oxygen concentrators

[Doug Fattic]

I have a feeling that with the need for oxygen assistance for those that contracted the coronavirus there will be a run on oxygen concentrators. Someone checked with M&M Medical in Beaverdale, PA that repairs and sells them and they were all out. The news talks about ventilators which does the work of a patient's lungs but the next step down is a concentrator that refines air into more pure oxygen. Has anyone tried to get one recently?

[Guillaume]

Just before taking the plunge, a respironics eveflo Q opi should work?

https://www.usa.philips.com/healthca...specifications

Oxygen Concentration*(at 5 LPM) 93% +/-3 %
Outlet Pressure 5.5 PSI
Liter Flow 0.5-5 l/min

[Doug Fattic]

Just before taking the plunge, a respironics eveflo Q opi should work?

https://www.usa.philips.com/healthca...specifications

Oxygen Concentration*(at 5 LPM) 93% +/-3 %
Outlet Pressure 5.5 PSI
Liter Flow 0.5-5 l/min

I'm not familiar with that particular model but the specs look right. The ultra quiet feature advertised is nice. I get used to the background noise on my concentrators but obviously less is better.

You will have to convert the barb outlet to a welding hose "B" fitting. The oxygen concentrator industry has plastic ones. The Welding trades make brass ones. Let me know if you want more information about how to do this. I get the clear plastic hose that attaches to the barb on the concentrator on one end and to the brass (or plastic) convertor on the other from Lowe's.

[Arttu]

Hi everyone, first post.

After years and years of wanting to build frames, I'm finally plunging for an oxy-propane setup. I've found a DeVilbiss 525ks locally for a fairly good price. It's not cheap (470 euros), but these things are fairly rare around here and and that seems to be about the average price the go for on ebay in europe. The seller claimed it didn't have a hour meter but after some digging I found out that on some models, the meter is hidden behind the filter cover on the back of the machine. The seller checked and it has 425 hours on it. I'm not an expert but that doesn't like too much considering that the machines are rated for tens of thousands of hours. 425 hours isn't much, right?


This might be slightly off-topic, but I'm looking at getting the Messer Minitherm torch to go with the machine. Some local builders use this setup, but I'm wondering if anyone here has any input on torch selection for a beginner?

[Mark Kelly]

I thought that was an interesting question so I took a look at the literature.

Apart from mechanical failures (compressor, valves) the life limit is the absorbent bed becoming poisoned because there are impurities in the infeed stream which are not removed by the pressure cycle so they stay in the bed. The cure for this is to replace the bed.

That being said: this paper says that with basic maintenance in a "low resource enviroment" the average age of the functioning concentrator was around 6500 hours and the oldest was more than twice that.

[Doug Fattic]

After years and years of wanting to build frames, I'm finally plunging for an oxy-propane setup. I've found a DeVilbiss 525ks locally for a fairly good price. It's not cheap (470 euros), but these things are fairly rare around here and and that seems to be about the average price the go for on ebay in europe. The seller claimed it didn't have a hour meter but after some digging I found out that on some models, the meter is hidden behind the filter cover on the back of the machine. The seller checked and it has 425 hours on it. I'm not an expert but that doesn't like too much considering that the machines are rated for tens of thousands of hours. 425 hours isn't much, right?

That's right, the low number of hours makes it almost like being brand new. They are designed to run constantly for years. When I got my Devilbiss model 515 (an older model than yours) it already had 20 some thousand hours of use and now I've been using it for years since.

The company that refurbished mine (M&M Medical) said the Devilbiss models 515/525 have fewer repair problems than other brands. I've tried out several other makers and I get more output on my Devilbiss (even though they are all rated at 5 lam). So it is a good choice. Covid 19 has made finding used concentrators more difficult. Damaged lungs have increased demand.

[Guillaume]

I scored a Devilbiss 515 for 300$ canadian, 1282 hours!! I knew patience would pay off someday.
Have been stored a few years but it seems to works as it should.

