Oxygen Concentrator Repair
I had quite a large amount of glass left over from a previous project so I thought I would give glass working a shot. The type of glass I have is borosilicate which is also known as hard glass, it requires an oxygen rich flame. One option is to use bottled oxygen and the other is to use what is known as an oxygen concentrator. An oxygen concentrator draws in air and then outputs oxygen with around a 90 to 95% purity. The great thing is that these machines are used in all hospitals and nursing homes meaning that I'm likely to find a deal.
I have a source that will sell me them for $50 CAD each, labeled as spares and repair. These machine normally start around the $1000 CAD mark new and sell for around $300 upwards on the used market. The fact is that these machines are very basic and there's very little to go wrong. Another point is that they are built for the medical industry which means that they are normally replaced after a certain number of hours have been met, after this they are scrapped even when in perfectly good working order.
There are a few conditions when buying one otherwise it could cost more than it's worth to repair. The first is to make sure that it turns on and that the motor starts. The rest is just a visual check, if possible the cases should be removed to ensure all components are present. Case are normally held on via two cam-lock screws so it's relatively easy to check. The most common problem with these units is an oxygen warning light, this can be caused by a few different issues but all can be solved easily.
Mechanics
So firstly what is a unit compromised of and how does it work? The technology is very simple and there are few working parts in the whole of the machine. Air is first drawn in via a pump, it is normally a piston pump and will always be lubricant free. The air is directed into a molecular filter via a valve, the filter will only allow oxygen and other smaller molecules to pass through whereas the nitrogen molecules will get stuck. Air is almost 78% nitrogen, 21% oxygen and 1% other. After a few seconds the filter will begin to saturate and the pressure will build, normally to around 30psi. A valve will open and vent this nitrogen to the atmosphere. There are two filters which means that while one is filling the other will be venting, this allows a constant supply of oxygen.
After the filters the oxygen will go to a buffer tank which keeps the pressure constant. The oxygen is then passed through a pressure regulator which is set at 8psi. The oxygen passes through a sensor to check the concentration, if not at 90% then a warning light will illuminate. The molecular sieve is normally made in two parts, one is a desiccant (around 10 to 20%) and the rest is the molecular sieve. The main molecular sieve is also a desiccant but it's much more difficult for it to release moisture, a cheat way is to cook them in an oven otherwise they will have to be replaced.
If a unit has not been turned on for even a day it will have an oxygen warning light. The main reason is that the sieve is already saturated with nitrogen and will take a few minutes for it to vent and allow oxygen to pass. Another reason could be that the desiccant is full of moisture and blocks the flow of air, again the machine just needs running for a couple of minutes.
What I Bought - Vision Aire 5
Here are the two units that I bought, they are Vision Aire 5. The reason I went for them is because they were compact, in good condition and black. One thing I really hate with medical equipment is that it always looks either dirty, faded or old because they are a horrible cream colour. Anyway they were shown to be working, each with 15,000 hours, both had the oxygen warning light and both had a repair label.
Since they both had a repair label I thought it would be wise to give them a visual on the inside before running them. They were both a little dusty so I gave them a clean, other than that they looked to be in great condition. I powered them both up and noticed straight away that one of the fans did not work, I shut that unit down. After around five minutes the oxygen warning light went out meaning I just paid $50 CAD for a perfectly functioning unit. I did a quick comparison of the two and noticed that the perfect unit had a 110V fan spliced into the supply whereas the other unit had a 48Vdc fan connected to the main board. It's obvious that this was a design fault, instead of changing the board it was much easier to use a different source for the fan.
The next step is to remove the fan. For what I will be doing with this unit I will also do the same with the functional unit. I should make sure all filters are clean, no dust on the internals and check for splits in all the tubes. This unit is a bit of a jigsaw puzzle, the only way to remove the fan is to remove the compressor. To access the compressor I had to remove the front panel, by enclosing the compressor it helps to reduce sound.
Here is the compressor, note that it is sat on springs. These help to reduce noise, vibration and in turn the service life of all the components in the unit. The polystyrene pads are to prevent damage and also keep the compressor located on the springs. When the motor is turned on or off it will cause the compressor to knock against the walls.
The compressor was very simple to remove since nothing held it in place. A quick inspection and the compressor looked to be in perfect condition. Something to note is that all oxygen systems have to be free of any substance combustible such as oil or grease. These compressors run dry their whole life and literally need no maintenance until they ultimately fail.
Here is the exposed fan.
The fan is sandwiched between the plastic box that held the motor and the frame at the rear. This box has to be removed to allow access.
The first step was to remove anything connected to the box, it just so happens that this is where the control board is fastened.
The board was left in place and just the wires and tubes were removed.
At the rear of the unit is the pneumatic system. The valve assembly is also connected to this box, it is much easier to remove a couple of hoses than to remove the valve bodies.
