I built this for myself. A few strips of Gorilla Tape and some ebay 4 pack for $45 MERV 15 20x20x2 filters taped together as a support for a 20" box fan from Lowes.
2600 sq ft home, filter is on the floor on the most open area of the house, which covers about 1/2 of the total space (39' Dome Home, 2 story plus a half story master, the area from the living room / dining room / kitchen to the master is all open area)
My air quality meter showed me in the low warning area for PM2.5 and PM10 before I started using this, half a day after starting it up I am in the good range across the board, plus the climate is more uniform throughout the house and I'm sleeping better from the gentle breeze and mild white noise.
I don't know if it would work as well in a boxier home, but a lot of people have said that it does for them as well, so it's probably worth it if you have $65 to burn and the space for the filter cube setup thing.
You can also make this with a box fan, construct a box with 4 filters of roughly the same size as the fan, the bottom of the box can be cardboard. Did this back in 2019 and it works great, though I ended up switching to a Coway a few years later because of the noise.
It's constructed using a regular 3 speed box fan, so it isn't going to be significantly louder than that and possibly quieter depending on how it is shrouded and insulated.
I have a few box fans filters and one cuboid fan, it's not actually the cuboid though, I used a circular hepa filter with the inline duct booster fan.
I leave the fan duct fan running on low (right at the midpoint of the speed dial so technically medium I guess) 24/7 in my kitchen out of eye site because it's so quiet.
So, in my experience, much quieter, quiet enough that people don't notice it running all the time. I have the dial in reach though so I can crank it all the way if something burns on the stove, clears the air out shockingly fast.
I got some sound proof panels and crappily bungee corded them around the fan and it didn't have a noticeable effect. I think if I put more effort into the sound baffles it would reduce the sound, but the whole setup without baffles is quiet enough I haven't bothered.
-EDIT- I should also mention that you don't need 5 filters for this (MERV 13 is expensive), you can use 4 and a piece of cardboard for the 6th side that is touching the ground.
Yeah mine has 4 filters with a box fan on top and cardboard that the box fan came in as the base. All taped together with duct tape. Brain dead simple and cheap. Used some more duct tape to fill in the "corners" of the box fan. Mine blows positive pressure into the box, I just saw some designs that suck negative pressure out of the box, I don't know if that matters, but I've got a PM2.5 meter and mine worked fine last year during the smoke days in Seattle.
I would think that you want to keep dust out of your fan motor/bearings etc. so you want the fan to be sucking out of the filter, not the other way around.
... they'll still function, but if you've ever had one you'll know that dust does matter. They're a pain to clean and the various crap in your air builds up on every edge. Definitely want it sucking filtered if that's that's an option.
I think it matters because with negative pressure you are causing the box to close and cover the gaps, whereas when you blow you are pushing on the box causing the gaps to expand.
I see this all the time, why use multiple 1" thick filters to build a box, when you could just use 1x 4" filter. They are pleated, and the surface area should be equivalent. I documented my setup here before: https://news.ycombinator.com/item?id=28395232
Can confirm this works great. I wish there were more options for higher quality 20x20 box fans though. I’ve had Lasko fans that got loud after a few months, and I assume that’s a bearing going bad.
Having a greater surface area to filter the air through will decrease the static pressure and be nicer on the fan motor. If OP was putting the filters in series, then yes, it should have no effect. But they are putting the filter in parallel.
Agreed, but it might cause your fan to burn out much quicker, or be less productive if you're using a box fan, since the static pressure differential will be approximately 16x between the two.
Imagine calculating all possible routes an air molecule could take to escape after it is blown in. In the box case, the air molecule has 4 potential escape routes instead of 1 in the direct stacked filter case. And each escape route has it traversing 1/4 the amount of filter material.
Respectfully, I think you don't understand how air filters work. Air does not pass through 4x the filter material in a 4 inch filter. Instead, the 4 inch filter is pleated, folded like an accordion, so that each air molecule passes through the exact same thickness of filter as a 1 inch filter of the same MERV rating, but ~3x more surface area allows for ~3x more air to pass through. This is the only reason for making a thicker filter - to allow greater air flow. A box fan is generally operating below the capacity of the 4 inch filter, which is designed to allow for high cfm whole-house airflow.
