วีดีโอ

Hello Jerome, Thank you for your answer. I think that your idea to heat up the box is good. This should be done using the heat produced by the Peltier Module. Concerning the temperature of the Peltier, it depends on the desired humidity level and it's different from commercial air dryers whose function is to draw water from very humid air. For an air temperature of 20°C, the dew point is 0°C for a humidity level of 25%. I tried to obtain the lowest possible humidity level and i get a rate of 14% with the Peltier module at -16°C while the dew point is then at -9°C. In all my measurements, there is a difference between the theoretical dew point and the temperature on the cold side. Let me know if you success.

Ah sorry, I missed your very good point about the dew point. I will definitely keep you updated when I will do it.

Yes, I have seen videos on this. I chose the solution that seemed the simplest to achieve.

Simpler but ineffective. In this case you'd better use silica gel and regenerate it regularly.

@@workshopfeedback How? All you're doing that this solution wouldn't is turning it off when the box is opened and slowing down the heatsink fan to reduce noise, neither of which impacts performance.

I was reacting to "it runs when the printer runs, good enough.". I don't agree with it. For the rest, yes it can certainly be simpler, more noisy also. For me it was also a testing device as I didn't know where I was going when I started and I needed to be able to control all the parameters. There is a fairly fine regulation of ventilation in the box which allows me to obtain a fairly low humidity level. This would be difficult to do without a little software. And then, an Arduino nano is not the end of the world...

Thanks

J'ai essayé mais j'avais de meilleurs résultats sans radiateur. J'ai eu l'impression qu'il valait mieux avoir une petite surface très froide plutôt qu'une grande moins froide pour obtenir le taux d'humidité le plus faible. C'est vrai que les déshumidificateurs du commerce ont un radiateur mais ils cherchent à faire de l'eau dans un grand volume d'air, pas à faire descendre le taux d'humidité très bas dans un petit volume. Il y a un gros enjeux de résistance thermique de contact entre le module Peltier et le radiateur extérieur. Je pense qu'il y a le même pour un radiateur côté froid. Ca fonctionne pas mal comme ça, la taux d'humidité descend jusqu'à 14%. L'amélioration à apporter serait de renvoyer une partie de la chaleur dégagée vers l'intérieur de la boîte pour remonter le point de rosée.

@@workshopfeedback j'ai essayé et remerqué la même chose! c'est fou ça gèle vraiment^^

Great ! Thanks

@@workshopfeedback I came up with this design using a permanent electromagnet. There will be no heating of magnet and no dropping of head on power loss th-cam.com/video/8qZWVmBb7m0/w-d-xo.htmlsi=ROECuBAK_hz26QeN

Thanks, you're right. In fact I don't reverse the voltage but just switch the power off and let the device temp between both faces to stabilize. It has a real effect on the efficiency. By flushing every 4 hours I obtain a humidity rate below 15%.

@@workshopfeedback Nice!!!

@@workshopfeedback I recently had this idea pop into my head aswell, and then I looked it up if others did it before me, only I had a slightly different flow in my head. As I understand it, you are now effectively cooling your filament chamber right? If the temperature drops, the relative humidity rises, and the dew point also lowers (to sub zero). My idea was to guide the air from inside the filament chamber past the peltier element (ideally also with a small heatsink on it) where it condensates, then the colder dryer air should flow past the hot side of the peltier to heat it back up before it re-enters the chamber. Hot air will be more capable of taking up moist from your filament, and you might raise your dewpoint above the freezing point, eliminating the frost issue. At a chamber temperature of 40 degrees, your dew point will be at 0 degrees with a relative humidity of 8.3%. I think you should steer on having the cold side at around 0 degrees by splitting the airflow after the cold part in two channels, one directly into the chamber, and one past the heat side of the peltier. If the cold side is going sub zero, redirect more air directly back into the chamber, and if the cold side is getting to close to the dew point, cool the hot side with the cold air. At worst you simply have a heated chamber, which is how most filament dryers work anyways, at best you have an active dehumidifier

Hi Henkheijmen, I confirm the system slightly cools the inside of the box and I agree that it is necessary to use part of the heat produced by the Peltier to warm it up. We need a large heatsink on the Peltier and another small one that heats the inside of the box. By regulating the speed of the fans located on each side, we should be able to maintain between 30 and 40°C in the box. This would raise the dew point and provide better results. I understand the benefit of keeping the cold side above zero to avoid ice, but experience has, for the moment, told me the opposite: the colder the Peltier, the better it works. I simply let the peltier warm up regularly to let the ice melt. Concerning the radiator on the Peltier, I never managed to get good results, probably for the same reasons as before.