The box then slid out to reveal the cooling fan.
The original cooling fan rated at 48V, 3.2 Watt.
I got a replacement 110VAC fan, the difference is that it's metal and rated at 25W.
I soldered the fan supply wires to the terminals behind the power switch. It must be noted that if wires are to be soldered then one end must contain a connector, that be on the fan.
The fan held in place via a piece of foam, it also helps to reduce vibration and noise, something screws would unlikely achieve.
The unit was then assembled in reverse order.
The unit powered up and ran without an issue. The original purpose of these concentrators were to supply the oxygen to a glass working torch. They performed perfectly however the glass working part was a disaster, I sold the all of the equipment shortly after.
AirSep Newlife Elite
Since I had a local supplier of these units and managed to make a profit on the last units I thought I should give it another go. This particular unit is extremely popular and what will be seen in most nursing homes. The unit is almost double the weight, height and uses 25% more power than the previous units I serviced. Strangely both units give out the same oxygen output, to me it is clear which I would prefer to buy. This unit again was for spares and repairs but seemed to be in perfect working condition. The only thing apparent which I have not shown in the images is that it was covered in bodily fluids, nice !
Removing the cases shows the internals are very similar to the last units. A notable difference is the molecular sieves, these are twice the length. The sieves are normally made up a zeolite, maybe the other units used a more efficient type of material.
A lot of dust ! This is the pressure regulator for the unit, this should be set at 8 psi. The other two pipes are used in the calibration and testing of the unit, this could be pressure and also the efficiency of the sieves - if they need to be replaced.
More dust ! The majority of these sieves have a bolt running through the centre, this holds the end caps in place. A lot of these units are not in service long enough to even need the media replacing, it is much more common to see sealed units in newer models. Another reason to use sealed units is because they cannot be serviced meaning the supplier can still make money from the product, being able to replace the media yourself does not benefit the manufacturer.
Even more dust ! A lot of these units will contain a battery for safety reasons. If there is a power cut to the device, then an alarm will sound indicating the unit is not able to supply oxygen. The alarm is also used to tell the user about other issues with the unit.
To the compressor, I first removed the supply hose and the electrical wires.
The compressor is spring mounted on a tray, removing two simple screws allows the tray to slide out.
The output from the compressor has to be unbolted, it is very much more common to see a metal tube. In the process of compressing air heat is produced, this metal coil helps to dissipate a little of the heat.
The compressor. You'll notice that all of the compressors will have four wires, two for each coil. A capacitor is used to shift a phase in one of the coils, the motor then therefore acts like a two phase motor. This method is simple and reliable since there are no moving parts, if a motor fails to start then 99% of the time it will be the start capacitor, this must be matched exactly.
Here is an output filter for the unit, it should be washed or replaced.
These motors have a very long life span and literally require no maintenance.
Here I have removed the compressor from the tray just so that I can give it a good clean.
I chose to remove the top case off the compressor. There will be two variations to a compressor in an oxygen concentrator, one will have the inlet and exhaust on the top whereas the other will have the exhaust on the top and the inlet through the crank case (The first units). This compressor has them both on the cylinder head, the valves being reed valves, these are the standard choice for almost all low pressure compressors. A stuck reed can cause a large drop in pressure, the chance of one sticking is low but it can be caused by a deteriorating piston seal.
Here is the head, you can see a build up of black soot. This dust is the residue from the piston seal, this is what can cause a reed valve to stick open.
Here are the pistons, they probably only have a stroke of about 10mm. Reducing the distance traveled on the piston reduces the amount of heat produced and also reduces seal wear. The only reason to go for a longer stroke is to achieve a higher pressure and a more efficient use of the pump. Low compression ratios achieve low heat, low wear and low pressure whereas a high compression ratio introduces more heat, more wear and potential for a higher pressure. These pumps have to run dry and use no lubrication at all, this is the reason they have to be low compression.
Here is the cylinder and the valve plate. A quick clean but nothing crazy, the seal leaves a coating in the cylinder which reduces wear. A great way of reducing seal life is by cleaning the cylinder liner, so try not to go clinical.
Assembling the pump back to the tray I use a thread lock, this is a must on vibrating parts.
Here is the intake filter, it is sealed so cannot be cleaned properly. This should be replaced, however if it is for hobbyist use then I would be inclined to blast air backwards through the filter, however this will only remove a little dirt. A common fault on these units is a low pressure alarm which is caused by this filter being blocked, just look at the first couple of pictures of how much dust has been accumulated in the unit. The pump was placed back in the unit and all of the correct lines were hooked in place.
I removed the front face of the unit in order to remove all of the dust. The white puck is known as a HEPA filter which is used to filter out fine particles, these filters can cause a high pressure alarm or a low output if they are blocked. It is hard to tell how dirty they really are inside so should always be replaced through regular maintenance.
