To say that a lot of things have happened in the year since the beginning of the COVID-19 pandemic, this is an understatement of epic events, so much so that it is difficult to remember the early days of the hardware hacker community that used mass-produced PPE reaction. , Homemade ventilator and so on. However, we do not remember that there were too many attempts to build this DIY oxygen concentrator during the initial expansion phase.
Given the simplicity and effectiveness of the design called OxiKit, it seems strange that we have not seen more such devices. OxiKit uses zeolite, a porous mineral that can be used as a molecular sieve. The tiny beads are packed into a cylinder made of PVC pipes and fittings from a hardware store, and connected to an oil-free air compressor through a pneumatic valve controlled by a number of solenoid valves. After cooling in the copper tube coil, the compressed air is forced to pass through a zeolite column that preferentially retains nitrogen while allowing oxygen to pass through. The oxygen stream is split, one part enters the buffer tank, and the other part enters the outlet of the second zeolite tower, where the forcedly adsorbed nitrogen is released. The Arduino controls the valve to alternately flow the gas back and forth to produce 15 liters of 96% pure oxygen per minute.
OxiKit is not optimized like commercial oxygen generators, so it is not particularly quiet. But this is much cheaper than a commercial unit, and for most hackers, it is easy to build. OxiKit designs are all open source, but they do sell toolkits and some difficult-to-procure parts and consumables, such as zeolite. We will try to build something like this because the technology is so neat. Having a high-flow oxygen source is not a bad idea either.
15 liters per minute seems very impressive. In terms of scale, it is enough to sustain the lives of 7 people under normal circumstances (each person @ 2 liters per minute).
I have always wanted to know how these work. Interesting. It seems to almost violate the laws of thermodynamics, but it is not the case.
With such a large amount of oxygen produced, I want to know what will happen if you hang this baby on a car engine and/or enlarge it. It may be like nitrite. This will be quite safe, because you can set it up so that the “pure” oxygen produced is consumed immediately near the engine instead of being stored anywhere. However, I need to adjust the car first. Backfired… “It will be bad.”
I think this is good for welding/brazing/cutting of oxygen/propane, oxygen/hydrogen or oxygen/acetylene.
Yes, after I watched this video, YT popped up Dalbor Farny’s suggestion video on the O2 concentrator. The purpose is to provide the oxygen fuel torch he needs for the glass blowing lathe. Manufacture your own customized digital tube. In fact, six of them combine to produce 30 lpm O2.
I guess a 2-liter engine running at a few thousand RPM might consume the 15-liter engine instead of 1 minute. However, could this increase the oxygen level in the intake air to a sufficient level? really do not know
Nitrite can provide energy because it releases a nitrogen molecule for each decomposed nitrous oxide molecule (it maintains its volume as oxygen is consumed), just as it increases the effective oxygen concentration ( Release will also give off heat). Pumping pure oxygen is not that beneficial, because you still lose volume and have to deal with issues that might ignite the engine block.
You will need to scale up seriously. A 2-liter car engine with a speed of 2500 rpm “breathes” approximately 2.5 cubic meters of air per minute (21% O²). It is about 600 times that of a human being at rest. The respiratory volume consumed by humans is about 25% of O², while the respiratory volume consumed by cars is about 90%…
It also burns very hot and molten pistons. By tilting the mixed fuel, you can actually get more power from any engine. But the piston will melt due to the increase in heat. The lower oxygen content prevents the metal from melting.
Ordinary car engines are restricted by airflow and will produce maximum power when combusting all the oxygen in the air. This is achieved by slightly enriching the mixture, which does not burn some gasoline. Unless the maximum power is required, car engines usually run at a slight tilt, because fuel-rich operation means reduced fuel economy and increased hydrocarbon pollution.
If you want to use this feature to increase power, you need a way to trick the engine computer into adding a certain percentage of fuel at the same time.
If you can keep the air-fuel ratio constant, it is roughly similar to opening the throttle by only a few percent.
However, if you exceed “a few percent” (intentionally ambiguity…), you may reach the limit of the ECU’s ability to understand how much air enters, or control how much fuel flows out, or set the correct ignition timing regardless of What speed and airflow are you using.
The flow rate needed to keep someone alive depends largely on their condition! 2 l/min is fairly simple. Many patients requiring intensive care require 15 l/min.
Just be careful to run out of oxygen. High concentrations of oxygen can make many things flammable and promote spontaneous combustion of many oils and lubricants. This is why they use oil-free compressors.
That, and many other “not immediately intuitive” O2 processing methods can hurt you, especially under increasing pressure.
If you are playing O2, you can use Vance Harlow’s Oxygen Hacker’s Companion (nitrox divers may already have this companion): http://www.airspeedpress.com/newoxyhacker .html
I don’t know the book, it’s the user, not the tuner. However, thanks for your reference, I will order a copy as soon as the form becomes effective!
Yes, I will mention. The failure mode of PVC compressed air is a shrapnel explosion, so watch these pressure ratings carefully-as the diameter of the pipe increases, the pressure rating will decrease.
