Building an Oxikit DIY Oxygen Concentrator
29,338
Published 2024-06-14
Errata and Clarifications:
- If you are going to use PVC pipe to store pressurized gas of any kind at any pressure, you MUST use pressure rated pipe and not "thinwall" or "DWV" pipe. I chose Schedule 40 pipe, which is the thinnest walled pipe that has a pressure rating, but there are such things as Schedule 80 and 120 pipe. Using pressure rated pipe does not make it safe to do, it just makes it safer. (Cheers, @Spirit532)
- Some concern has been raised about material fatigue caused by pressure cycling. I'm not convinced that a 30 psi pressure swing meets the threshold to cause this kind of failure, but the concern is probably well-founded. A failure of this kind at 30 psi could be scary, but probably not lethal. The chances are good that a failure would be unexciting, if it ever occurred. PVC is not a brittle material and is also very resistant to oxidative reactions. I will be sheathing my cannisters in something to prevent shrapnel just in case. (Cheers, @NicolasBana)
- Sieve is pronounced /sɪv/ and I know that in any other context (I pronounce it correctly in the idiom "leaks like a sieve") but I learned to pronounce "molecular sieve" from @NurdRage, for better or worse, and his incorrect pronunciation has stuck with me. (Cheers, @nilo70)
- My description of the PSA cycle is not entirely accurate. I state that the equalizing valve is mostly responsible for purging the nitrogen from one tank to another, but that's not exactly true. Most of the purge actually happens via the orifices in the output junction. The equalizing valve theoretically "primes" the unpressurized canister with the last bits of pressurized O2 from its counterpart (because the tank will adsorb nitrogen better under pressure, so if you already have some O2 pressure in there when you start pumping in air, you'll pass less nitrogen as the pressure ramps up) I may address this in a follow-up video, but this is still not a bad introduction to the PSA cycle.
- Some of the PSA cycle diagrams show out of order, mostly when I cycle through them quickly. What I'm describing is still correct and any diagram that is discussed for any length of time is correct, it's just some of the quick ones. This section is a Google Slides stack that I recorded with a screen recorder and at times I advanced the slides in the wrong direction without noticing.
- If it is working properly, the oil in an oil-filled pressure gauge will never contact the process gas. What I describe as a hazard is absolutely still true, however. Please do not use oil-filled gauges in oxygen service. In fact, if I were building this machine for someone else to operate, I would take apart the cheap pneumatic valves, wash them in degreaser, and re-oil them with a fluorinated oil for reactive gas service. Take care to look into the oxygen compatibility of anything that has to carry pressurized O2 (Cheers, @AdvancedTinkering)
- When describing the tool used to agitate the sieves in the OxiKit video, I call it a "pneumatic vibrator." In my head, they used a concrete vibrator. Watching the footage back, they're clearly using an air hammer. If you go looking for that tool, it's called an air hammer. I don't necessarily recommend using it for this, however. They make a point to run it at low pressure and to only contact the fittings with it and not the pipe wall directly, but it still seems easy to damage your PVC this way. Buy a muscle massager instead, you deserve it.
- I describe the gas separation mechanism as "having to do with the relative size of the molecules" which is sometimes the case with molecular sieves but in this case, the mechanism is actually electrostatic and ion exchange interactions! This is why the metal cation (Li vs Na) bound to the zeolite affects the selectivity. (Cheers, @justliberty4072)
Check out the Oxikit project here:
oxikit.com/
You can stay up to date and support me on Patreon here:
www.patreon.com/integratedtherm
And if you want more info on my build, you can read the blogposts here:
www.nickpoole.me/
And check out the Github Repository:
github.com/NPoole/PSA_Oxygen_Concentrator
All Comments (21)
-
Excellent build! We definitely need more oxygen concentrators instead of cylinders. There's one REALLY important thing to mention - you must use pressure-rated PVC pipe if you're going down that route. NOT drain or other unrated pipe! The difference between the two is a very violent explosion that sends large shards a few inches deep inside you. Also, a simple hack to increase your oxygen purity and performance - run the concentrator at max throughput for the desired pressure/purity, then use an overpressure valve(of the not-safety variety) to backfeed the waste oxygen into the inlet when you close the valve on the torch. It helps increase purity when you're not using the concentrator at its limit, which for most small glassblowing will be almost never, unless you use extra torches/burners. It also helps unload the sieves, and since it's already been through the system, extends the life of the air dryer.