4 x 1" filters, each with a surface area of 10 square megamicrons, provide a total of 40 square megamicrons when assembled into a box.
A single 1 x 4" filter strapped to a fan gives a total of 10 square megamicrons.
How are those the same surface area?
Ah! I get it, because of the pleating. I guess it depends on whether the greater incoming airflow interferes with itself once it gets squeezed between pleats.
Exactly what I do, but with a 5", 20x25" MERV 16 filter. I run two of these on box fans when it's smokey in the summer, and it works very well. Box fans do get loud though.
> why use multiple 1" thick filters to build a box, when you could just use 1x 4" filter.
Probably uncertainty over how effective fans are when immediately obstructed, and maybe concerns about particles possibly getting deflected backwards if a filter messes with a smooth, laminar flow.
then it seems like a box-fan might effect a laminar-flow further away from the blades (under normal circumstances), but could it be more turbulent near the blades?
The boxed-in design seems to try to dodge assumptions about how air flows near the blades, reducing uncertainty about how turbulent-flows might screw with it.
Note that the OP uses ASHRAE's definition of CADR (nominally airflow rate * removal efficiency) and not AHAM's more complicated definition which accounts for contamination rate.
The article didn't mention the static pressure of the fan. Typically, for a "suck" type fan, you want higher static pressure and there are fans which are designed expressly for this.
What is that calculation based off of? The article lists the 4 filters as $70 (which I'd assume you would replace all at once).
The Levoit uses 1 filter "unit", and filters are $55 for a 2-pack on Amazon ($27.5/ea), or half that if you trust the third-party ones.
Levoit recommends to replace the filters every 6-8 months, so you could get 12-16 months of filtration for $55. I can't find any calculation in the article for how long the author estimates their filters would last.
Levoit filters are advertised as lasting 6–8 months and cost $23 each.
Meanwhile a set of four filters like the one in the tutorial cost $70 and looking up comparables online at that price point indicates they're good for 3 months.
So even if using 4 at a time extends to 12 months, that's still more expensive than the up-to-2 Levoit filters you'd use in that time.
I see zero cost benefit here over Levoit -- it seems to be more expensive actually?
The Levoit mentioned (I have one) is tiny and only good for a small room. The comparison box fan filter can clean a lot more air over the same time period.
It's good for a large room, not just a small one. I have one too.
The CR box is tremendous overkill for something like an apartment. It's gigantic and I don't even know where most people would put one even in a house.
But it's also not what the original post was talking about. The original post is a device akin to a Levoit in terms of size and power.
Seriously. Plus the Levoit gives you adjustable fan speed to balance noise vs effectiveness, set a timer, it tells you when to replace your filter, it looks nice and professional rather than garage hardware, etc.
If it were a quarter the price then I appreciate a DIY thing, but this is already in the same ballpark price-wise as the best-selling air purifier on Amazon. I don't get it.
These sorts of fan setups also use a lot more electricity than a dedicated air filter with similar stats. Not a big deal if you're just running it for a few days, but something to keep in mind.
The Levoit, and basically all 'consumer' purifiers, are extremely under-powered and generally have very low air clearance rates & CFM. Usually relatively/overly expensive filters, too.
To get a remotely apples-to-apples comparison, look at commercial air purification systems -- think devices intended for areas that allows indoor smoking, such as bars/casinos/smoking rooms/etc.
My experience was that a relatively modest 20x20 (x12' high) room requires a massive amount of air movement, and products actually designed for spaces of that size generally run $3-6k or more.
A Corsi-Rosenthal box can delivery very good PM2.5/PM10 numbers, even with challenging environments & when "stretched" to provide purification for an entire home.
PurpleAir or a similar quality AQ monitor provides some very objective data to guide you; I trashed my Levoit purifiers rapidly after realizing just how ineffective (and expensive) they were versus a CR box.
The Levoit moves a lot of air on its highest setting. The filter is $23 and lasts for 6-8 months. It works amazingly and effectively in a bedroom or living room or office. I know, I use them and measure my indoor air quality with an Awair.