Thanks, I hope it will help you to success

Great, let me know if you success

I may share the fusion360 file with you if you think you ca use it

@@workshopfeedback yeah i think i could fiddle with it!

@@birdmananyweather1681 Send me an email, I'll return a download link. You can find my address here : www.youtube.com/@workshopfeedback/about

Thanks for your comment, have fun with your design.

I have no real problems to print flex filament with a bowden printhead.

I used this one in 24V fr.aliexpress.com/item/4000766066541.html

Hi, it is a carbon fiber plate, 0.5mm x 10mm

@@workshopfeedback where did you buy it?

Amazon a long time ago. I found this which should be similar www.amazon.fr/XMRISE-Bandes-Carbone-Barres-Accessoire/dp/B094QQRN4P/ref=sr_1_5

Hi Matheus, yes, I think that it should be ok

@@workshopfeedback thank you

Hi James, Using part of the power to heat the inside should be nice. You may be able to get a very low humidity rate. Let me know what you did. Thanks for the diodes, you're right, they are missing on the fan drivers.

Thanks for watching GeekDetour. I'm glad you liked it and watched your video about Sunlu Filament Box. I wonder which solution, condensation or heating, is better in terms of electricity consumption. With the condenser, I maintain two coils at 15% humidity with a consumption of 30W and I was impressed by the 350W of the S4.

Ah Ah ! subliminal propaganda ?

By the way i have been working on a similar idea...where you would dry the filament as it is being extruded.

My design files are not clean enough to be wildly shared but I can help and share some parts if you have questions on specific points.

Hi John, and thanks for your comment. I have to say that I don't feel like I'm violating anyone while I'm tinkering in my workshop. I'm curious to know what good idea I had that isn't the simple application of knowledge shared by everyone for a long time. Unfortunately, I don't know Stratus.

@@workshopfeedback Here is just 2 of their plethora of Patents. th-cam.com/video/nB2AFsNMAg8/w-d-xo.html As to what you might be violating. The minute you share it. :)

@@johnbeima6413 Hello John, Hello John, Thanks for these informations. I am not a PI engineer, but after looking at the patents mentioned, I identified several significant differences in the description of the processes compared to those used on my printer. On the other hand my lawyers, who are not IP specialists either, do not seem very worried about the claims that can be made in the context of this video. You may be an engineer or a lawyer specializing in IP, in which case I would be very interested in your opinion on either of these points.

@@workshopfeedback By the way, this is because we are looking at approaching this as well. So this is something over time we would also come up against. Hence asking, since many others feel there is no way around these.

Hello Liam, thanks for your comment. I already discussed this point in the comments. The idea of using the consumed power to heat the inside of the box is good. This will increase the dew point and the efficiency of the condenser. But I think that the idea of keeping all the power inside is not good. The principle of the Peltier effect is to maintain a temperature difference between the two faces (inside and outside) If you put outside into inside the all system will heat up and the temperature difference between the air and the cold surface will not change or worth decrease. I presume that the good way to do it would be to have a bigger heatsink with an outer fan connected to a smaller one heating the inside with the inside fan. A smart fan speed management should be able to maintain a target temp inside the box while maintaining a low temperature of the Peltier.