Almost completely dust free. It must be noted that care should be taken when vacuuming up dust as this is very likely to cause a static charge. Sometimes a great way of damaging a circuit board is by using a vacuum on it, clean compressed air is the only method that should be used to clean electronics.
The unit is clean as a whistle, literally not a spec of dust in sight.
The unit looking like new again. I would always advise to wear gloves when handling or cleaning medical equipment, this one for example was covered in nastiness, I certainly don't want to get sick.
I turned the unit on to find that it was really quite noisy, a quick check with the db meter and it was well within standard range. The two previous units I worked on were super quiet, why would anyone go for a heavier louder model.
Just for a test I wanted to see how quiet I could make the unit,
I got some acoustic foam.
I padded the underneath of the motor where it vents to the outside. This foam is known as open-wall meaning that it can breath.
Here you can see that the unit is much more quieter. However, the stock unit did not come with this foam so I chose to remove it. I don't ever intend on this unit being but back into medical circulation, but it is always best to cover your back, never modify medical equipment.
I hope this was interesting to read. This page was by no means an instructional or how-to fix an oxygen concentrator. It is simply part of my portfolio on showing the different items I am able to understand and repair. This unit is not intended on being introduced back into the medical field, it is intended on being use as an oxygen supply for glass blowing torches.
Respironics Millennium
Here is another pretty common model.
The only thing I don't like about it is that it is so bulky. The things I do like about these particular units is that they do not feel flimsy as all of the cases are bolted together. The plastic also feels to be of better quality and is less likely to become discoloured due to it's colour and texture.
The inside is nicely laid out and everything is easy to access.
The construction is a little strange since there are only two valves whereas every other unit has had five. I would have liked to take it apart to find out how it worked but I did not want to risk damaging any gaskets since it was in such good condition. I would assume that the top two valves are bi-directional and that in the bottom of the unit where the oxygen passes will be some heavy one way valves, they allow flow when a certain pressure is hit. There was some light dust in all of the unit but not a single spec on or inside of the filter meaning it recently must have had a filter change.
The cover was removed to access the compressor, again a little different as this is a single cylinder. The reason to use two in the other units was to maintain a constant air flow, this unit achieves this by having a ballast / buffer tank and restrictor.
All of the filters in this particular unit are so much easier to access.
The majority of the units on the market are rated for 5LPM at around 90% oxygen concentration, it is quite common for units to be able to flow up to 8LPM.
This unit is in perfect working condition and has a low number of hours.
What amazes me with all of these units is how they can differ in construction and still achieve the same output. I feel like this particular unit is a lot more refined just because it's a lot more basic on the inside.
Airsep Quietlife 5
This is the quiet version of the of the Elite, immediately you can see that the molecular sieves are a third smaller and the compressor is encased in a box.
Dust ! This is probably the dustiest unit I have come across yet.
I gave the whole of the outside a clean and then ran the unit to see if there were any issues. The unit was certainly quiet but it lacked flow of air and the oxygen sensor stayed illuminated, there were no alarms on the control board.
I removed the front to clear yet more dust and checked to see if there was anything obvious, there was not.
It was time to see if there was anything obvious around the compressor, failing that it would likely be the piston seals or a stuck reed valve.
The compressors in all of the units are mounted on a set of springs to reduce vibration, noise and ultimately extend the life of the unit. Between the compressor and the valve manifold is a flexible piece of tube, in this particular unit the tube is very short. It is not the tube that flexes but the connection itself which has lead to it leaking.
I'm not quite sure why the manufacturer chose to change this part of the design as originally it used compression fittings, the lines from the compressor to this fitting also chose a different path. I chose to replace it with a compression fitting and some flexible hose, it meant coiling the hose in a loop in order to make room.
The pipe sealed perfectly and allowed movement in the compressor. This simple fix solved the flow issue and also the oxygen warning light, these molecular sieves cannot function properly without pressure.
I have also had a few units where the start capacitor has failed, this is a simple diagnosis as the motor will not start. It is extremely unlikely that the motor windings would fail in this situation since the manufacturer implemented a thermal and a current overload. Here you can see below the new capacitor on the left and the failed on the right. I'm quite surprised that the failed capacitor didn't explode, the case has certainly bowed a lot.
More Random Pictures
Since I wrote this page I have repaired and sold many different types of oxygen concentrator. Most notably on the more common types I started to replace all start capacitors and compressor parts to ensure a reliable sale. Here are some more pictures.
Hello, if you have enjoyed reading this project, have taken an interest in another or want me to progress one further then please consider donating or even sponsoring a small amount every month, for more information on why you may like to help me out then follow the sponsor link to the left. Otherwise you can donate any amount with the link below, thank you!