The CR box is probably far more powerful, yes, but it's also gigantic. You'd never put it in a bedroom or living room. I'm honestly not sure where you'd put it.
The haphazard assembly of his box fan doesn't look ideal. If you're going to do it that way at least get a 20x20 size furnace filter and use some duct tape to seal it.
The design should also fix the worst parts of using a box fan:
1. Make less noise.
2. Use less electricity.
Point #1 is not the worst part of a box fan for many people, myself included. I have a lot of fond memories from growing up where a box fan was running at night due to the outside heat and lack of central AC. It made sleeping a much easier and more pleasant experience.
I both need the noise and the air (I sleep warm), but I had what had to be a 1950s version of this when I was a teenager , which was just a small enclosed fan that had a top you could regulate to change the sound. Still miss it sometimes.
Yep. Bought one at one of the donation stores for $2. Made in USA, must be about double my age, and still works great. Amazed that a little fan like that can just keep spinning for that long.
Hooked it up to a wifi power switch. Old meets new.
What this and the box fans seem to miss is that prop fans have very little ability to generate static pressure. You will move more volume over the static pressure loss of the filters if you use a centrifugal type of fan. A little looking on Amazon shows those at about 3x the price of the prop boosters for the same rated cfm.
If you add even a little bit of duct to extend approximately 3 duct diameters on the outlet it will make the fan a lot more efficient by allowing the airflow to stabilize and lower buffeting noise at the outlet. There athere are some of the centrifugal fans that include a noise suppressor for more money that's basically a short duct that's double walled and perforated on the inside.
Having tried both, prop fans seem to beat centrifugal fans for the pressure drop across typical filters, and deliver more air per watt through a filter.
I suspect the pleated nature of the filter means that the static pressure drop isn't all that much. Haven't got a manometer to measure it though.
Thanks for the feedback. Thinking about it some more, a general purpose 1" HVAC filters are rated at about .12 in. water column at 300 feet per minute. If you're running them a lot slower by having 4x-6x area, then I could see that your tests would bear out.
I often reference this manufacturer when designing a duct system and selecting filter sizes. This is for their MERV10 filter. MERV8 is lower initial pressure. They have a velocity vs pressure chart on page 4. At 50 FPM it's down to about 0.03 in. w.c.
In a ducted system the ductwork and air terminals reach an economic happy medium that does not really balance with trying to make the filter pressure loss the driving factor.
Cost is a huge concern. A trivial design - filter strapped to a box fan costs $50. A product specifically design for the purpose could cost upwards of $500. For something that most people don't pay attention, it's an unnecessary purchase. With most Americans being unable to afford an emergency $500 purchase, of course a cheap DIY solution would be at the forefront. It's only due to the emergency nature of the need that most people are buying one in the first place, so there's also that.
The only problem is that instead of re-circulating and filtering conditioned air, a furnace fan is going to bring in external unconditioned air (and is only filtered once).
I’ve never seen a furnace fan that pulls in outside air, unless it was some DIY garbage that wasn’t to code, or a huge money waster from 100 years ago.
Most older homes simple have an outdoor draft/damper for makeup air.
You’ll draw in a lot of dirty outside air with that setup if you don’t block/reduce outside air intake. It will still go through a filter, but it’s going to cause the filter to dirty very quickly.
Furnace fans often use a lot more electricity than you’d expect. You can check by looking at your electric meter with it on and off (or maybe by looking at a sticker on the furnace).
Multiple HVAC techs I've talked to have recommended against installing good/high quality filters in a forced air system. They've explained that high-MERV filters reduce airflow which causes the fan to wear out faster and may affect the performance of the cooling/heating. I'm not sure how accurate that is, researching it seems to suggest a MERV 10 to 12 is a good balance, but I've been told to use a MERV 3 filter.
It is not actually as simple as looking at the manual. You need to take ducting layout into account, since there is a lot of air flow reduction from any ducting system, and some can be much worse than others. It definitely does cause your furnace to overheat if you drop below the required airflow rate, something that can lead to short cycling and even premature failure by cracking your heat exchanger. All of this is calculated out using HVAC software when a system is commissioned (in theory and in many building codes, but not so much in practice).
