I built this HDRY module cabinet by CC3D Projects in a previous video on the channel. But as is often the case on this channel, and I do promise I am working on being better in this regard, I wasn't able to finish everything before publication. So I first set about gluing in the magnets that finally arrived, plugging the holes intended for the PolyDryers that could not make it to me, given everything that's going on in my part of the world, and adding in the foam tape to the areas that were missing it. I even put some hot glue in the PTFE tubes just so no moisture could creep in that way. I also did have to adjust the doors a bit because my build is made of a hodgepodge of different filaments that I had lying around. Eventually after a bit of tinkering, I did get everything closed up reasonably nicely. At least I thought I did, more on that in a second.
But before building our desiccant loop dehumidifier, I did want to test the seal and at the same time do a kind of baseline test of how effective just stationary desiccant is. So I filled a large tray with three pounds of desiccant, I placed it inside the cabinet and I left it overnight. When I came back in the morning, the cabinet had gone from an ambient relative humidity of 58% down to 38% in the top, and 40% in the bottom. So, lots of room for improvement, both in actually sealing the box and in removing the humidity inside of it. Which meant it was time to get printing.
Enrico, the designer of this HDRY unit, was kind enough to turn my ideas into reality, and let me just say, I'm very glad he did because, wow. I mean, look how freaking cool this design is. I legitimately could never. The design is composed of four parts for each unit, and I set about building two of these units to go along with the two mid-sized 24 volt blower fans that I happen to have lying around. First we have this housing, which holds the fan in place and holds it up against the silica box that is affixed to the back wall of the HDRY with a nice little offset bracket that will allow airflow from every single direction. And finally, we have this nice two-part silica box which is held together by magnets. Now this box slides in and out, allowing you to remove and replace the box as a cartridge, should you wish to print extras. Additionally, you can open it up quickly and easily with the magnets, should you want to pour the desiccant out for drying or replacement. Personally, I'm incredibly lazy. I wanna be able to just plop this thing in a dryer as is and dry it just like this. So I chose to print it out of Polymaker's PolyLite clear polycarbonate. God, I feel like I'm a burning man, that is a lot of polys in one sentence. But because the design does have three colors with this nice little logo on the front, I did decide to print it on the Snapmaker U1. Of course, as everyone knows, the s Snapmaker U1. is really only suitable for pretty aesthetic filaments like PLA and PETG. It is not at all suited for serious engineering-grade filaments. So even though this video is sponsored by Snapmaker, I still had to make some pretty heavy modifications to the printer in order to get it to actually print polycarbonate.
As for the print results, I do think just this one part of the design might be a little bit too thin and delicate because it moved under the nozzle. But other than that, the U1 performed absolutely amazingly considering what a nightmare polycarbonate is to print. From there, it was time to go ahead and print the housings, which considering the fact that Snapmaker now finally offers hardened steel nozzles in their web store, I decided to print in Prusament PETG-CF, with the accents in Prusament Orange PETG. Plus, since the U1 does allow me to mix and match different filaments with zero added waste, I opted to actually support the print with PLA, since it doesn't bond to either of these filaments. And I just have to say the finish on these prints is absolutely stunning. The automatic flow calibration on the U1 is really second to none. In case you can't tell, I do really, really love my Snapmaker U1. They are some of the best printers that I have ever owned, especially at the price point. Plus, I do really love how Snapmaker is embracing and abiding by open source firmware and software, and they're even embracing a thriving third party ecosystem, instead of trying to shut down and control every aspect. So if you are looking for a true multimaterial 3D printer that can handle everything from PLA and PETG to nylon and even polycarbonate, look no further than the link in the description where you can pick yours up today. Now they are in stock and shipping now, and if you're very quick, you might still be able to catch the pre-order price. Before that goes away on April 10th.
But now let's get back to the project where it was finally time to go ahead and get the first unit assembled and to test it. Assembly of this unit is definitely a non-event. It's two screws to hold the fan, eight magnets, and some desiccant. Now, Enrico did warn me to cut a slice and check the fan holes first, but I was way too impatient and way too crunched for time, so I just had to kind of force the screws in and chamfer the inside of the fan. But don't worry about all that, I have let Enrico know so this will be fixed in the final design.
