The most common type of 3D printer is termed "fused deposition modelling", FDM. This works by forcing plastic into a melting zone and then out through a nozzle. The object is then "drawn" by moving the nozzle around in the x and y planes, once the drawing is complete, the z axis moves and the object
builds up layer-by layer.
Usually, the plastic is in the form of filament, 1.75mm x a few hundred meters. The range of filaments is expanding constantly, allowing designers to match the properties to their application. Many filaments absorb water from the air, this causes problems. First, hydrated plastics are brittle. Causing the filament to break and wear printer components. Second, that plastic will be forced into the printer's hot end where water will rapidly convert to steam. This causes sputtering and clogs.
So filament needs to be dried, dehydrated, specifically. In the lab I have a few options. A lab oven, set it to any temperature above the boiling point of water. I can lower the boiling point of water with a vacuum. I could use a chamber with desiccant or use a steady flow of a completely dry gas, Nitrogen is dry because it's made by cooling air and the water freezes out a long time before the N2 does.
The oven is the simplest option, and this is what most commercially available filament driers are, little boxes that heat up to 50°C or so. Crucially, they have a fan to circulate air and are somewhat leaky so that as the small volume of air in the chamber is saturated, it is steadily replaced.
Raising the temperature of air increases it's capacity for water. So, take air at 20°C with 50% humidity and raise it to 50°C, now the humidity is 15% or so. Same amount of water. Similarly, take air at 70% humidity and drop the temperature 10°C and now the relative humidity is well above 100%, so the balance above 100% will precipitate out as fog/cloud/rain/snow etc. I'm from NW England, I'm familiar with the process.
Herein lies the problem. If you live in a humid climate, raising the temp to 50°C isn't enough for some filaments. It's also painfully slow, having to dry a spool of filament for 12hrs before printing is a huge lead time in what is supposed to be a tool for rapid prototyping. Why not raise the temperature? It takes exponentially more energy to 100°C vs 50°C, the little fans don't like that temperature. Ultimately you end up building a device similar to an air fryer, which actually make good filament driers if they have good temperature control. However, they're noisy, not really built to run for many hours at a time and at 1500W or so, they gobble power and will heat up a room noticeably.
So, a more elegant solution is to use lower temperature with a dry gas. The gas I'm choosing, is air. It's availability and cost are unbeatable. Often, air is humid, which was the whole problem in the first place. Conveniently, there is lots of freely-available, very dry air in the world. However, while moving my printer to the Atacama desert or Antarctica would work, the logistical complexity of collecting my completed prints would change from walking to my basement to a week-long expedition with expensive clothing. So why is the air dry in the Atacama/Antarctica? Because it was, or is cooled to the point where all the water precipitated out, it's that white stuff they leave lying around on the Andes/under penguins. So let's cool the air down so the water precipitates, drain it off, then re-heat the air so that it's relative humidity is super low.
So, take a small air pump, e.g. aquarium style. Use it to move air into a chamber with an aluminum heatsink. This is cooled to ~0°C by a Peltier device. If you move the air slowly, this allows water to precipitate out on the heatsink where it can drain away to a small tank through a small hole in the bottom. Now, we take our cool, dried air and move it to another chamber where it will be heated. Conveniently, in cooling one side to 0°C the Peltier will pump the heat to it's hot side. We can use the hot side of the Peltier to heat air & the chamber. Peltier devices can generate temperature differences up to 125°C, so 50-70°C is easy.
So by cooling the air to 0°C, and re-heating it to >50°C we can now pump super dry air into the chamber. But why not go further? You no longer need slightly leaky chamber to ensure a supply of fresh air, you can recirculate inside a sealed system. Much more efficient. As a further sophistication, the chamber should weigh the filament. Digital scales are cheap, all you need to do is monitor the weight over time, as water is removed, its weight falls. When the weight no longer changes, the filament is as dry as it's going to get. So you can stop the process. Because the chamber is sealed, the filament won't immediately start re-hydrating.