As can be seen, this is an issue during the heating season, not the cooling season - which is when many people are fighting smoke. If you actually pay attention to home maintenance, you can go for the better filter in the summer, and drop down to the minimum or close to it during the winter.
As filters age, they change colour. Mine start white and turn dark grey when old.
However, I noticed that even the outlet side got darker. That made me wonder if a 'blocked' filter was actually allowing more small particulates through (for example perhaps because all the surface area that a bit of PM2.5 could electrolytically stick to has already been stuck to).
Does this sound plausible? Anyone got test results for small particulates of different materials and how the filter performs as it ages?
Not leaks. On filter exit. Air moves around on clean side of filter but it's not a straight line. Dirty air will flow into the clean side from the room due to turbulence of the air on exit.
> If you take the top pick from the Wirecutter and read user reviews carefully, you’ll see that roughly one person a week reporting that their unit exploded.
I mean, if you read colander reviews on Amazon, you'll find people who somehow lost a hand to it.
I love the reviews and comments where you really hope the person left it on the wrong product page, such as the person asking how big reppelent helps with babies teething.
It's usually Amazon's fault. A listing can change the product being sold entirely. I suspect it's a well-rated product that sells its listing to a new product so it starts out with high ratings and popularity (even though the reviews are for a different product entirely).
I see that in the test he compares four filters being used completely in the cube setup to three being used incompletely on a box fan. But that doesn't seem like it can be the entire explanation -- particularly given the reduced airflow of the duct fan.
They have just thrown 3 filters on top of the box fan with lots of gaps on the side. Air will mostly follow the path of lease resistance and not go through the filters. You need a filter that is the same size, 20x20, as the box fan and then the filter(s) will need to be sealed, via duct tape or so, around the fan for it to be effective.
You are correct from an engineering perspective. From a cheap/GSD perspective the box fan filters are easy to build and provide a measurable result at a low cost.
The internet would welcome any new designs you might have. I'm sure OP would be happy to have started a build-off.
Yeah but there are better box fan filters out there. You can do exactly what OP did with his cube and throw a box fan on top of it instead of the duct fan, and with the higher airflow from the box fan combined with the exact same filter setup you'd likely see better results.
Most of the box fan filters I've seen these days look like this cuboid fan- as in, they have four filters, a base, and a fan. Many also include some cardboard curved corners to improve airflow and reduce noise. I kind of assumed that's what we were comparing until I saw the image.
And the construction is just as shoddy. The gaps where air can bypass the filter are important. Most of the air will bypass the filters because the air will go through the path of least resistance.
I couldn't afford this without having my own shop space. The parts are cheap, but after setting up something the size of a mini fridge that's covered in bungee cords in my little apartment, the marriage counseling would be costly.
If aesthetics are more important than cost, I have a blueair 211+ that does a great job and is quite silent. I would buy it again if a DIY filter was out of the question, tho my next filter will probably be a DIY job with a bit of time spent to make it look nice.
142 comments
[ 4.9 ms ] story [ 215 ms ] threadCan recommend.
2600 sq ft home, filter is on the floor on the most open area of the house, which covers about 1/2 of the total space (39' Dome Home, 2 story plus a half story master, the area from the living room / dining room / kitchen to the master is all open area)
My air quality meter showed me in the low warning area for PM2.5 and PM10 before I started using this, half a day after starting it up I am in the good range across the board, plus the climate is more uniform throughout the house and I'm sleeping better from the gentle breeze and mild white noise.
I don't know if it would work as well in a boxier home, but a lot of people have said that it does for them as well, so it's probably worth it if you have $65 to burn and the space for the filter cube setup thing.
I'm currently dreaming up a ceiling fan replacement that has a much lower profile and also cleans the air with a standard furnace filter.
Post: https://www.jefftk.com/p/ceiling-air-purifier
HN Discussion: https://news.ycombinator.com/item?id=31590773
[0] https://cleanaircrew.org/box-fan-filters/
I leave the fan duct fan running on low (right at the midpoint of the speed dial so technically medium I guess) 24/7 in my kitchen out of eye site because it's so quiet.
So, in my experience, much quieter, quiet enough that people don't notice it running all the time. I have the dial in reach though so I can crank it all the way if something burns on the stove, clears the air out shockingly fast.
