Those ball type flow meters will change the flow rate. Not sure how much but could be significant. For measuring low flows without too much back pressure (almost none), look up soap bubble flow meters. They are basically a graduated cylinder with a nipple to introduce air and a small reservoir of soapy water. You just time the time for the soap bubbles to cross a start and finish line. Easy to make your own, too. Just need a clear plastic tube as the measuring cylinder and a little math if you want to use regular units.
Pressure is definitely a way to spot leaks. Depends on the pump, though, how easy it will be. Iāve just been measuring how far under water I can put the end of the tube from the silica gel dryer. When I had all the leaks and with the tiny pump I couldnāt even get 1cm of water pressure. As I sealed it up Iād get more and more. Havenāt measured on the new version yet but itās definitely air tight and should develop full pressure and max flow.
One way to characterize pressure is where the flow stops like with the water column measurement. With flow stopped, flow restrictions and such donāt matter. You can test the pump pressure alone and compare with your apparatus installed - they should be equal. If they arenāt itās because there are leaks. The aquarium pumps I used were very interdependent on flow and pressure.
And totally agree on all the adapters and such. Itās why the test stand looked like something out of a Frankenstein set and why I put it almost all in a 3D print. Saved so much hassle and neatened it up a lot. Too bad the print isnāt quite air tight. It would be nice to be able to guarantee air tight but every print has leaked to some extent.
I had never heard of such a thing before. Very clever.
Provided the air pump has enough pressure though, I still like your original idea the best. A hose with even just a little water over the end of it. Inspect for bubbles to get a visual indicator. Especially for a first pass, itās elegantly simple.
Regarding the silica gel tower, Iām not sure which direction to go on that. Yours has the advantage of packing in a lot of desiccant for far less cost than commercial grade ones, which have to be a lot stronger to handle potentially very high pressures in their typical use-case for drying compressed air. And packing a lot of desiccant is good, to handle higher flow and avoid the burn through you mentioned.
On the other hand, Iām leaning toward metal, because in my case this will most likely sit around unused most of the time until those times when I need to run the filament dryer. Metal would guarantee moisture stays out of it. On the other hand, the opacity would lose visibility onto the indicator dye that you enjoy, and that is an interesting effect to see the moisture loading inch up the column. The see-through column offers simpler monitoring, and simpler is good. So, for now, Iām undecided. Maybe I can decide better after the first pass. Meanwhile, if you notice it gaining moistureāor notāwhile inactive, that would be good to know. Perhaps itās insignificant. AFAIK, the drierite column is plastic (though what kind of plastic I donāt know), so that would tend to support the see-through idea. Then again, if I ever do go closed loop, temperature might suddenly become a factor, so perhaps glass would be the answer to that. A big mason jar could cheaply handle the heat (Iām guessing to pressure cooker temperatures, maybe higher) and still be see throughāand guaranteed impermeable. I need to look into what the upper limits might be, because the air temps for drying nylon or whatever else might get even higher than that. Soā¦after writing up these trade-offs, Iām changing my mind: high temperature glass would likely trump metal. In case you were wondering, thatās the angle Iām approaching this from. For drying PLA, yours makes total sense.
I was initially worried diffusion while the air dryer was off would take water to the desiccant but at least short term itās not much of an issue. I just checked the inlet and outlet RH meters and with the pump stopped overnight the inlet meter showed 37% RH (down from about 45%). The pump should block a lot of air movement on the inlet side so the inlet hygrometer would be expected to drop. On the outlet, Iāve used little air valves and clamping but the exit tube is about a foot and a half long now and the exit hygrometer was still reading 10 with no clamping or valve. Unfortunately donāt know how much RH has risen in the exit tubing thanks to the lower limit on the hygrometer display but probably not much.
What Iām more wondering about are the beads in the column but so far (now literally a couple of weeks) Iām not seeing the effect Iām worried about - water diffusion through the column. Right now the water is concentrated around the air inlet and the beads are getting really dark there. Up at the top (where air exits the reservoir), the beads are still the bright dry orange. Over time (itās a slow process) the water should move throughout the column and ālevel outā.
That would poison the column. Same if I shook up the reservoir and just mixed up the beads. As it is now, the beads are holding more water than they could (with a 10% output RH) if the water was evenly distributed. The color change happens at around 1/3 of the total water capacity of the silica gel beads so at the dark end the beads are probably close to saturating. The air leaving the column is dryer than it should be because so much water is held up at the inlet end and the still dry beads near the outlet are polishing the humidity down.