From there, it was time to actually test the thing out. Now, for the initial test, I didn't want to go ahead and cut a hole in the back of the unit to run the wires just yet, so I instead opted to just run the wire outta the front door, which probably didn't help with the seal. From there, I set up a timelapse, turned on the 24 volt power supply, and started tidying up the studio.
Now, I'm not sure what I was expecting here, but it definitely wasn't this. Immediately, the humidity inside the chamber started dropping and fast. In just the 15 minutes it took me to tidy up around the studio, it went from 42% down to just 27%. Mind you, that was with just one of these units, and we're going to be installing two of these units, and I might still store some of my filament with the desiccant containers in the middle of the spool. It's also worth noting that the cabinet was completely empty, meaning that on the one hand, there was a lot more volume of air inside than there ultimately will be, but on the other hand, there was also no filament or cardboard spools potentially leaching out their moisture. Still, this was pretty darn exciting. But 27% still isn't low enough to store exotic filaments like nylon. So I decided to leave the fan running overnight and head home for the evening to see how low it would go. When I came back in the next day. The results were pretty fantastic for this early on in the testing process. 22% on the top and a respectable 18% on the bottom where the desiccant actually was. Honestly, that's already pretty darn good, but I think we can still do better.
My next step then was to go ahead and check for any leaks, allowing moisture back into the box. As you may recall, my HDRY was printed out of a variety of materials I had on hand on a variety of different printers, and with some creative workarounds when I ran out of the foam ceiling gaskets, so I knew that it wasn't going to be hermetically sealed as built. To test it out, though, I placed a heat gun inside, locked the trigger and broke out the thermal camera. Now, this was interesting for a couple of reasons. Number one, it reiterated to me that this box is not designed to be a filament dryer. It is very poorly thermally insulated, and you would waste a ton of energy trying to heat it up enough to actually dry your filament, which makes me very glad that we're gonna be shelving the polydryers for now. Additionally, it was interesting for me to see just exactly where the heat was escaping, and therefore I assume where moisture was seeping in. First, I could clearly see that some of the parts that I printed on my belt printer were actually slightly smaller than they should be, which meant that the foam gaskets were not creating a proper seal. Additionally, the thermal camera essentially acted like one of those UV lights in a sleazy hotel room, making me face each and every one of my shameful little mistakes.Tthe areas where I had to improvise and 3D print my own foam seals. The areas where the doors still needed adjusting. It was all as clear as day, and that made it quick and easy to fix. So fix it. I did. Ah, yes.
But there was one other thing that was bothering me with Enrico's otherwise perfect design, this diamond-shaped hole in the bottom. Clearly it's designed to redirect some of the air from the fan through the box so that it can recirculate throughout the entire cabinet, but it seems to take up a pretty large portion of the area where the fan is actually blowing in. What's more, I kind of wondered when the majority of the air just hit the top layer of the desiccant and take the quickest path outwards, rather than dissipating throughout the entire box. So I got to thinking, wouldn't you want some kind of internal vein structure to kind of direct the air down and distribute it more across more of the beads? Before building and installing the unit then, I decided to try and modify one of the cartridges on my own, rather than bothering Enrico on a Sunday, or reprinting the entire thing. So I quickly drilled a bunch of holes in the central vein, and then I went ahead and just plugged the bottom of it with hot glue. I then inserted this new cartridge and the tray of desiccant back into the newly-sealed cabinet, taped up the area where the wires came out, and took off for the afternoon. But when I came back in the morning, it was bad news. The results were exactly the same, 18% and 22%. Now, funnily enough, when I called Enrico, he reminded me that his HDRY unit with the polydryers does only ever reach 20%. So we're already on par with that result, but we don't want parody, we want improvements. And when I started this project, I set a somewhat unrealistic goal of 10% relative humidity, which meant it was time to go nuclear.