-
At the beverage company I work for we’re installing a large one of these to create a pure nitrogen gas flow to replace Co2 in some of our processes.
-
I usually don't interact much on youtube but wanted to leave a comment because your channel is criminally underrated. Great production value and extremely interesting projects, keep it up!
-
I never heard of the Oxikit, think I will build one but just use my shop compressor. Never fails my oxygen tank is empty when i need it. You did a really good explanation of how it works ! You can clean out and use a condensor from a car air conditioner, got mine from junk yard. Far more efficient than the copper pipe. We get a lot of high humidity days so water in the compressor tank was always an issue for me. I run the output of the condensor into 2" black iron pipe that is jacketed with 4" PVC pipe using rubber couplers. It's filled with coolant that circulates when compressor kicks on. It's rare I ever have water in my main air tank since making that. I have a 60 gallon compressor so you may be able to scale down the size of the pipe and still catch a lot of the condensation.
-
I'm not sure why YT graced me with this video, but I'm glad they did! Great project, great video format, and that got you a new sub. 👍
-
When I clicked on this video I wasnt sure what to expect. It sure wasnt Bad Ass Applied Technologies, but here we are and I must say, you do a great job explaining things.
-
lesson learned, always throw in an extra audio recorder as a backup
-
Another thing to note because I didn't hear you say it in your warning about breathing this air: make sure you are using a specific breathing air air compressor. Most air compressors, even oil free ones, are not meant to make breathing air and can be dangerous to breathe from. There are specific kinds of air compressors for these applications you should look for
-
I want to make a really small one of these but in reverse, you can make a fairly inert nitrogen supply using the same hardware. It's total overkill but i want a nitrogen purge on my soldering iron. 😂
-
Really interesting project. I've saved it in my playlist, passed it on to a couple of mates to watch and subbed.
-
Never thought anything good would have come out of Covid. Nice video.
-
For something that got randomly recommended to me on YouTube this is great stuff! Love the content!
-
Awesome video. May the algorithm favor you.
-
What a wholesome build, and great quality of the project and everything. This channel deserves 10x or 100x the audience and I am so glad the algorithm sent me here!
-
That's a damn nice project - should get you going with those oxy-propane burners for sure! I really like the pneumatic and automation design, and placing it all on a dolly is very practical too. Take it right where you need it :).
-
I really love that this is built on a typical hand cart! EDIT: I'd love to see if the atmosphere input would benefit from a heat pump drying pack as is found in a system like the GE ventless Washer/dryer combo.... EDIT 2: You've got me so hooked in with this vid that I keep thinking of things to say... For the sensor hookups that you've yet to use... I bet that data could be integrated as feedback into a controller that gives a more steady/efficient output!!! I'd LOVE to see a video about that!
-
Very very interesting! I was a little lost on your diagrams of the machine, i didn’t get time to find the bits you were talking about, if you highlight your mouse you could use it as a pointer to point to the bit your talking about, which would help the older folk like me :) this was an awesome project. Very very well done bravo!
-
Pro tip- fill the copper pipe with sand....if requiring to create nice round, tight bends or loops. Shape the pipe with the sand filled within it. This will almost completely remove the chance the pipe will crimp down on tighter bends.... you will have nice and smooth and round piping...remove sand when finished.
-
As always, a great video and an impressive build! I also wanted to build a concentrator, but considering the fact that you can get used medical devices for very little money, it wasn't worth it for me. However, if you need a very large or modular concentrator, a DIY build is certainly the better option. Regarding the oil-filled pressure gauges. I thought the oil doesn't come into contact with the gas (if the pressure gauge is intact). Or am I mistaken? Of course, it's still better to completely avoid (regular) oil in a system with oxygen. Anyway, great video!
-
15:40 It's been so long since i've since someone reference steve1989 ! Love what you made, this is a really cool apparatus and professional-grade ! I woukld have added put the safety valve right side up because i don't think having water standing in it is very good for it. I'm curious about the performance of it ! Also, even pressure-rated PVC will have fatigue stress, i've seen a video of a guy doing his DIY water/air separator/dryer combo system entirely out of pressure rated PVC and it had a rapid unscheduled disassembly after a few thousand pressure cycles. At 30 psi it may take a very long time and it will not be extremely violent but still, you need to be careful of it when it happens. Maybe wrap a couple turns of tape aroud the tubes to be extra safe.