The flow from diffusion of water in the column is in the same direction as the air flow so using it to dry air just helps the process along and will help spread the water. Letting it sit will let the water spread through diffusion. So in a way there is a limit to how long the column will last once it is used and captures enough water. I donāt know how big the effect will be in practice.
If you use Drierite in a column, thatās a chemical reaction that captures the water and it is much more tightly held than water on silica gel. A Drierite column wonāt let water diffuse around anything like silica gel.
In perspective, though, the water diffusion issue in silica gel is one of those fortunate problems. Like I said, the column is holding more water than it should is thanks to the really lopsided water distribution.
The water bottle has a silicone sleeve on it to protect it but the way the beads are changing color itās actually pretty cool. Might just take that sleeve off. Hereās the column yesterday. The dark area is pretty striking. Itās almost a kinetic sculpture. It would be an interesting experiment to put some saturated beads and fresh beads in a sealed jar and see how long it takes for the water to even out among all the beads. That will give an idea how long a used column can sit unused.
That was exactly the same unknown that had me wondering as well.
If it does become observable over a longer timeframe, it occurs to me now that there would be an easy preventative: when youāre done drying your last roll of filament and donāt foresee doing more drying anytime soon, maybe de-mount the desiccant column and put it inside an impermeable container or some other kind of drybox. Or, if more convenient, keep it mounted but put your entire device inside such a container, for easy removal later. That would presumably halt any meaningful moisture intrusion over long periods of time while not in active use. And if not sure, then itās maybe cheap insurance. Pro-active belt and suspenders preventative measure. Not that keeping it as-is is much concern, since youāve designed it to recharge anyway, so maybe you just do that when you see that it needs it. Minimalist approach, and still easy. And you donāt need to decide now. You can just wait and see.
I think you donāt even have to worry about it with Drierite since itās chemical (but reversible). But for silica gel it might be a situation where you charge it up and then dry a pile of spools, let it sit until the next pile, recharge the silica, etc. Thatās worst case though. It all depends on how fast the water tries to migrate and so far I havenāt seen much evidence. Again, hygrometers that can read lower would let me see a little more what is happening.
The thing about demounting and mounting the bead reservoir is it will cause some physical bead mixing. There is a fitting on the end of the straw that extends to the bottom of the glass jar (to the top with the jar mounted upside down) that will mix some amount of beads as it gets pulled out and reinserted. Might not be too bad, though with the new design - I made it smaller. The older design in the air dryer currently running is a bigger disk that will really mix them up.
This is another question that will take time to answer but so far itās not really a factor. As long as the beads hold the water tight when at lower humidity, the dry beads will at least slow water migration I think but that could also be a wrong way to think about it.
Anyway, at least for now the way itās behaving is really good and water movement isnāt really noticeable yet. At room temperature it could be very slow to the point of not being an issue during the time people might use it before the beads need regenerating anyway.
And lastly there is just how much of an effect will it be on filament drying times. Water diffusion in the column will cause breakthrough to be more pronounced but that will probably only extend drying times unless it gets really bad. Infrequent use may cause the beads to need regenerating more often to control that, but I havenāt gotten there yet to know how bad that will be or how it will affect drying.
What I do know is spool 18 is almost done on the original bead charge and itās still working great. Amortized out itās about $0.50 per spool if I just threw the beads out and bought fresh. With the way the bead column is looking, it might even be able to hit 40 spools dried but Iām running out of filament to dry. Still have more spools but not that many. Once I run out of spools Iāll use it to dry all the silica gel packs Iāve accumulated but it may be a while before I can get these answers.
Oh, I see now: if Iām reading you correctly, youāre actually more concerned about the already trapped moisture spreading out, leading to migration upward within the column than you are about outside moisture penetrating through the shell and adding to whatever is already adsorbed. Not something I can remember ever reading about. Are you sure it would be a bad thing and not a wash? I suppose you could possibly prevent that from occuring by segmenting your column into connected chambers like a string of sausages. If you put valves between the chambers, then you could isolate the upper chambers by shutting the valves when not in use. Or you could maybe detach whichever lower chambers were showing any green and replace with fresh ones, and that would reduce the amount of desiccant that needed to be changed out. But, as you say, your longevity data proves the amortized material cost of simply dumping the whole load in the round bin and refilling is so low, maybe such a partitioning and compartmentalization just isnāt worth the bother. Itās great that youāve already established that itās so cost effective, because your dump option would also save you time while adding convenience, thereby reducing your soft dollar cost as well . Some of the desiccant columns I saw listed on Amazon arenāt rechargeable and are meant to be disposable, similar to how many water filters are disposable.