[Ahmdo Molah]

Doug (and those familiar with Devilbiss units),

*** My attempts to send this message to your 'current email'(as of Nov.'21) failed(along with VS PM too). I'm attempting email via VS and posting here too. ***

Someone near my location is actually selling a Devilbiss 515KZ for about $250 (Oh My Got!). Sorry for the limited photo attached which is all that the seller posted online.

Devilbiss 515KZ.jpg

It does not have the OSD (O2 sensing) feature, but I have an O2 analyzer that can also check flow(lpm)and pressure(psi).

If I recall correctly you had a similar model(US version) until you replaced it recently. Other than what I can check with the O2 analyzer, and checking the hours of use by opening the filter cover behind the handle, I'm hoping that you might be able to give some pointers on what to look out for to determine if it is even worth getting. I will be visiting a total stranger's home so I don't think I can impose and spend too much time there taking off the panels and looking inside like I'd like too.

Perhaps you noticed some distinct sounds or other symptoms before your unit needed to be retired, and how it was different from when the unit was operating normally?

BTW, this one seems like a unit grandma once used (the seller is also selling one of those fancy 4-wheeled electric mobility scooters). : )

[Doug Fattic]

Doug (and those familiar with Devilbiss units),

*** My attempts to send this message to your 'current email'(as of Nov.'21) failed(along with VS PM too). I'm attempting email via VS and posting here too. ***

Someone near my location is actually selling a Devilbiss 515KZ for about $250 (Oh My Got!). Sorry for the limited photo attached which is all that the seller posted online.

Devilbiss 515KZ.jpg

It does not have the OSD (O2 sensing) feature, but I have an O2 analyzer that can also check flow(lpm)and pressure(psi).

If I recall correctly you had a similar model(US version) until you replaced it recently. Other than what I can check with the O2 analyzer, and checking the hours of use by opening the filter cover behind the handle, I'm hoping that you might be able to give some pointers on what to look out for to determine if it is even worth getting. I will be visiting a total stranger's home so I don't think I can impose and spend too much time there taking off the panels and looking inside like I'd like too.

Perhaps you noticed some distinct sounds or other symptoms before your unit needed to be retired, and how it was different from when the unit was operating normally?

BTW, this one seems like a unit grandma once used (the seller is also selling one of those fancy 4-wheeled electric mobility scooters). : )

My internet provider decided to quit offering email service so they shut mine off October 31. While I'm setting up a new gmail account, you can email me at my very old juno.com account. Just type in my name before the AT in all lower case letters without any spaces or dots or anything else and you'll reach me.

I haven't actually retired my Devilbiss 515. It has been working fine for 10 years. The filter in the back has to be changed once in awhile. It is behind the back door that when opened up also shows the number of hours used. Mine is over 17,000 hours. I just added a Devilbiss 525. Unfortunately I don't know much about their inner working. They run fairly quiet when working properly but still make a noticeable background hum. It is enough that those using them for breathing while sleeping figure out a way to place them in another room and run a long extension tube.

I know they can be tested to see what percentage of oxygen they can produce. They are supposed to put out 92% when working properly. However one of my students bought one that had an output of 80% (according to the seller) and we couldn't tell the difference in flame quality between his and mine. I'm not sure how the seller tested the oxygen output. Does your oxygen analyzer check oxygen percentage? His 525 runs noisier than mine but still works fine.

As mentioned before, the Devilbiss 5 lpm models put out more pressure than other brands. $250 is a good value. It is convenient that the output fitting on the 515 and 525 are the same threads as an American size "B" fitting so it is possible to screw in a welding hose directly into the concentrator.

[Ahmdo Molah]

... I haven't actually retired my Devilbiss 515. It has been working fine for 10 years. ... Mine is over 17,000 hours.

... Does your oxygen analyzer check oxygen percentage? ... .