I got some sound proof panels and crappily bungee corded them around the fan and it didn't have a noticeable effect. I think if I put more effort into the sound baffles it would reduce the sound, but the whole setup without baffles is quiet enough I haven't bothered.
https://www.texairfilters.com/a-variation-on-the-box-fan-wit...
-EDIT- I should also mention that you don't need 5 filters for this (MERV 13 is expensive), you can use 4 and a piece of cardboard for the 6th side that is touching the ground.
Or even all the excess dust kicked up from a digging out old things we didn’t know he still had…
If there’s a belt from the motor to the fan you’re in for a good time.
4 x 1" filters, each with a surface area of 10 square megamicrons, provide a total of 40 square megamicrons when assembled into a box.
A single 1 x 4" filter strapped to a fan gives a total of 10 square megamicrons.
How are those the same surface area?
Ah! I get it, because of the pleating. I guess it depends on whether the greater incoming airflow interferes with itself once it gets squeezed between pleats.
Probably uncertainty over how effective fans are when immediately obstructed, and maybe concerns about particles possibly getting deflected backwards if a filter messes with a smooth, laminar flow.
This is, if we think of two kinds of airflow:
1. [laminar flow](https://en.wikipedia.org/wiki/Laminar_flow ), which is like a smoothly flowing river;
2. [turbulent flow](https://en.wikipedia.org/wiki/Turbulence ), which is like erratic flow inside a mixer;
then it seems like a box-fan might effect a laminar-flow further away from the blades (under normal circumstances), but could it be more turbulent near the blades?
The boxed-in design seems to try to dodge assumptions about how air flows near the blades, reducing uncertainty about how turbulent-flows might screw with it.
Some details about AHAM's CADR calculation are here: https://frdmtoplay.com/nagivating-air-purification/
Or don't let the perfect be the enemy of the good -- buy 10x box fans for cheaper, because they're manufactured and shipped in the millions of units.
If it's not going to catch on fire, why are we worrying about the longevity of something that costs $30?
They should list a CFM at Inches of Water Column in the 1/4 to 1/2" range to be considered "high static" pressure.
Those blowers from the other comment are like for your iron smelter.
The Levoit uses 1 filter "unit", and filters are $55 for a 2-pack on Amazon ($27.5/ea), or half that if you trust the third-party ones.
Levoit recommends to replace the filters every 6-8 months, so you could get 12-16 months of filtration for $55. I can't find any calculation in the article for how long the author estimates their filters would last.
Levoit filters are advertised as lasting 6–8 months and cost $23 each.
Meanwhile a set of four filters like the one in the tutorial cost $70 and looking up comparables online at that price point indicates they're good for 3 months.
So even if using 4 at a time extends to 12 months, that's still more expensive than the up-to-2 Levoit filters you'd use in that time.
I see zero cost benefit here over Levoit -- it seems to be more expensive actually?
The CR box is tremendous overkill for something like an apartment. It's gigantic and I don't even know where most people would put one even in a house.
But it's also not what the original post was talking about. The original post is a device akin to a Levoit in terms of size and power.
If it were a quarter the price then I appreciate a DIY thing, but this is already in the same ballpark price-wise as the best-selling air purifier on Amazon. I don't get it.
To get a remotely apples-to-apples comparison, look at commercial air purification systems -- think devices intended for areas that allows indoor smoking, such as bars/casinos/smoking rooms/etc.
My experience was that a relatively modest 20x20 (x12' high) room requires a massive amount of air movement, and products actually designed for spaces of that size generally run $3-6k or more.
A Corsi-Rosenthal box can delivery very good PM2.5/PM10 numbers, even with challenging environments & when "stretched" to provide purification for an entire home.
PurpleAir or a similar quality AQ monitor provides some very objective data to guide you; I trashed my Levoit purifiers rapidly after realizing just how ineffective (and expensive) they were versus a CR box.
The Levoit moves a lot of air on its highest setting. The filter is $23 and lasts for 6-8 months. It works amazingly and effectively in a bedroom or living room or office. I know, I use them and measure my indoor air quality with an Awair.