Oh, one other thing: I received the Drierite desiccant column, and I can confirm that it is plastic. I donāt know what kind, but Iād guess either polycarbonate or acrylic.
Eyeballing the cylinder wall thickness, it looks to be about 6mm wall thickness. Pretty thick! Maybe about the thickness of a Schedule 80 PVC pipe of similar diameter.
The idea is really good in concept but might be tough to do and be sturdy with no leaks. It would be really handy to pop off the lower chambers and add fresh at the top. But yeah, with this going long on a fill of beads it might not be worth it.
Iām pretty sure the water spreading out would be a bad thing. Hereās whyā¦
The curved blue line is just a guess at what the water distribution looks like along the column length but probably is somewhat close to how it is now with the dark indicator where the water enters. The light orange beads tend to point to the fall off of water content as you get closer to the outlet. In operation the blue curve will shift more and more to the right.
But as the water moves right a little it hits dryer beads and some sticks down, etc. but the water keeps seeing dryer and dryer beads which keep removing more and more water. At the outlet, at least at first, the air is really dry. But as the curve moves right, that tail on the right starts lifting. Thatās the breakthrough. But that distribution will have an average value if the water was distributed evenly throughout the column. The tail initially is below what the exit hygrometer can register because of the way later beads dry the air more and more. But the average water value can be well above the exit RH%. Over time that blue curve will look more and more like the average line for water content as the water spreads out.
But so far, that seems to be a slow process. It will happen though. The water is basically evaporating from and condensing on the beads. Itās the time when itās in the vapor phase thatās the problem. A net flow (like when the pump is running) will move water along the column until it sticks back down again. If the pump is off, itās more random but there will be a concentration gradient which will tend to push the water in vapor phase something like when the pump is running but slower because the water can be moving either direction. Itās just when moving towards the inlet, the water in vapor phase is more likely to hit other water molecules (since there are more of them / higher concentration) and bounce back the other way.
Thatās probably clear as mud but is pretty much how it works. I wasnāt counting on the water sticking as hard as it appears to, but first, I wasnāt counting on so much water getting trapped right where it enters. I thought the water wouldnāt stick as hard and more would be in the vapor phase to spread out or move with the air flow. But it ends up a big bonus and either reduces the amount of desiccant needed, or lets it go longer between bead changes.
Started drying spool #20 this morning. Air output still pegged at 10% RH (as low as the hygrometers Iām using go). The dark from the indicator changing color is now halfway up the silica gel column. You canāt see it in the photo but there is very slight discoloration above whatās visible in the photo to right about halfway.
Not sure about a 40 spool capacity with the silica gel but maybe depending on how much breakthrough starts slowing down drying and the ambient RH when itās running. All my running has been between 45 and 60% RH ambient. But I think 30 spools actually is looking possible now except Iāll have to open new boxes to get to 30 spools tested.
The act of pumping/compressing the air raises the humidity frequently to the point of water condensing out. Itās why compressors have water traps and drain valves. But letting air out of the tank reduces pressure and humidity falls. Itās definitely a way to dehumidify air.
An air brush pump will give volume and flow and an ability to purge with a temporary high flow that the pump canāt deliver steady state. Even better, because it was compressed, its humidity should be lower.
High purge flow could be problematic for desiccants because the flow could sweep water through the column and disturb the gradient that sets up at lower flow. But since the air out of a compressor tank should be lower humidity anyway, it may not need to blow through the desiccant column to be used for purging.
Depending on how high the pressure gets in that tank, you might have pretty dry air anyway. For drying filaments like PLA, who knows? Only way to know would be to find humidity documentation or just try it. I doubt an airbrush compressor would get you to air dry enough to hit a -40 dewpoint though but havenāt run the numbers.