That's good to hear that your older Devilbiss 515 is still working fine, as the one that I'm considering is likely as old. As the used units I'll be checking out may have never been refurbished properly like yours, I feel at ease that I can check their condition with my ultrasonic O2 analyzer;

Ultrasonic O2 Analyzer.jpg

The analyzer has 2 nipples on top, the left nipple analyzes O2 percentage and flow(LPM) and the right nipple analyzes pressure(PSI). It's one of the relatively affordable units, and I felt it a necessity to prevent purchasing a non-performing unit. And for units that are not equipped with O2 sensors, I figure the analyzer would be handy for checking O2 output & pressure from time to time.

[bulgie]

Is a malfunctioning Invacare Platinum 5 worth trying to repair/rebuild?

I got it cheap a couple years ago, never used it other than turning it on once I got it home. The O2 sensor light came on green, but other than that I didn't test it. I've been using up my bottled O2 in the meantime, and not doing very much brazing, so my two bottles have been hanging in there.

But now I'm interested in setting up the Invacare for use. When I turned it on, it seemed to run well at first — O2 sensor lit green, and flow right at 5 lpm. But after maybe ten minutes, a piercing alarm went off, and a red LED lit up solid. Power-cycled it, and now there's no alarm or red LED, and the flow is still at 5 lpm, but the O2 sensor light does not come on. I have not hooked it up to a torch, but I don't expect it's making O2.

Hours meter says 21,000 and change, but of course it could also be 121k or 221k since there's no sixth digit for hours.

Has anyone here gotten theirs repaired, and if so how expensive was it? Is such a high "mileage" unit even worth trying to fix?

I'm in Seattle in case that matters. Shipping it would be exhorbitant, so unless there's a local repair option, I'm probably screwed. That is unless diagnosing, ordering a part and installing it myself is an option. But I wouldn't know where to start.

Thanks for any advice
Mark B

[Scheisserad61]

Few years back I bought a Sequal 10 LPM concentrator from an estate sale. It had low hours, was super clean, and seemed to work fine for about half a frame build, but then started giving me issues with purity and very low output. I think the seive bed might have gotten contaminated and needed new Zeolite, and/or the control panel electronics have gone South. The compressor is solid and no, it isn't the filters or any of the easy fixes.

I couldn't find anyone nearby that worked on concentrators, and much to my chagrin -I found that the original Sequal company had gone belly-up and basically I'm S-O-L on this P-O-S. Oh well, I've lost lots more $ than this $200 clams to be overly bothered by the trial and error of this experiment. When the thing was working it was pretty cool to think that I wouldn't need oxygen bottles anymore. I went back to the bottle (for now) until someday down the road when I can get a more mainstream (probably much more easily fixable) Devilbiss for reasonable money.

I can easily deal with a Regulator rebuild problem (especially on nice, simple Victor250's), but my concentrator was more of a PITA than I want to deal with.
If there is any advice I can give - it would be to hang on to at least one oxygen regulator just in case you ever need a backup.

Regards,
Michael Fabian

[bulgie]

If there is any advice I can give - it would be to hang on to at least one oxygen regulator just in case you ever need a backup.

Yes, wise words. I'm now shopping for my cheapest local hydro-testing option, so I can keep and refill my two O2 bottles that are both past their pull-date. I want the redundancy of two complete setups. Maybe I could have gotten along with just one O2 bottle if I'd gotten the concentrator working, but I'd never trust it enough to ditch the bottles completely.

Sorry to hear that yours has bitten it. No one around you that even wants it for spare parts?

Mark B

[bulgie]

OK, I've watched some Youtubes where they tear down an OC and talk about the parts and what they do. And I found this doc with schematics and part numbers, so maybe I can just start replacing the most-likely things to go wrong — if they'll sell me the parts. They might well only sell to "certified" repair shops though, we'll see.

Since the flow is still good, I assume that means the compressor is fine and the input filter isn't clogged. I hear the valves switching normally as they go from one sieve bed to the other, so I think the circuit board and valves are probably OK. I'm guessing the most likely thing to need replacing is the 'sieve bed assembly', those two cannisters with the zeolite crystals.