In the early 1980s, I worked for a medical equipment leasing company that leased and serviced Devilbiss oxygen generators. At the time, these units were only the size of a small beer refrigerator. I clearly remember the “hardware storage” nature of its internal structure. I still remember that the sieve bed was made with 4-inch PVC pipe and cover, so the structure described in this project is consistent with previous historical (but obviously practical) technology.
The compressor is a double-oscillating piston/diaphragm type, so there is no oil in the compressed air. The valve in the compressor head is a thin stainless steel reed.
Stream sorting is done by a mechanical timer, no Arduino is required. The timer has a synchronization (clock gear motor) that drives a shaft with multiple cam wheels. A micro switch riding on the cam fires a solenoid valve, causing the gas to move around.
The biggest enemy of these machines is high humidity. The adsorption of water molecules destroys the sieve bed.
Just before I left the company, we started to acquire a concentrator from a competitor of Devilbiss (the name is now unknown to me), and the company has shown great progress. In addition to the smaller and quieter new concentrator, the company also built the sieve bed using aluminum tubes. The tube is covered with a plate with machined grooves for O-rings. I seem to think of the full-threaded support that combines assemblies. The advantage of this design is that if necessary, the bed can be separated and the sieve material can be replaced. They also eliminated mechanical timers and replaced them with simple electronic devices and SSRs to trigger solenoids.
They do require the use of SCH40 piping (rated pressure 260psi @ 3″) and are clearly equipped with a 40psi safety valve and a 20-30psi regulator before the PVC is pressurized, so there is a good safety factor. Not sure how it will be exposed to O2 Change the intensity.
The burst pressure of the SCH40 is many times the rated pressure-depending on the diameter. A 3-inch pipe is approximately 850 psi, and a 6-inch pipe is approximately 500 psi. 1/2 inch is close to 2000 psi. Double the number of SCH80. This is why PVC tennis launchers don’t explode-too many. Enlarging them to a 6 or 8 inch combustion chamber will increase your luck. But in general, the hacker community tends to seriously underestimate the strength of plastic piles. https://www.pvcfittingsonline.com/resource-center/strength-of-pvc-pipe-with-strength-chart/
I would be interested in reducing the amateur’s ability to use fireworks (and possibly purity). The hobby market usually buys retired medical oxygen cylinders. That was my first idea, but the cost of the kit + BOM far exceeded the price of a retired medical unit.
A 2 liter car engine can consume 9,000 liters/minute of oxygen (high speed), so 15 liters/minute of oxygen is about 600 times shorter. , This is a cool device. I bought several refurbished concentrators of 5 liters per minute for $300 each (the price seems to be rising). It produces 5 liters/minute. A few hundred watts are used, so it is extrapolated that 9000 liters per minute (for entertainment purposes only) requires approximately 360 kW (480 hp).
Because their algorithm was written by the Berlin band. (Calculate one and you will get a gold star.)
Check out the company’s website… well, the specifications in their store are a bit vague, but they will sell you 5 pounds for $75.00. So let’s take a look at github. Do not. There is no BOM there.
We have an open source electromechanical design that can tell you how to build it instead of how to fill it. I call this a place where key information is missing. It’s like a character raises eyebrows… it’s fascinating.
OxiKit mentioned in a comment on one of their videos (the one I linked to in the story, namely IIRC) that this is sodium zeolite.
Just like any other molecular sieve, you tell the manufacturer what you want to use it for, not what it is for. Because they are the same thing, but the aperture is different.
O2 concentrators usually use 13X zeolite 0.4 mm-0.8 mm or JLOX 101 zeolite, the second is the most expensive. When rebuilding the craigslist o2 concentrator, I used 13X. The green light is always on, so the purity of o2 is at least 94%.

https://catalysts.basf.com/files/literature-library/BASF_13X-Molecular-Sieve_Datasheet_Rev.08-2020.pdf

5A (5 angstrom) molecular sieves can also be used. I think it is less selective for nitrogen, but it can still be used.
There is a good animation on Wikipedia that can intuitively help you understand the working principle of the device: https://upload.wikimedia.org/wikipedia/commons/7/76/Pressure_swing_adsorption_principle.svg I compressed air input A adsorption O oxygen Output D desorption E exhaust
When a zeolite column is almost full of nitrogen, all the valves are turned over to release the nitrogen adsorbed by the column.
Thank you very much for your brief explanation. I have always wondered if the nitrogen generator can be used for DIY projects of nitrogen welding at home. Therefore, the waste output of the oxygen concentrator is basically nitrogen: perfect, I will use it in my lead-free soldering station.
Indeed, for amateurs, it is very useful to be able to convert air into mostly pure oxygen and mostly pure nitrogen. I want to know if you can use “mostly nitrogen” as a shielding gas for welding.
For TIG (also known as GTAW), since the plasma plume is very sensitive, I am not sure. Argon gas is mainly used, sometimes with a little helium gas to penetrate into materials such as aluminum and titanium. The flow is about 6 to 8l/min, which may be too large for a standard compressor.