20 spools of filament dried so far on one charge of beads. Iāve got some more poly containers on the way to have a place to put more dry spools to see if it can get closer to 30 spools dried on a charge. In the mean time the column will just sit for a few days and there have already been a few days and all the nights where the pump didnāt run to give diffusion in the column a chance to rear its head. Exit is still stuck at 10% so still holding up as far as can tell.
Yeah, I think it just reduces the moisture load, not really dry it all the way, but isolating it in that manner is better that than have it condense in the airline and then sputter the the condensate into someoneās airbrush.
From what Iām reading, the real purists in all this seem to be the guys with the plasma torches. Seemingly they want super dry air and no contaminants of any kind, including no bits of desiccant dust either.
On spool 21 now. Output air still below 10% RH. Some more containers arrived so have a place to put dry filament that keeps it dry.
I also ordered some PETG-HF so will get to see how that stuff dries. Bambu says drying temp is 65C and the Sunlu goes to 70C. Will be some days before I get it to try drying it.
IIRC, you mentioned you were filling up your desiccant beads and then struggling a tiny bit with jamming the tube down into it? What Iām noticing is that with a number of desiccant systems, the tube gets loosely positioned in the housing and then the desiccant beads get packed in around it. This guy illustrates that with his particular DIY approach:
Iām not saying one way is better or worse than another, but just thought you might find it interesting. He had almost the right idea, but in the end he seemed to forget that the entire housing was meant to be a compression fitting: the threads were never meant to form an interference seal. Heās missing one or two o-rings or gaskets needed for it to seal properly.
Iāve run across a couple of additional ways of doing it. Iāll post about them later. One is similar, but the inverse, to what this youtube showed. All of them have their proās and conās.
My design is a little clunky around bead filling. The bulk of the base gets in the way of partially inserting the straw and then pouring beads in. What I did to make it easier to just stab the straw with the end fitting down through the beads was making that end fitting smaller, a little bit pointy, and put a raised design on the end that with twisting back and forth helps drill down by moving beads out of the way.
Iāve got two of these things now - the one Iāve been testing and the one I built in the final configuration. If I take the beads from the one thatās running, the big end fitting on the straw will mix the beads as I pull it out and ruin the column. I need more beads to try stabbing the new end fitting to make sure that works like I think it will and theyāll be here tomorrow. Then I can get final pressure/performance numbers and post.
I just loaded spool #23 this morning. Exit air is still pegging the exit hygrometer at 10% but drying time is getting a little longer. Instead of the 8-10 hours it was taking to reach 19%. Spool #21 was nearing 12 hours to hit 19% RH. So pretty sure thereās breakthrough now and exit air humidity is rising. Still showing less than 10% exit RH but the dark region in the beads is really dark and itās no stretch to say it extends halfway up the column.
Nearing the end of this set of beads I think but not quite there yet. With all that water in the reservoir now and in the beads, this is the worst time for mechanically mixing up the beads or for water moving through the column on its own driven by the concentration gradient. So 40 spools on a charge of beads is now looking doubtful. 30 might still be possible.
As an aside, I put spool #22 in a cereal box last night but forgot to put a pack of desiccant in to help keep it dry. I checked it this morning and the hygrometer was pegged at 10% RH. On cooling, the filament dryer 19% RH I dried the spool to turned into less than 10% RH after it cooled.
Iāve got a project now that will use white PETG-HF and a spool is on the way. It sets me up for one more test on the breakthrough issue. I am planning to dry the PETG using the beads that are nearing exhaustion to see how much it dries / how much water gets driven off, then like you suggested before, switch over to air from a fresh charge of beads to dry it more and see if / how much difference that makes / how much additional water can be removed. That will give an idea for how much difference there is between drying with nearly spent beads and drying with fresh.
I think for me the most interesting question right now is whether cranking up your dry air output would reduce drying time by a lot or a little. Even if youāre happy with your current speed, it would be good to know in case youāre ever in a pinch and need a spool of filament dried as fast as possible. Also, some filament types (not sure which) may have a tendency to become brittle if exposed to heat for ātoo long.ā Since how long is ātoo longā is usually an unknown until itās too late, I think my bias is toward as fast as possible.
At a purely intuitive level, it makes sense it would be faster. If you get out of a swimming pool in Arizona, the air is most of the time dry enough that the water will evaporate so quickly that youāll be literally shivering, even if itās the heat of the summer. Contrast that with Florida, where outdoors your cloths stick to your body because even your own sweat is slow to evaporate. And compared to a truly tropical location, like southeast asia, Florida seems dry!
