It keeps the chamber free of deposits for over 1000 hours when printing ASA / ABS, and keeps my small room mostly odor-free for 4 to 6 months before needing to rejuvenate the 1kg (2.2lbs) of activated carbon.
The STL and 3MF files are on Github and MakerWorld
Now working on a P1/X1 and a P2 version, with option for a beefy heater so we can finally heat up their chamber fast to 55-60°C for high temp filaments and avoid warping.
Happy, clean and odorless printing to all !
Scrubby-Doo versions
- H2: Github MakerWorld
- P1/X1: Github MakerWorld
- P2: Github MakerWorld
- HOT.P1/X1 with heater: (coming)
- HOT.P2 with heater: (coming)
______________________________________________
Chapters (sorry, long post...)
- Overview
- Why ?
- Printing
- Assembly
- Parts
Overview
The hat just fits between the top glass and the top of the printer. It is held in place by the printer's rim, no screws, tape or magnets needed.
Changing the carbon is also toolless: just remove the glass, slide out the filter chamber's lid, swap the carbon bag, done.
It doesn't interfere with the AMS, which can still be installed on top of the hat.
The recessed front face also ensures no interference with the opening of the automatic vent when printing PLA / PETG:
The toolhead, chute and bed plate remain visible when looking down into the printer. Even when an AMS sits on top of the glass. Very useful when monitoring or debugging prints.
A few storage spots are available, to keep nozzles / tools / bits handy. 2 removable storage bins can also be positioned in the front, though they'll obstruct vision:
The filtered air exhausts on the right side, away from the prints, to avoid warping when printing ABS. That approach has been working wonders on my P1S for the last 3 years. View from inside:
However, note that this scrubber, although very effective for ABS / ASA / PC / PA, is less effective at containing odors with PLA / PETG than the scrubber I made for my P1S.
That's because the H2's integrated air filter's implementation is a joke. Not only does it not even properly capture the deposits produced by ABS / ASA (pics below), but when printing PLA / PETG the H2 filter is bypassed and the chamber fan actively dumps fumes out into the room. Which results in a stronger smell than with P1 as there's no easy way to turn off the H2 exhaust fan, unlike on P1. How dumb is this implementation ?
So people have developed exhaust filters for H2, like the Vento from Voxel and this filter from Chuck240. But of course they contain very little activated carbon and don't multi-pass recirculate, so their effectiveness and longevity are limited.
I might end up developing a more effective solution one day for PLA / PETG, combining recirculation and cutting off the exhaust fan, or adding an active vent to the scrubber's exhaust, but too busy at the moment...
Why ?
Not only does 3D printing emit toxic fumes (PLA included) but printing ABS / ASA also results in everything inside the printer being caked with a thin translucent coating. Here is what my H2's glass panels looked like after 350 hours of printing ABS:
On the other hand my P1S after 1500 hours, 1/3 of the time printing with ABS, is squeaky clean. Never had to wipe the glass even once. The difference ? The previous home-made recirculating air filter & heater hat holding 1.5kg (3lbs) of activated carbon on the P1.
Unlike many other onboard fumes-scrubbers, like the Nevermore or the Bento, the intake is right above the toolhead, and the 120mm blower moves a lot of air. So I suppose this immediately captures most of the fumes, thus preventing deposits.
Worse, not only was the H2 fogging up fast but, the few times the window in the room was closed, I had to leave the room after ~15 minutes due to a too-strong PLA odor during prints with High-Flow nozzles. In 3 years that had never happened with the P1+Hat.
Yes, one should never stay in a room with an active 3D printer, I don't, but getting a headache in a matter of minutes is definitely wrong.
And yes the H2 has an integrated carbon filter but it only holds a few grams of activated carbon, which probably needs changing every couple of weeks to remain effective when printing 12/5 (though Bambu claims 60 days). Note that it is also bypassed when printing PLA / PETG, with the H2 directly venting the fumes out...
Heck, look at what the Creality K1C's carbon filter looks like, what a joke:
Screw that. Hence my need for a recirculating filter holding tons of carbon like on my P1. But 3D printed this time, no more woodworking, and no more 120V fan / heater for safety reasons.
Printing
Settings for H2:
- 0.4 standard nozzle, 0.24 layer height
- default Bambu ABS settings: 0.95 flow, 100% Shrinkage, etc
Note: this model was designed without prior shrinkage calibration. The CAD model was unfortunately tweaked dimensionally via trial and error on my printer, instead of using absolute dimensions + shrink calibration. So YMMV and you might have to tweak your filament's Shrinkage setting for proper glass clearance. That is not optimal, live and learn.
- 2 walls, 3 top, 3 bottom
- 20% Support Cubic infill, Infill Combination on
- 0 Minimum Sparse Infill Threshold (reduces print time on grille / tall narrow features)
- Support enabled, Critical Regions only
It takes 3 plates, ~30 hours and ~1.1kg of filament to print. Or more when also printing 1 or 2 of the optional front storage bins.
This filter was designed to be printed in ABS so as to support 65°C chamber temps. Otherwise it would soon become a Dali masterpiece.
Once printed, there should be 1 to 2mm (~1/16") of clearance all around the printer's rim, and around the glass.
If one never prints with high temp filaments, then using PLA or PETG is fine. However, those are much less prone to shrinkage than ABS is, so the parts might print a couple of millimeters larger. Consequently, once dry assembled, carefully check the top-of-printer clearance and adjust the slicer's Shrinkage setting if need be so the hat sits inside the top rim.