For welding, it must be that the major welding station brands all sell nitrogen shielding gas for rohs production, but the price of the kit is between 1-2k euros. Their flow rate is about 1l/​​min, which is very suitable for molecular sieves. So let’s assemble some hardware and do flux-free lead-free soldering at home!
Welders want to be able to use pure nitrogen as a shielding gas. It is cheaper than argon or cheaper helium. Unfortunately, it is sufficiently reactive at the temperature reached by the arc and tends to form undesirable nitrides in the weld.
It is used for welding shielding gas, but only a small amount can change the characteristics of the weld.
Obviously, it is feasible to use it in laser welding, but even a well-equipped fab may not have this function.
Therefore, in theory, at least one PSA can be used to reduce nitrogen, and then another PSA (using another zeolite) to reduce oxygen, leaving a higher concentration of substances that are neither oxygen nor nitrogen.
When you are right, at that point, I suggest that you condense the air and then distill it to separate out the gas you want/unwanted.
@Foldi-A folding point in terms of energy input and gas output. I totally agree that the efficiency will be much higher on a larger scale because you can use evaporation for pre-cooling.
But on a very small scale, you will have 1 compressor, 4 zeolite towers and a bunch of electronic pressure valves and the initial cost of a cheap controller (The Brain), which I think will be less.
@irox can by analogy with certainty, but no one using 2 liters of oxygen will quickly die/deteriorate without getting oxygen. For comparison, our intensive care unit (ICU) patients who have secondary high flow due to COVID, get 45-55L when FIO2 is 60-90%. These are our “stable” patients. If there is no high flow, they will definitely deteriorate quickly, but they will not be so sick that we will be intubated. You will see similar or higher numbers for other ARDS patients or most other situations that require a larger nasal cannula than a conventional nasal cannula.
For me, usage is a niche. This can reasonably keep 2 patients at a pressure of 6-8 L, which is actually a place where high flow is irradiated above the conventional nasal cannula or NIPPV. I would like to say that this is very effective for a small hospital with limited oxygen supply, and can provide medical services to patients with chronic diseases in short-term emergency situations.
Does the patient consume 6 liters (or 45-55 liters) of oxygen per minute, or is it partly lost, exhaled to the environment or something?
My background/experience is just a limited life support system for healthy people (with carbon dioxide removed and about 2 liters of carbon dioxide added per person per minute), so thanks to the number of medical uses, this is an eye-opener!
It is important to remember that they are taking oxygen, because their lungs are very cramped when taking oxygen. Therefore, compared with the theoretical needs of the human body, the cost is very high, because in fact, very few people enter.
I don’t know if the person who spoke was the one who designed it, but this does not match the way he described it. Molecular sieves and zeolites do not trap N2, they can trap O2. To capture N2, you need a nitrogen absorber, which is a completely different animal. The sieve traps the O2 under pressure while the nitrogen continues to pass through. This must be correct, because when you release the pressure and use it to dump the N2 in another column, it makes no sense to try to remove the N2 with N2. . These are pressure swing adsorption units (PSA), they work by trapping O2. Higher pressure and larger cylinders can bring higher efficiency (4 cylinders have an efficiency of up to 85%). This does condense O2, but it doesn’t work as he says (or the article says)
You must provide the requested information source, because you can absolutely adsorb N2 on 13X and 5A zeolite molecular sieves. http://www.phys.ufl.edu/REU/2008/reports/magee.pdf
The Wikipedia PSA article also confirms that the zeolite absorbs nitrogen. https://en.wikipedia.org/wiki/Pressure_swing_adsorption#Process
“However, it is much cheaper than a commercial unit.” Since the BOM exceeds $1,000, it is difficult for me to support this statement. The bill of materials for household (non-portable) commercial concentrators costs close to 1/3, is easy to find, and requires no labor. I know 17LPM is cool, but no one outside the hospital will request such traffic. Anyone with such a request is about to check out or be intubated.
Yes, this is a cool project, but yes, its cost-effectiveness is negligible to a certain extent. In Australia, the new 10l/pm equipment is only about $1500AUD. Assuming that $1000 is U.S. dollars, this reduces the cost of purchasing new equipment.
Before the pandemic, I bought one on eBay at a price of around £160 with a flow of 1.5 liters per minute at a price of 98%. And this thing is much quieter than this one! In this way, you can really fall asleep.
But having said that, this is a huge effort. Put it in the room next to the long pipe to avoid noise and explosion hazards…
I want to know if it is possible for you to use it as an almost pure nitrogen source, in protective environments or even in welding?
How about nitrogen-filled tires. Considering the fees they charge for this service, nitrogen must be very expensive…
The next step may be interesting-get the output of this concentrator and separate a 95% O2 + 5% Ar mixture. This can be done by kinetic separation using the CMS molecular sieve in the PSA system. Then set up a 150 bar pump to fill the argon cylinder.
Now, we only need someone to perform the Linde process at home to have real explosive fun
By using our website and services, you explicitly agree to our placement of performance, functionality and advertising cookies. Learn more