Peltiers are interesting things. Inefficient interesting things.
Basics of a Peltier driven Filament Dryer
Bambu has shared some drying settings on their Wiki (here). Some basic numbers for when drying filament on the heated bed inside the machine:
PLA - 60 C - 12 hrs
ABS - 90 C - 12 hrs
PC - 90 C - 12 hrs
PA6 - 90 C - 12 hrs
Interesting... and of course it's different when using a dedicated dryer 😅
Let's assume we build a dryer from ABS. Now the softening temp of ABS is about 105, but the heat deflection temp is a bit lower at 80 C.
So, let's limit ourself to 70 C, which means our dryer will need quite a bit longer than a 90 C dryer, but it will still dry.
Now, imagine that a Peltier element sits inside a closed box with both the hot and cold side, and we try to get the hot side to generate enough heat to heat up the filament, and use the cold side to condensate any water in the air. That's a double win, right? Hot air, that's actively dried...
Hmm... if it was that easy, everyone would do that... where's the catch?
In theory you could dry something at room temperature in an extremely dry environment, but that might result in excessive long drying times. (Water molecules inside the filament need sufficient reason to leave their comfy plate, and you can't do that without heating the filament, right?) Hmmm... that means we do need some heat...
What can a Peltier element deliver?
Coding scheme
TEC1 - 12706
TEC - Thermo Electric Cooler
127 - 127 pairs
06 - 6 Amps
Each junction is about 0.1 V, so a TEC12706 is designed for:
127 x 0.1 ~ 12V
06 = 6A
12V x 6A = 72W
In other words, if I take a TEC12706 and connect it to 12V and have sufficiently large heatsinks for the hot side and something smaller for the cold side it will consume 72W, right?
Well... not exactly. When I browse some datasheets (like this one, copied below) it's a bit below 70 W, but hey, close enough.
Datasheet (example)
Note: not all TEC modules are the same quality, or have the same specs...
dT
The max temperature difference I can create is 80 degrees. To reach 70 degrees I'd need to supply around 16.5 V, with results in an Imax of 6.1A.
My TEC will burn out at somewhere near 18 V.
So I can reach 70 C DT, ie. 70 C on the hot side, and 0 at the cold side, OR 90 at the hot side and 20 at the cold side. Theoretically.
There is no max temperature listed for these Peltier elements, but unless you have a high temp unit they're limited to 80 C absolute. So yeah, let's use the 70 C as our max. Most of the data sheets use 50 C as a reference in their graphs.
Efficiency
There things are notoriously inefficient, so we can look at them in two different ways:
A. We use them primarily as a heating source, and any condensation is a bonus, or
B. We use them primarily as to dehumidify, and the temperature is a bonus, or
C. We use them to heat up and dehumidify at the same time.
In scenario A we don't care about the efficiency. These things would be just a heat source, and you might consider an electric heating elements to do the job just as well. If you have a very well insulated box the 70 W might be a bit much, but otherwise it would be fine.
In scenario B the cold side sits inside the enclosure, and the hot side outside the enclosure. That way we could make sure that the cold side (in absolute temperature) would be low enough to create condensation.
In scenario C your would run the unit at 12 V, create a dT of 30 C, and create condens on the cold side. Something like that 😁 Now if the box gets too warm - for example 50 C - then the cold side would get perhaps 20 C, which isn't enough to dehumidify. It's a little more complicated than that, but I plan to do some experiments to see how far I can get.
My idea it two put one or more elements in series, so the first module cools the air from 50 to 25, and the second one from 25 to 0.
50 C temperature in the dryer
50 C hot / 25 C cold on the first element
25 C hot / 0 C cold on the second element
Note that I would have to bring some cold air to the hot side of the second element, as it otherwise would need to bridge more than 25 C.
Looking at the COP charts I would be running somewhere around 5 V to create a dT of 25 C with a COP of 1, and would have to arrange the airflow to do something like this:
hot air -> PT1 Cold -> PT2 Cold -> cold air -> PT2 Hot -> PT1 Hot -> hot air
Summarizing: I might increase the efficiency by running TEC modules in series (to lower the voltage) but it will reduce capacity and the lowest temperature the system can reach.
3 elements, air coupled
I might use 3 elements, each bridging around 30 C, each running at 4 to 5 V, to reach 60 C vs 0 C, with the elements being coupled by air.
I also would have to redirect the air flow like this:
hot air -> PT1 Cold -> PT2 Cold -> PT3 Cold -> cold air -> PT3 Hot -> PT2 Hot -> PT1 Hot -> hot air
That's... complicated, and might not work well.
3 elements, sink coupled
Many examples of Peltier stacks use different sized elements, ever decreasing the capacity the lower you get in the stack.
Now, what if I would get as close to a COP of 1 as I could get, and I would combine multiple, same-sized TEC modules. Instead of going like big module -> smaller module -> even smaller module, I could 'build' my own 'larger' modules by putting two regular TEC modules together.
If I create a stack like this, and run the modules in series, I should be able to create something that A. gets cold enough, no matter how hot it gets 🤔 and B. be a bit more efficient, simply because I'm putting multiple TEC modules in series (thus reducing the current going through them).
Is this the cheapest solution? No. No serious dryer manufacturer will use this approach. But I might...
I might give it a shot with 3 TEC12706, each running at around 6 V, which should give me a dT of 30 C. An old laptop power supply should give me enough power at about 18 V. Heat generated would be about 18 x 2 A = 36 W. Which might be enough, assuming the box isn't too large. I could build more than one module for more heat, or use 12703s for less. I might have to tilt or angle the design to make sure there's no water condensing in places where I don't want it, but it seems workable. Options, options, and crazy ideas 😅...
Disadvantage
There's one clear disadvantage (ignoring the complexity 😊) and that is that at the start of a drying session the internal temperature will be low, and the 'mid' plate may already show condens. In other words, such a solution must be build in such a way that that condens isn't affecting anything else in the dyer.
More dehumidification
Adding two TEC12703s in series with their heatsinks OUTSIDE the dryer would allow some additional dehumidification, even when there's no heat needed and / or the drying session is done. More components, more expensive, more complicated... more fun? 🤔
More heat
If heat is the primary goal, then efficiency doesn't matter. Theoretically, I could put a power regulator on a heat sink and add it to the mix. But then I won't have the advantage of maximal dehumidification... but I would have a dedicated / combined heatsink to figure out 😂
Air valves
Nearly all commercial dryers simply heat up, then - frequently - open two vents. One to let the hot, moist air escape, and the other to let colder air with a lower humidity back in.
I could add some air valves to experiment with. More complications! 🤣
Time to test!
I think it's about time to put the theory to the test... What other bits and pieces (aluminum, heat sinks, fans) do I have in the attic... Hmmm...
No comments:
Post a Comment