Cleaning Up Smoke With An Electrostatic Precipitator

[Steve Dufresne’s] got another great project for us — a device that effectively gets rid of smoke!

It’s called an electrostatic precipitator and it works similar to the way many cars are painted today using a process called electrostatic coating. Electrostatic coating works by giving the paint particles an electrostatic charge, opposite to the charge on the vehicle’s body panels — this makes the two attract and results in using around 95% of sprayed paint — barely any over-spray, and a better bond to boot!

[Steve’s] tried this experiment of creating a smoke precipitator with the eventual goal of using it on a car’s exhaust. He’s been through a few designs so far, and finally has one that works quite well. It’s not even that complicated, just take a look at the following diagram.


As the smoke enters the tube, it passes through a fine metal mesh which is negatively charged from his homemade high voltage power supply. This gives the smoke particles a negative charge, and as they continue to float up they get attracted to a positively charged metal tube. Low and behold — no smoke rises beyond this point. Don’t believe us? Check out the demonstration:

56 thoughts on “Cleaning Up Smoke With An Electrostatic Precipitator

  1. What all does this actually capture? Can it capture all byproducts of combustion? I would like to see more tests with various sensors to demonstrate what it actually captures, and how much of it. Interesting work, I will definitely keep an eye on this.

        1. Just the solid particles I’d think. Anything gaseous or liquid isn’t going to stick to the plate for long. They’re meant for the same thing in industrial chimneys.

          1. The dust particles too can leave the electrode after their charge is neutralized, by simply being blown away by the ionic wind or becoming positively charged and repelling each other.

    1. Depends on the distance between inner conductor and outter tube. I designed a prototype one for an pilot scale plant, to parcipitate nanopartices. I’f I remember the papers correctly, 5kV is usually the minimum to make an effective parcipitator. But you don’t need any current to speak off to develop the corona.

    1. Maybe I’m thinking cars have dirtier output than they do, or that the tests I remember are from older cars. I don’t have one (lucky enough to live in a part of a city where there’s really no need), hence it taking time to borrow a friend’s to try it out.

  2. could a fan be used to pull air through this, then put behind a soldering station to filter the smoke? might be an interesting alternative (though more power hungry) than using a passive filter.

    1. A filter using traditional filter medium that slowly clogs will require a powerful fan. This system, though it will need power for the power supply, only needs a low power fan as all it needs is just enough power to pull air through.

  3. If you car is emitting particles for this to capture then there is something wrong or it is very old.

    We use very large electrostatic precipitators at work to capture airborne coolant mist, the brand is “smog hog”. It charges the plates to about 10kv.

        1. Diesels still do emit particles. The manufacturers largely loophole their way through the emissions standards, as they don’t count the number of particles but total particle mass. As a result, diesel exhaust particle emissions have reduced in mass but increased in numbers due to new diesel cars emitting nanoparticles instead of visible soot.

          Plus, in practice they do smoke when you put your foot down, especially when the engine is cold.

      1. A smoking car can be an easy fix, my neon started smoking and it was just valve stem seals. Made up a tool to remove the keepers with the head still on and had them changed in less than an hour. No more smoke.

    1. The majority of those typically have a mechanical filter, fan, and an ionizer to add “healthy” negative ions to the air. But not a precipitating plate, as that would remove the negative ions. Typically some nearby household item, like the surface the cleaner is on or a wall, inadvertently becomes the plate and requires frequent cleaning. ;)

      You can salvage the ionizer module from them. It’s a separate part and fully potted. You can also sometimes find the modules sold surplus. I have about a dozen of them, most with 120VAC input, and a few with 12VDC. Fun to play with, sometimes useful too. I’ve “misused” them to put static charges on houseplants, Christmas trees, and other items I want to quickly trains pets not to molest. Touching the item is like walking across carpet in socks on a dry day and touching a doorknob, harmless, but pets don’t like or understand it and learn quickly, typically after just one or two zaps. Just make sure the protected item isn’t grounded, that will render it ineffective and possibly burn out the module. Also make sure the item doesn’t have too much capacitance or resemble a Leyden jar, you don’t want to accumulate a harmful charge.

      Back on topic… If you supply your own HV source like [Steve] did, know this. There is a voltage at which point you start producing ozone. The higher you go above it, the greater the ozone production. A little ozone can be beneficial, it oxidizes and breaks down some pollutants that being non-particulate, precipitation won’t catch. But too much and it becomes a pollutant itself, you don’t want to breathe much of it. I forget the exact voltage at which ozone production starts, but I have never seen a commercial ionizer module higher than 7.5KV, and 4-6KV are more common.