Just that assembly is probably more expensive than what I paid for the whole OC though, and I'd hate to invest that much into it only to find that it's still not working... I'll research it some more and think on it. Any advice or experience (even guesses) would be most welcome.

[Doug Fattic]

Mark, is there a chance that you have turned off your oxygen knob (like you would if you are using a tank) because when you do, the machine will start screaming to let you know you are about to die because you aren't getting any oxygen any longer. You have to leave the oxygen flowing all the time without ever turning it off or it gets real noisy.

You might call M&M Medical in Beaverdale PA just to see what they charge for a repair. I don't think it is much cheaper than buying a refurbished unit from them for $300. Sometimes you have to ask for that price if they quote a higher one. My refurbished unit from them has been working fine for over 10 years. It has 21,000 hours on it now.

You should try hooking it up to your torch just to see if it will work. I had a unit that only put out less than 80% oxygen and it operated a torch just fine.

Jihoon from South Korea rebuilt his unit if I remember correctly. He is up a few posts from yours. You might email him to get his advice. He was pretty clever about making his work again.

[bulgie]

Mark, is there a chance that you have turned off your oxygen knob

No chance, I had nothing attached to the OC, and the flow was solidly on 5 lpm when the alarm went off.

You should try hooking it up to your torch just to see if it will work. I had a unit that only put out less than 80% oxygen and it operated a torch just fine.

Yes I plan on doing that but haven't yet. My O2 purity green light went out, but it could be that the O2 amount is still adequate, no repair needed.

Thanks for all your help on this!

[Ahmdo Molah]

bulgie,

I recently rebuilt that 'grandma's old DeVilbiss' unit I mentioned above, with the help of an Invacare Platinum Series service manual.
https://www.manualslib.com/manual/82...latinum-5.html

*Note: My DeVilbiss 515 and the Platinum 5 both use the same compressor - a Thomas 2650. The Invacare Platinum series service manual is so detailed (with clear 3D drawings and even torque settings) it filled in all that the DeVilss service manual left out, at least with regard to a compressor head rebuild.

When I turned it on, it seemed to run well at first — O2 sensor lit green, and flow right at 5 lpm. But after maybe ten minutes, a piercing alarm went off, and a red LED lit up solid. Power-cycled it, and now there's no alarm or red LED, and the flow is still at 5 lpm, but the O2 sensor light does not come on. I have not hooked it up to a torch, but I don't expect it's making O2.

The flow rate inversely affects the O2 concentration (i.e. higher flow = lower concentration, lower flow = higher concentration). Your unit may be producing O2, but with the flow maxed at 5lpm the concentration may not reach the unit's set medical grade concentration.

I noticed with my DeVilbiss, if I set my flow low, say 1~2lpm, at initial power ON, the concentration goes from 21% to 90% after about 15 minutes, but if I set it to 5lpm, it takes longer. Also, if I crank the knob more so that the flow is maxed beyond 5lpm, the concentration drops to as low as 40%. It makes sense that the DeVilbiss service manual indicates that concentration tests should be performed after 20 minutes from initial power ON, to allow for normal rise in O2 concentration.

In your specific case, I've quoted the Invacare Platinum service manual's 'Operating Sequence' (p. 14-15) below for your convenience;

"Operating Sequence
Once the power switch has been turned on, the SensO 2 circuit will wait five minutes for the concentrator to begin producing clinically acceptable oxygen and the oxygen sensor to stabilize. The GREEN light will illuminate (indicating normal system operation) while the oxygen sensor is warming up.

After five minutes, if the oxygen purity exceeds 85% ± 2%, the GREEN light will continue to illuminate.

If the oxygen level is not above 85% ± 2% after the first five minutes, the system will continue to monitor the O 2 and wait for a maximum of 30 minutes from start-up to reach 85% ± 2% before activating an alarm. Environmental factors such as low voltage, high altitude, or age of the machine will affect the time required to reach 85% ± 2%.