Assembly
The glueing order is as follows, so I also chose to print in that order:
Once all 4 pieces are printed, dry assemble them, and check for clearance all around the printer's top rim, and around the glass. If there's not enough clearance, change the slicer's filament Shrinkage setting as needed.
After printing piece #1, break away the support wall of the blower exhaust:
When glueing 2 pieces together glue should be applied only to the piece that shows a protruding chamfered foot at the bottom of the face to glue. This ensures the glue won't run-off as the pieces slide against each other. Hence the glueing logic below: #2 → #1+glue, #4 → #3+glue, 4+3 → 2+1+glue.
The foot also ensures vertical alignment: once the sliding-in piece bottoms out we know it is now aligned.
CyanoAcrylate glue (Superglue) should be applied along all the ridges of the dovetails that connect the pieces together, as well as across all the faces on piece #1 to ensure that pressurized air doesn't escape through cracks before entering the filter:
So, first, piece #2 must be inserted top-down into #1:
Then #4 is glued to #3, making sure the assembly doesn't bow out or in. A ruler can help with that:
Finally, 3+4 is glued to 1+2.
Note that at each glueing step, clamps should be used to ensure parallelism and avoid gaps between the pieces while the glue cures:
Then the foam tape is placed around the blower exhaust opening:
And the blower can now be installed. A good trick to ensure it is tightly pressed against the tape is to wedge a tool between the blower and the wall opposite the opening, before tightening the screws:
Make sure to manually check that the fan spins freely. Depending on the brand, dimensions can be off by a millimeter or so, and the center inlet hole can lack concentricity with the spinning cage. If there is any sign of rubbing, the thin ring lip can be broken away. It reduces air rejection at the inlet when the carbon filter presents too much resistance, but it's not necessary with carbon pellets:
The assembly is done at this point, if no front switch is needed. In that case, just glue the switch plug in the front:
(optional) Installation of a front switch:
- solder 1 male and 1 female barrel connector to the switch
- feed the barrels + cables through the opening in the front of the box
- connect in series between the blower and the speed controler
- make a simple knot in the cable that exits the box, such that the knot rests against the slit in the back wall and provides strain relief
The speed controler can be left either inside the box, or outside. However, since this a cheapo device (most aren't UL / CE certified) and since a lot of my printing is done at 60°C (i.e. way past most consumer electronics' spec for ambient temp) mine will stay outside. I don't want it to fail fast or catch on fire...
Before filling the chamber with carbon, install a 32x11cm (12x4.5") piece of MERV13 fabric against the exit openings on the right side of the filter chamber. Tape it down so it stays put when filling the chamber:
Filling the chamber takes about 1kg of carbon. Spread the pellets evenly and flush with the top of the chamber:
Finally, fold the fabric and close the lid. Do not force the closing or it'll make the lid bow up, letting air escape. Instead remove some pellets and try again.
Now the airflow can be tested. At full blower speed there should be a noticeable amount of air coming out all along the right side of the filter. If not, remove some pellets.
Das ist alles. Muy fácil. Ça casse pas trois pattes à un canard, non ?
Enjoy your now crisp fresh air ! (but always leave a window open, or exhaust air out of the room)
Parts
Any 12V 120x32mm air blower will do. I chose one with a speed controler that directly plugs into 120V, and with a quick disconnect so a 12V switch can easily be inserted in series:
The blower is held in place with 3x 40mm M4 flat head screws and locknuts.
Foam tape is used to seal around the blower exhaust. Any tape ~1mm thick will do.
It takes about 1kg (2.4lbs) of activated carbon to fill the filter. It must be acid-free (make sure to test it!) and in the form of pellets, about 4mm in diameter
Anything smaller will dramatically reduce the airflow and possibly stall the blower. Do not use granulated carbon
To hold the carbon I used a 16x12" (40x30cm) muslin drawstring bag. But a bag is quite inconvenient as it makes it hard to fill the corners of the chamber, and the drawstring can't easily be tucked in flat when closing the lid. A cheese cloth would likely work better. In any case, it needs to be made of a very low density and aerated fabric, or airflow will be impeded.
The front switch is optional. But I wanted to control the blower from the front, and easily tell if the fan is on or off. So I used a pre-wired SPST 12V LED-illuminated rocker switch for 13x20mm openings. A 120mm blower draws less than 2A, so any 5A switch or up will do.
The blower's quick-disconnect uses standard 5.5x2.1mm barrel connectors, which made inserting the switch super easy, barely an inconvenience. It also facilitates swapping fan /switch / controler without having to cut and resolder wires, which helped a ton during prototyping.
To prevent carbon dust from being blown into the printer some MERV13 filter fabric is laid down on the exhaust side of the carbon chamber. This notably impedes the airflow, but letting carbon dust in would bring a fast end to anything lubricated in the printer, like the bearings riding on the X rail and Y&Z rods, and the ballscrews.
MERV13 fabric uses the same mechanical + electrostatic filtering principles as HEPA filters and N95 masks. And it's much cheaper over time: just cut a piece out of the large sheet each time the carbon gets refreshed, done.
I suspect that MERV11 wouldn't be enough to catch some of the dust, and that MERV16 would kill most of the flow. But these are just a backyard mechanic's guess...
Comments
Post a Comment