      As an example of a very fancy air cleaner, I have an EcoQuest Breeze AT which costs about $300 new, I paid $5 for it at an estate sale. It has mechanical filtration, for which I build my own replacement cartridges to avoid paying the OEM a ridiculous price. It seems to generates negative ions and internally precipitate particulates. It also has ozone output which can be turned off, on, or adjusted. A timed “sterilization” mode cranks out massive amounts of ozone, and requires all living creatures leave the area for a while. I was told the previous owner used two of these in a nightclub to control cigarette smoke, and used the sterilization mode after hours to remove lingering odors from smoke on surfaces. In a sealed and vacated residential room I can say it successfully removes virtually all smoke and pet odors. Theoretically this might destroy some bacteria and virii too.

        1. The Breeze filters are carbon foam, mounted in a metal frame, and covered by a mostly open metal mesh to distribute charge across it (as it’s not just mechanical, but part of the electrostatic particulate collection). The only part of this that wears out is the carbon, so I buy “cut-to-fit” carbon for air filters, of the same thickness, from Amazon. Then gently pry open the metal and swap the carbon. The carbon, although it was visually identical, didn’t work as well as the original at fine particulate removal. I figured the original must be treated to enhance conductivity, and a warning in the manual not to wash this filter with soap provided possible verification. So I sprayed it with a little aerosol laundry static cling remover. That did the trick, good as new OEM! After a rebuild, it can be reused about a half dozen times with minimal loss of efficiency by washing, letting it dry fully, and recoating.

      1. I’ve got a beer fermenting carboy that I cover with a space blanket to help with temperature fluctuations. My cat likes to rub against it.

        This seems like a good idea…

  4. Anybody using this for soldering fumes mediation?

    My lab bench is away from any windows, so I cannot vent fumes outside, and conventional filter-based solutions do not work well. This looks promising.

    1. You should get sick off flux fumes someday just so you know how much is too much. I’ll say this, the wisp coming off a soldering iron tip is no where near enough to worry about. If you dump a pound of flux core solder into a pool of molten lead you might have something to worry about though. Don’t ask me how I know this :)

  5. Can you clean lasercutter smoke with this?

    This sound like a fun project. I have a slightly larger smoke problem to clear though: our lasercutters produce quite a lot of smoke. Some of it gets trapped in our pre-filter and some of it is blown outside by the large exhaust fan. But some of the smoke also precipitates on the fan itself, unbalancing it. I wonder if this method would be a good way to clear most of the smoke before it hits the fan?

  6. This technology has been used for years in residential HVAC systems. Electrostatic precipitation works well with some particles. In lab testing, there are two types of dirt that test filter efficiencies. One is plain brown dust, or house dust. The other is carbon. Carbon is very difficult to ionize and therefore, it is difficult to capture. Typically, carbon must be captured with a mechanical filter rather than electrolytically.

    This is a fun experiment, but will not likely work well for removal of particulates from an automotive exhaust system.

    1. well there goes that idea… Am going to build one for the house HVAC. AM NOT going to build one for the shop. I’ll go with a more standard filter for that or simply evacuate the air to the exterior.

    1. What is the output from your ignition coil like? The esp needs a flat DC output or maybe a high frequency pulsed DC output would work too. I imagine a low frequency DC output wouldn’t work and certainly not AC.

    1. Since a lot of people had been asking me what power supplies work, I recently tried my flyback power supply with a HV diode built in the flyback case. That sounds much like what you’ve tried. It didn’t work for me either. I didn’t measure the output high voltage during the test but based on the input voltage, I suspect it was around 20,000 volts, pulsed DC.

      However, I did further tests with the power supply in the video above. It has a multiplier after the flyback and produces a high voltage flat DC. It’s this one By decreasing the spacing between the cylinder and the mesh to 27mm/1 inch, I managed to get it to work with 21,000 volts, flat DC. Decreasing the spacing any further than that just resulted in sparks.

      So for this geometry of cylinder and mesh, it looks like the minimum voltage needed is 21,000 volts flat DC. Your geometry might be slightly different so you might have a different minimum. You could try decreasing the spacing between your cylinder and mesh, but the fact that it’s pulsed DC might prevent it working, I’m not sure if that’s the case or why.

      Here’s the video of my recent tests.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.