@@workshopfeedback I was toying with the idea of using peltier for a dry box. Never got to building it. The idea was that the box would be vented and the incoming air would pass over the peltier to remove moisture. The hot side would have 2 radiators. One of them enclosed in a duct so that the box air could be recirculated over it to heat the box. Second half would be cooled by external air as needed. Why I ditched the idea was that peltier can keep around 60degree temp difference. If I want the cold side to be -10C then I'd need to keep the hot side not over 50C. The box temperature would not be nearly enough to dry nylon etc. Maybe it would work if I would use only external air for cooling the peltier and use something else for heating. For storeage I use silica gel that I throw in the kitchen oven once in a while. Keeps the box humidity around 15-20%.

Thanks Doug. I had a look to your work which is also very interesting. It seems that we share a design addiction.

@@workshopfeedback 🙂Yes!

@@workshopfeedback Any chance you have shared, or would share, your printable files?

I just sent you a download link.

@@workshopfeedback thank you! Also, by using your approach, I designed a magnetic kinematic tool changer for the LowRider v3 CNC. I’m almost ready to publish it.

Merci !

I'm curious to know more about it !

Hi, thanks for your comment. The magnet consumes 6W and heats a bit because it's enclosed in the carriage. I added a 30mm fan and left a down to top air path in the carriage to avoid overheating which is sufficient.

@@workshopfeedback Yes, thank you for your kind reply, out of curiosity I asked the question before fully watching the video, in fact you mentioned it, my fault :) if I may ask what g-code command do you use to activate the magnet? i would like to come up with something like this in the project to make a home printer for me thanks.

M42

It's a 200N one (24V / 6W) Didn't find any suitable an affordable "no-power" one.

Thanks

I'm sorry, the industrial file isn't in a shareable state at this time and I don't think that it would be a good idea to start a new design on this basis. I'll do my best to answer to your questions.

No that was my first try !

Thanks, I'll try to share more.

Hi, here is where I bought them : www.magnet-shop.com/adhesion-grounds-adhesive-surfaces/metallic-grounds-for-screw-mounting/2407/metal-disc-oe-34-mm-with-hole-and-countersink-nickel With a M5 conical screw

@@workshopfeedback thank you! I didn't know such plate exist. Again thank you for inspiring me with you weel thought design. Have a nice day

Using the hot side to heat up the chamber seems to be a good idea. Using the heated air to refresh the peltier module is not as this will also increase the temperature of the cold side.

@@workshopfeedback My idea for a similar project was to use the hot air to actively drive moisture out of the filament, either on the spool or in the path to printer. But since I learned about molecular sieve, I've just been printing out of passive MS dry boxes.

Thanks

Thanks Allan

Sure ! It's carbon fiber 0.5 mm x 10 mm x 1 m bought on Amazon in 2016 but no more available.

@@workshopfeedback Thanks! I have very little experience with CF do you think 1mm would be too thick to bend properly in this application? Thanks for the info!

@@nerddub I think the main problem with the thickness of this part is the elasticity and the varying force it applies to the carriage depending on its position. The more flexible the better, you must feel the influence it will have on the kinematics of the printhead 🙃 Difficult to know without having it in hand but I think that going from .5 to 1 mm makes quite a difference. For the wires going to the carriage itself, as there's no bowden tube to support, I cut the strip in half lengthwise with a Dremel, it makes it softer.

You are right we could use the hot side to heat the box but I don't think that putting it inside would be a good idea as it will probably overheat. Using a commercial one could also be a solution but still outside of the box.

And I guess commercial dehumidifiers keep their cool surface above 0°C to prevent freezing. This gives a dew point around 25% or 30% humidity at room temperature.

@@workshopfeedback hm yes. thinking about it some more I suspect the move might be to initially draw outside air over the cold side to condense the water and then have the now cold and dry air go over the hot side to raise the temprature, drop relative humidity further and recover some of the energy used to cool down the air, before releasing it into the box. still feels like continually introducing fresh air into the system can't be quite optimal as you are also displacing the allready dehumidified air by doing so but you are probably right that keeping everything closed naively would lead to overheating before raching optimal dehumidification. Then again having the inside of the box quite warm means that the relative humidity reached at a dew point just above freezing woud be quite a bit lower than at room temprature. maybe a dehumidifier plus a thermostat to prevent overheating could be a nice and lazy solution as one could just get some off the shelf devices, drop them in a box with filament and be done with it.