This seems cool, but I wasn’t a big fan of the part where the guy put his fingers right up near the 300 degrees nozzle. 
Regarding silica gel desiccant, if I dry it at 140C and put it in a sealed container with a hygrometer, the ambient air inside the container might get as low as 2% RH.
At the other end of the spectrum, when it is nearly saturated, it might get as high as 40% RH.
So, plainly there’s a continuium inbetween as the desiccant picks up more and more moisture.
So, lately, after drying a spool, I’ll put it in a sealed cereal container with a hygrometer and no desiccant just to see what the RH levels out to when it cools off. Afterward, I may have some not still fully dry desiccant (perhaps leftover from protecting some other spool), and I have to decide whether to re-use it again before re-drying it. So, how I decide is this: if the RH in a sealed container at ambient of that desiccant is lower than the RH of the sealed filament alone, I can use it. My guess is that if the steady state RH of the desiccant is already above that of the steady state filament by itself, then it would probably be donating moisture to the filament, and not the other way around.
I haven’t actually tested this to confirm whether or not it’s true, but common sense seems to indicate it would be true. What do you think?
I suppose in your case, if the desiccant has an RH < 20%, it would still be worthwhile (though obviously the lower the better)? In this case I’m just referring to packaged or unpackaged desiccant that cohabitates with the filament, not desiccant in a column used in a filament dryer.
The sealed container test works great. They are small in volume especially with a spool in there displacing air. The small volume of air means very little water can have a big effect on humidity. In general, whatever you put in there will dominate a humidity measurement.
So if you put silica gel in a container with a hygrometer, the RH will move to the RH characteristic for whatever water content is in the beads. If the beads carry more moisture the humidity will be higher, less water and it will be lower. It’s a physical characteristic of the silica gel beads and manufacturers generate humidity/moisture content curves that show that relationship. If you specify a humidity, you’ve specified a moisture content because of that water/humidity relationship and you can find lots of those curves are available on the web.
Filament is the same way but we don’t have many published moisture content / humidity curves. Thing is, even if we don’t know those curves there is a relationship very similar to silica gel.
If you test silica gel and filament separately, and their RH readings are different, water will move from higher RH/moisture content to lower to establish equilibrium.
1 Like
Clarification - if you put them in the same container then the one that would have a higher equilibrium humidity would donate water to the air to where the air would move to that equilibrium humidity value.
This is done all the time in the museum and archival storage fields. Companies sell silica gel preconditioned to moisture contents that will produce equilibrium humidities that help preserve documents and artifacts.
Humidity higher than equilibrium for the filament moisture content will just allow the filament to absorb water until it’s in equilibrium with the new high humidity value.
This is why I commented before that people who leave silica gel sitting out are letting it absorb moisture from the air to try to equilibrate to ambient humidity. To then take that silica gel and store it with filament just provides a ready reservoir of water to hydrate the filament with. Silica gel can never dry to lower humidity than its equilibrium value.
When the silica gel is used and already has moisture you get into the situation where it doesn’t have so much water capacity in reserve to hold water that the filament might release. It may scavenge some water but the humidity will rise quicker than if there was lots of remaining capacity to hold water.
Used/partially hydrated silica gel will still scavenge water as long as its equilibrium humidity is lower than the filament’s. You just need to keep a closer eye on it because the humidity will rise faster as water infiltrates and it can become ineffective sooner and need regeneration.
1 Like
Tangential, but this gizmo’s approach is interesting: Speculation about new printer announcements - #2883 by NoahKatz
That looks interesting. I think it could work. Only thing is like you said is it a deep dry or surface dry. They soaked the filament for a fair while so I don’t know their results are biased to just surface moisture. Using their PLA numbers it looks like they got the equivalent of a little over 1.5g of water off a kg spool which is in the ball park but maybe a little low from what I was seeing with PLA and that was with them loading up with extra water.
Grain of salt. I haven’t seen it and have some questions. But it should work. The question is how well.
And the inline dryers are stuck with the feed rate you use. That sets residence time in the dryer. The faster you print the less time water has to get out.
Also, it’s something else to set and check if you use different filaments with different drying parameters.
I don’t know. It could be good.
After drying 3 rolls of Elegoo High Speed PETG, I put each one in its own sealed cereal container with a TH sensor and checked them the next day.
Doing that, the resuts were: At 70F, two of the three registered 2%RH, and the third reggistered 4%RH.
Those RH numbers are much lower than I would have guessed. That means that anything other than very thoroughly dried desiccant would likely be donating moisture to the spool of filament.
Now, maybe it doesn’t matter, but I’m posting it because I was surprised by it.
1 Like
How long at what temp did you dry the PETG?
Yup silica desiccant will likely be a donor at that RH.
They do make sense and nice results! To answer your question, what you did is the humidity test I have tried to explain here. You got great results and this is why well-dried spools extend the life of desiccant in the AMS, in cereal storage boxes, etc. The filament actually will grab water from (reversible) desiccants which is why I’ve also tried to explain why you don’t use silica gel that’s been left out, etc.
