Ever since I started my DIY filter experiments, I’ve been wondering whether I could create a super DIY by using a stronger fan. In the past three months, my collaborator Anna has helped me test a half dozen different fans. She now has a home full of fans:
After months of tests, one fan has proven itself to be the king of fans. I’m now ready to unveil a newer, more powerful DIY:
Informally I’ve been calling it the “大炮,” the Cannon. That describes how it looks, but also what it does to air pollution. Simply put, this thing kicks butt:
Even when you look at the really small stuff–0.5 micron particles–the Cannon is equally impressive:
Over five nights of testing, the Cannon removed an average of 97% of 0.5 micron particles and 96% of 2.5 micron particles. I think this may be close to the upper limit of how much particulate pollution you can remove from your air with a home-use filter.
What’s the difference between the Original DIY and the Cannon?
In 8-hour tests, the Cannon removes 4-13% more particulates.
In the overnight tests, the Cannon is removing somewhat more particulates–especially the smaller 0.5 micron particles.
Remember, this is an average, so the difference is larger on very polluted days and smaller on clean days. (You can see the data for individual days in the raw data at the end of this post.)
The Cannon cleans faster.
Because the Cannon circulates air a lot faster, it reaches those numbers faster than the original DIY.
This is harder to illustrate than average effectiveness, but you can see it clearly in our controlled tests. In the controlled tests, we burn a source of 0.5 micron particles (a stick of incense) in a small closed room. After 20 minutes, we turn on the filter. In those tests, the Cannon is clearly faster:
Now, this speed test is only to demonstrate relative speed in controlled conditions. The speed test room was small–a bathroom. It’s not meant to say the Cannon will clean your whole bedroom in 1 minute. But it does say the Cannon will clean your room faster than the Original DIY.
The Cannon is better suited for very large rooms.
Some people have told me they want to use the original DIY to clean large spaces, like stores and multi-person offices, but the original DIY is not really made for large spaces. If you want to clean a large area, the Cannon has the power to get the job done. (See tests in a 30.5 m2 room.)
Is it still affordable?
The stronger fan is more expensive and heavier, but we’ve still managed to get the price at Smart Air to 450 RMB. At that rate, you’d be removing 97% of your 0.5 micron particulate at just 4.1% the price of the 11,000 RMB IQ Air.
When should I choose the Cannon vs. the Original DIY?
I would recommend the Cannon for:
- People with larger spaces, like large offices or bigger-than-normal living rooms.
- People who are very concerned about air pollution and therefore really value that extra speed and effectiveness–for example, people with health problems, pregnant women, and children.
But for most normal-sized bedrooms (<15m2), the Original DIY will still do the trick.
As always, I’m posting the original data and methods below for fellow data nerds.
[accordion-item title=”Cannon Original Data and Methods” state=closed]
Here is the raw data for the five Cannon room tests.
Calculation method: As always (1,2), to calculate effectiveness, I average the last four hours and divide that by the baseline number. As I’ve said before, I think this method of calculating effectiveness is more rigorous than calculating as reduction from outside because indoor air is cleaner than outdoor air.
Sample timing: Anna took the baseline measurement before she went to bed, and she set the particle counter to take hourly measurements until she woke up in the morning.
In the raw data, you can see that some days the air pollution outside fluctuated a bit. The most stable day was probably 9/25. The effectiveness numbers that day are also very close to the average of the five tests.
Room setup: The DIY filter was on one side of the room. The particle counter was on the opposite side. Anna made this schematic of the experimental setup:
Anna’s room is 15 meters squared. The windows were closed at all times, and the bedroom door was mostly closed before Anna went to bed, and it was always closed during sleeping hours.
Particle Counter: My particle counter gives counts for (a) particles .5 microns and above and (b) particle 2.5 microns and above. For more information on how laser particle counts compare to government machines and AQIs, check out my data comparing the two over 70 measurement occasions.
Interpreting the speed of the drop: Thee time between (1) turning the filter on and (2) the nearest hourly measurement varies a lot based on when Anna took the baseline measurement and when she set the particle counter up for hourly mode. Therefore, I caution against interpreting the exact time it takes the counts to drop too closely.
Instead, to answer the question of how quickly the counts go down, the controlled, minute-by-minute test is much more useful:
Controlled minute-by-minute test
To get the minute-by-minute tests, we performed controlled tests. First we put the DIY filter (not turned on), particle counter, and a lit stick of incense into Anna’s bathroom. She sealed off the fan on the ceiling and the door to keep the environment controlled.
Pollution source: Anna ran tests to determine the length of incense that would burn completely in 20 minutes (8.3 centimeters). On average, burning the incense raises the 0.5 micron counts by 40,000. It raises the 2.5 micron counts by far less–by as little as 100. This demonstrates that the burning incense is releasing particles that are almost all smaller than 2.5 microns.
Turning on the filter: After 20 minutes (when the incense extinguishes), the DIY filter is turned on using an extension cord that goes under the door and outside the room. This means we do not have to open the door, which would affect the air quality and bring undesired noise into the experiment. After that, we let the DIY filter work for the remaining 40 minutes. (Unfortunately, the particle counter sometimes fails to save a few minutes of data at the beginning and end, so the data below is somewhat shorter than 60 minutes.)
Control condition: We also ran tests without a filter (column 1 above) to see whether the numbers drop naturally–even without a filter running. The no-filter test showed that the numbers do drop a little bit on their own (about 9%), but the numbers are quite different when a filter is turned on.
Thomas is an Associate Professor of Behavior Science at the University of Chicago Booth School of Business and the founder of, a social enterprise to help people in China breathe clean air without shelling out thousands of dollars for expensive purifiers.