At this point, I heavily suspected, as many of you commented in the last video, that the issue had nothing to do with the silica box design and everything to do with air leaking in. After all, we are dealing with a three-dimensional jigsaw puzzle covered in joints and made of a hollow plastic that itself absorbs moisture. So in addition to further beefing up the seals with thicker 3D printed ones and completely replacing any of the caps that had tiny little perforations in them, I decided to do as Enrico did on his own unit and clear coat the inside. It was at this point that I was really grateful that the HDRY is designed to come apart so easily so that I could remove all the inner workings very quickly to get it painted. Side note: You know that things are getting serious on this channel when I, A, actually go outside, and B, pay the ridiculous price for building supplies at the corner hardware store. 14 bucks for a can of spray paint, are you kidding me?
So now with about two bottles of spray paint coating the inside and upgraded seals throughout, there was nothing left to do but a full scale test of the entire system. So I'm gonna go ahead and get both of these assembled and installed. I'm for right now, only gonna leave the drilled holes in this one, and the top one I'm gonna allow to redirect the air back, and we can tweak the design once we figure out what works more effectively. But before I did the final installation, a quick repeat of the heat gun thermal test showed that we had closed up the majority of the leaks and that meant that it was finally time to put everything back together. So I'm gonna go ahead and load up the entire cabinet with all of my most hydroscopic filaments. We're talking nylons, polycarbonates, and TPUs - all without their vacuum bag protection. Honestly, this feels like absolute lunacy before actually knowing that this thing is gonna be reliable long term, but if we're going to give this thing a proper test, then we need some real-world conditions. And once I get all that loaded up, it's time for the Passover holiday here, so I'm gonna head out of town for a couple of days and when I get back, we'll see how the system performed.
Okay, I lied. I ended up coming in just for a half day before heading out of town for the weekend, and I'm very glad I did because things were not going to plan like at all. 32%. So I spent the better part of the morning racking my brain and chatting with Enrico, just trying to figure out, what the heck had gone wrong? I mean, it couldn't be leaks. The seal is at least as good, if not better, than it was during our last test. Could it be that the spray paint wasn't fully dry and was still off-gassing moisture? Or maybe the filament and the cardboard spools weren't fully dry and they were letting go of their moisture at a much faster rate than I could have possibly anticipated? And if that's the case, I mean, I might as well consider this entire idea failure, because if all it takes to contaminate the entire dry cabinet with moisture is unknowingly adding a couple spools of slightly-moist filament, then I mean, what good is this thing? Right?
But then I remembered something that I learned from Stefan from CNC Kitchen who explained that, you know, saturated desiccant also release moisture back into your filament. And if that's true of filament, then it's definitely true of a nice, big, low humidity container. And that's when I realized not only did I not dry the desiccant in our desiccant loop containers after the last two experiments, but I also didn't check each of the smaller spool insert desiccant boxes before shoving them back into the cabinet. And when I took each of them out to check, I was quickly able to verify the problem and another potential challenge with this type of system. A lot of the desiccant had already soaked up a ton of moisture and instead of continuing to wick up more of it in the chamber, it was now releasing that moisture into the dry box. So to combat this, I went on a complete rampage, emptying out, replacing and drying all of the silica that I could get my dirty little hands on. I then put fresh silica in both the loot boxes and in each one of those spool inserts, and I even added another large tray in the bottom of the machine. And pretty quickly I have now been able to determine that this was in fact the issue, because although the humidity had climbed back up to 50% while I had the chamber open, it immediately began dropping lower and lower quickly surpassing that 32% plateau. So, for now, I'm gonna head out of town for real this time, and when I get back we'll go ahead and see if this positive trend continues or if this solution just isn't as effective when you actually add filament and cardboard spools into the mix.