But I think you are also correct about print quality not being helped by really dry filament. I haven’t seen any differences here but maybe there are. As long as my filament is dry to a particular warm humidity, it prints great. It’s very nice to have headroom for filament to pick up moisture without showing issues from that water.
Now the question - how did you do that? 
Adam Savage has a humorous pithy saying that I like: “The difference between science and fooling around is writing it down.”
To that end, here’s what I wrote down, and this will probably answer the questions:
I dried more or less to asymptote by weight. I think it’s likely that the cardboard spools acted like a desiccant, and also probably led me to drying the Elegoo spools for a lot longer than I needed to. This probably also explains the abnormally low RH. Evidence in support of that will likely come when I get the equilibrium RH measurements from high speed PETG that came on plastic spools (albeit different brands).
Creality is selling what looks like some really cool aluminum spools for $20-30. Not exactly cheap, but they’d likely harbor nothing more than trace surface moisture, and they include bearings so that the filament will allegedly unload more smoothly. They’re meant for drying high temperature filaments that would dry at high enough temperatures that they’d melt most regular plastic spools. So, why doesn’t Creality want to use cardboard for those?
Perhaps a cheaper way would be to print one’s own spools out of PolyPropylene, which, IIRC, is the only filament type that never needs drying because it’s allegedly more or less hydrophobic. However, this may not be sufficient for filaments that require really high drying temperatures.
There’s a aluminum bambu spool here:
1 Like
Good eye! That looks to be the same as what creality is selling in their amazon store:
where the listing even said it was Creality “brand”. I was a bit thrown by the MarsWork logo on it, but it passed-by me because I hadn’t ever heard of MarsWork before. Over the weekend Creality was even bundling it at a discount if you purchased it along with some overpriced creality filament, but I can’t seem to find them now.
I suppose the sheer weight of it may also address one of the AMS problems that can happen of the spool getting pulled up when nearing the end of the filament on the spool.
Seems all-around better than the $59 Visioneer anodized aluminum spool or the $144 filabot metal spool.
I’m honing my skills on drying PLA and PETG and gaining the relevant knowledge in part so that I’ll be better prepared for printing some of the very expensive engineering grade filaments that are known to be aggressively hygroscopic.
1 Like
I did end up ordering one of those MarsWork spools, as I was reading comments on an 85A TPU that it works better if the usual spooling mechanism of such a soft TPU is augmented to be especially low friction. So, we’ll see…
Meanwhile, here are the final results for the two HS-PETG filaments:
The Sunlu HS-PETG RH% is consistent with what I remember measuring in the past for that brand and type of filament (back when I fooling around but not writing it down). I’ve never before used or even tried the Kingroon HS-PETG, as this was going to be a trial spool. 8% is lower than what I was expecting. Kingroon is perhaps the cheapest filament on the market, and so i wouldn’t be surprised if it were a dodgy formulation.
For all of the above filaments, including the Elegoo HS-PETG, the first weight measurements were taken right after first opening the sealed packaging. i.e. how it arrived from amazon.
Aluminum is one of the worst metals for adsorbed water. There’s multiple layers of water on any piece of aluminum at STP in regular air.
Only being surface layers I’m not sure how much water that actually is, but aluminum holds onto water big time.
But also not sure how much it matters except for water loading of drying setups. Similar to cardboard, it all moves to whatever amount of water remains at whatever humidity it’s in.
1 Like
What about anodized aluminum?
Most aluminum is anodized which just forces the growth of an oxide coat. Cut into a block of aluminum and it grows it’s own coating pretty quick. Aluminum oxide is what attracts the water so hard. Basically it’s all aluminum not otherwise protected with paint or other treatment.
Anyone that has had to deal with aluminum wiring in a house is probably very familiar with aluminum oxide. Tightening down on terminal screws breaks the oxide and allows current to flow but heat/cool cycles reduce screw pressure and let air into the joint to form an insulating oxide layer. That layer increases resistance which increases heating which makes the process snowball and houses would eventually burn.
Now that process is understood and mitigated better but aluminum is very reactive just on its own. The oxide coat protects the aluminum from further reactions so is essentially self limiting but attracts water like crazy.
1 Like
Relating to spools, and tangential to the topic, but worth mentioning as a kind of PSA: I’m noticing that the cardboard spools on super cheap filaments, like Sunlu’s recycled PETG, can shed loads of carboard bits and carboard dust particles all over the filament. I received a 4 pack yesterday, and I’m sending it back. Not worth the jamming risk.
2 Likes
Reporting back:
I received the Marswork aluminum spool. I’m immediately doubtful that it will be as useful as I had hoped, because it seems optimized for engagement only with Marswork replacement filament, which 1. comes on a cardboard core and 2. that cardboard core has a particular notch in it at a particular location, which you can see here:
The aluminum spool engages with that notch by screwing a bolt through the aluminum sidewall into that notch.
This is no bueno, as I don’t forsee myself switching to Marswork filament just so I can use this spool.
Strange that the amazon Creality store is selling it. 
Does it not work with Bambu filament?