The CR box is probably far more powerful, yes, but it's also gigantic. You'd never put it in a bedroom or living room. I'm honestly not sure where you'd put it.
https://www.amazon.com/Adaptive-Sound-Technologies-LectroFan...
Hooked it up to a wifi power switch. Old meets new.
That level of noise is pretty bad for most of the rest.
In order to meaningfully clean the room's air? Not on one PC, with any reasonable number of fans.
If you add even a little bit of duct to extend approximately 3 duct diameters on the outlet it will make the fan a lot more efficient by allowing the airflow to stabilize and lower buffeting noise at the outlet. There athere are some of the centrifugal fans that include a noise suppressor for more money that's basically a short duct that's double walled and perforated on the inside.
I suspect the pleated nature of the filter means that the static pressure drop isn't all that much. Haven't got a manometer to measure it though.
I often reference this manufacturer when designing a duct system and selecting filter sizes. This is for their MERV10 filter. MERV8 is lower initial pressure. They have a velocity vs pressure chart on page 4. At 50 FPM it's down to about 0.03 in. w.c.
https://www.glasfloss.com/product/zl-hv/
I think a centrifugal could still be an advantage for noise while having a reasonable dimension, but your feedback is sincerely appreciated.
A 10x increase in surface area would mean far far less powerful fans are needed, and the filters could last a decade before needing to be replaced.
I can imagine many theoretical reasons like sourcing and availability but are there things beyond actual conjecture here
https://www.usa.philips.com/c-m-ho/air-purifier-and-air-humi...
For many people on HN they won't get it. $500 isn't a lot for an appliance right? For many that's two months groceries...
I’m actually very curious about more intelligent systems. Those that can:
- bring in outside air on occasion to freshen things up
- negotiate the best way to cool a home depending on outside air. It drives me mad when my AC kicks on and it’s actually colder outside.
All HRV's do that. "single room recuperator" from aliexpress is in every flat after renovation in Lithuania.
You’ll draw in a lot of dirty outside air with that setup if you don’t block/reduce outside air intake. It will still go through a filter, but it’s going to cause the filter to dirty very quickly.
MERV 3 is really really minimal filtering. Your furnace almost certainly is designed to take better filters than that.
As can be seen, this is an issue during the heating season, not the cooling season - which is when many people are fighting smoke. If you actually pay attention to home maintenance, you can go for the better filter in the summer, and drop down to the minimum or close to it during the winter.
As filters age, they change colour. Mine start white and turn dark grey when old.
However, I noticed that even the outlet side got darker. That made me wonder if a 'blocked' filter was actually allowing more small particulates through (for example perhaps because all the surface area that a bit of PM2.5 could electrolytically stick to has already been stuck to).
Does this sound plausible? Anyone got test results for small particulates of different materials and how the filter performs as it ages?
I mean, if you read colander reviews on Amazon, you'll find people who somehow lost a hand to it.
Historically, Amazon doesn't seem to care.
I see that in the test he compares four filters being used completely in the cube setup to three being used incompletely on a box fan. But that doesn't seem like it can be the entire explanation -- particularly given the reduced airflow of the duct fan.
- Tape (free).
- Two pieces of packaging foam (free).
Not that those are expensive, but I'm not sure why the link says they are free.
So people will build heirloom-quality furniture pieces for "free" with "scraps."
But, as a DIYer myself, a lot of things do accumulate for Actually Free, you just have to watch out for good materials.
They have just thrown 3 filters on top of the box fan with lots of gaps on the side. Air will mostly follow the path of lease resistance and not go through the filters. You need a filter that is the same size, 20x20, as the box fan and then the filter(s) will need to be sealed, via duct tape or so, around the fan for it to be effective.
The internet would welcome any new designs you might have. I'm sure OP would be happy to have started a build-off.
Most of the box fan filters I've seen these days look like this cuboid fan- as in, they have four filters, a base, and a fan. Many also include some cardboard curved corners to improve airflow and reduce noise. I kind of assumed that's what we were comparing until I saw the image.
Like these from another thread: https://cleanaircrew.org/box-fan-filters/
This is not a perfect device, but it will capture more particulate than nothing.
Next time I’ll get one with washable prefilters even if the final one is disposable.