One of the more interesting things I took away from the above DIY video was how quiclky the guy was burning through desiccant. Thatās what motivated him to try the upgrade to a much larger volume of it.
Turning the air up now would be bad to do I think. Drying times are getting longer and the dark indicator color change may even be moving faster. Not from humidity since thatās been fairly constant recently at about 45%, but from so much water being in the silica gel. Itās now on spool 24 and is taking nearly 12 hours to hit 19% RH. I keep thinking Iāll check the output hygrometer to see it ticking to 11% but itās still pegged at 10%. Itās holding about 2.5 times the water now than I thought it would because of that lopsided distribution. I really thought Iād only get 10 spools dry.
But thatās why I think turning up flow now (not much headroom anyway - throttle is at 3/4 setting) would just show how much breakthrough you can get. Higher flows would definitely speed up drying with fresh desiccant, though.
This gets to why that guy might be changing his desiccant out so frequently - heās using it on a compressed air line. Just about anything he does with that air is going to be a high volume situation. At least much higher volume than an aquarium pump can push.
And he makes a great example of the opposing issues in this - more dry air will sweep out moist air faster but at the expense of using desiccant faster. This is where a recirculating dryer gets more efficient - it only has to collect the moisture in the filament. With drying the air into a filament dryer and pushing a lot of it you dry air that gets wasted more and more as you up the flow.
For needing to dry fast, more flow is one way. But more flow also encourages breakthrough which increases the RH of the air out of the dryer. Increased RH in the exit air increases drying times. Itās kind of diminishing returns with this method. But the best time to do it is with a fresh column if youāre going to do it. Lowest breakthrough and lowest exit air RH.
So Iām going with the slow and steady approach. It still gives way better results than propping the dryer lid open and especially not propping the lid open.
I am pretty far from that actually. The flow of air is very low from the aquarium pump and even when pumping direct from the pump into the Sunlu S2 where the flow would be higher, it still didnāt seem to be anywhere close to too high of flow for the heater.
The only time Iāve had trouble keeping temperature was when I had an earlier modified base on the Sunlu that had thin walls where the heat was escaping and I was trying to hold 70C (the Sunlu S2 max). I could get to 68C but heat losses were too high to get to 70C. I modified the base I printed for the Sunlu to have printed-in honeycomb insulation of sorts. Now it gets to 70C just fine.
I canāt say about pulsing vs steady state. It seems like pulsing might be better at getting rid of moisture but canāt say for certain.
About that air fryer thing - got to be careful with āfastā regeneration since it generally involves high temperature which can damage the silica gel beads and reduce their water holding ability. And if people have the blue silica gel beads, those are colored with a cobalt compound for an indicator. I wouldnāt use the air fryer for food afterwards and in an enclosed space without ventilation. Even with the orange beads, that indicator is supposed to be non-toxic but no idea if they stay non-toxic when heated. But dedicated to bead drying, itās a good idea as long as you go slower and lower temps. Air fryers seem to have good temperature control if you keep temps a little lower. Might be a nice system!
24 spools done now. Last one was a new black PLA refill that lost around 2.5g in 10.5 hours. Output air still at 10% RH (displayed).
Point taken. Regarding that, according to chatgpt:
The best temperature to dry silica desiccant is typically between 200Ā°F and 250Ā°F (93Ā°C to 121Ā°C). Heating the silica gel at this temperature for about 1 to 2 hours should effectively remove the absorbed moisture, restoring its drying capacity. Itās important not to exceed 300Ā°F (149Ā°C), as higher temperatures can damage the silica gel and reduce its effectiveness.
He set his air fryer to 116C, so heās in the range, though admittedly who knows how much temperature overshoot it had or even how accurate it is. So, to your point, he could set it even lower and probably not add all that much time.
Mainly I was impressed because my eva-dry, which I plug into the wall to recharge, takes 12 hours to recharge. Admittedly the eva-dry is stupid easy, but his method is 60x faster. But on the basis of purely easy-to-use, one could make an argument for pushing air through an eva-dry.
. Some of the desiccant columns I saw listed on Amazon arenāt rechargeable and are meant to be disposable, similar to how many water filters are disposable.
Maybe about the thickness of a Schedule 80 PVC pipe of similar diameter.