Alright, I'm back, but it's not the best news I'm afraid. So I ended up staying out of town for five days and when I came back in this morning, the cabinet was sitting at 34% and 35% humidity, which obviously is not ideal. So I dried and replaced all of the dessicant once again, and fortunately we are already at 26% and 27% and dropping. Oh my gosh, guys. Okay. Sorry I had to stop the video because it's been about three and a half hours since I changed the desiccant out and recorded this, and look at this, 18%, I don't know if it's gonna keep dropping, but this is super, super exciting. So as far as quickly wicking up all the ambient moisture in the chamber, this idea has already more than proven itself. But as far as long-term storage goes, something still isn't quite right. And just looking at the desiccant that came out of this and how much moisture it is soaking up before the humidity creeps up, I think it's pretty clear that one of three things is happening here.
The first and most likely candidate is that this solution actually is pulling moisture out of the filament and the cardboard spools, which despite making this whole thing look like a massive failure is actually kind of a huge win, and that's not just copia. I do actually have good reason to believe that this is a big part of the issue because I printed this part out of polycarbonate that I'd actually left out for days before storing it in the cabinet, and it printed absolutely beautifully, in fact, much, much better than it should have had the filament not been somewhat dried. So if the humidity in the box is actually going up because our little device is slowly dehydrating the filament, then the solution to that is actually pretty simple. A, we need to make sure to only put very dry filament and spools in, and B, if we're going to store this many spools of filament, then we're gonna need much more desiccant to absorb that much moisture. Now to that end, Enrico has already been hard at work developing a significantly larger V2 of this idea, which can hold up to 350 grams of desiccant per box.
Now the second potential cause for our elevated humidity has to do with the type of desiccant that we're using. Silica gel's ability to absorb water varies widely, depending on both temperature and the relative humidity of the environment. And while it's really good at bulk absorption of large amounts of moisture, as we're seeing, like we discussed earlier, when it gets saturated enough, it can actually release moisture back into the environment. So I did a little bit of preliminary research and it seems that in order to reach ultralow humidity, it might actually be more appropriate to use activated alumina desiccant. So as soon as I can finally get shipments from abroad again, I'm gonna go ahead and order a bunch more of this stuff and experiment with replacing or even supplementing the silica gel. I honestly wouldn't be surprised if in the end, the best solution ends up being silica gel in one unit and activated alumina in the other. Now, I'll likely do a casual update on that and how those experiments go over on our second channel, so do make sure you're subscribed over there as well if this is something that interests you.
And finally, the elephant in the room. It is entirely possible that my HDRY cabinet still just isn't sealed well enough to properly test this idea. It could also be the fact that the HDRY is designed to achieve 20% relative humidity, but it's really not intended to go lower for a reason. After all, I'm no scientist, but I am pretty sure that the lower the humidity inside the enclosure, the stronger the vapor pressure gradient pulling moisture in through the tiny gaps. So it's probably worth testing this idea with something that we know is a hundred percent airtight, like an old wine refrigerator. Drop a comment below and let me know if that's a video you'd like to see me do.
So yeah, there's still a lot of work and testing to be done, and that's before we even tackle some kind of nifty solution to automatically dry the dessicant without releasing the moisture back in the chamber. But like so many of the projects that I do on this channel, the point isn't really to finish it all by myself and wrap it up in a nice little bow. This is the 3D printing community, after all. So the whole point is to put our work and ideas out there for the entire community to build upon. And I'm quite sure that the comments down below are already full of all your brilliant ideas for improving V3. That's why after speaking with Enrico, we decided to make all of these designs available 100% for free to anyone in the community under a CC-NC-SA license. That means use it, improve it, and share your creativity back with the community. And if you are as grateful to Enrico as I am for creating these wonderful designs and sharing them with all of us, you can support him by visiting the link in the description to pick up some of his designs for yourself, including the HDRY cabinet. Or go ahead and check out the recent video that I did documenting my entire build. We'll go ahead and link that down in the description. But if you're wondering, okay, cool, but which filaments do I actually need to store in a dry cabinet? Well, good news. I actually did a whole bunch of testing to determine exactly that and the results honestly surprised me. So you can check that video out right here. That's all for this week, but I'll see all of you on The Next Layer.