If the oxygen level is not above 85% ± 2% within the first 30 minutes, the oxygen concentration alarm sequence will activate and the unit will shut down.

When oxygen concentration is above 85% ± 2%, the sensor measures oxygen purity every 10 minutes. If a reading falls below 85% ± 2%, a YELLOW light will illuminate. If the oxygen purity falls below 73% ± 3% the RED light/Alarm/Shut-Down mode will activate."

Since the flow is still good, I assume that means the compressor is fine and the input filter isn't clogged. I hear the valves switching normally as they go from one sieve bed to the other, so I think the circuit board and valves are probably OK. I'm guessing the most likely thing to need replacing is the 'sieve bed assembly', those two cannisters with the zeolite crystals.

The flow does not necessarily mean that the compressor is 'fine' as I discovered when I rebuilt my compressor's head. Refer to the service manual's section on "Rebuilding the Thomas Model 2650 and 2660 Compressor - Platinum 5 and 10 Models" (p.37~44).

On p.39, the drawing shows the 'Piston Cup'. It is essentially a Teflon plunger, that acts like a piston ring in an engine. These piston cups wear down very thin on either side where the WOB-L® pistons rub most against the 'Cylinder Sleeve'.

Wear on Piston Cups and Piston Sleeves
* Note the extremely thinned edges of the Piston Cups. As these edges wear thinner, the compression/air tightness becomes weaker resulting in lower pressure at the sieve beds and slower O2 concentration at the output.

Wear on Piston Cups and Piston Sleeves.jpg


Old & New Piston Cups
* Note the thinner than hair edges of the old cups(top) vs the 1mm-ish thick edges of the new cups(bottom).

Old & New Piston Cups.jpg

Even with the severe wear on the piston cups, my unit managed to reach 94% O2 concentration eventually, but decided to rebuild the head when I noticed the marks on the inner walls of the cylinder sleeves, visible from the outside through the fan blades. It would have been only a matter of time before damage to the pistons would take place, and that would likely be a more complicated repair or possibly require a compressor replacement.

MACK Industrial sells a 'Thomas 2650 and 2660 Compressor Rebuild Kit' for about $40.
https://store.mackusa.com/products/t...or-rebuild-kit

*Note: There is absolutely no need to remove the fan blades(counter to what the service manual may seem to suggest) when rebuilding the compressor head. Attempts to force the fan blades off often leads to to their breakage according to a compressor service technician that has done more than 1,000 rebuilds.

The valves switching are indeed a good sign they are working and that the PCB is good for that purpose. In my case, I opened the valve housing and replaced the gaskets and O-rings as the gasket had hardened due to the heat and pressure, much like a cylinder head gasket on an engine. Getting these parts took some time and effort but was definitely worth it in my case, as it was clearly a potential leak point.

4-Way Valve Body and Gasket.jpg

I'd bet that your 'sieve bed assembly' is fine as long as there are no leaks on their connections. Moisture is apparently not good for the zeolite (causes clumping/clogging?), so keeping them dry would be good.

Inspired by Invacare's Heat Exchanger, I used some 1/2" aluminum HVAC pipe and compression fittings to build one for my DeVilbiss. I made some 'J-shaped' hooks to hang the heat exchanger from the upper housing. I could not find enough room for a dedicated fan above the heat exchanger, but I did find juuust enough room between the compressor and the housing(with all the crumbling fudge-like noise damping foam removed) to fit a moisture separator with a DIY 'AI' shaped aluminum bracket assembly.

HotRod DeVi.jpg

So far I've not noticed any moisture released through the auto drain on the moisture separator, likely due to short sessions and low humidity air conditioned environment, but I'm comfortable knowing I can be at ease for longer sessions even in humid environments.

No chance, I had nothing attached to the OC, and the flow was solidly on 5 lpm when the alarm went off.

... My O2 purity green light went out, but it could be that the O2 amount is still adequate, no repair needed.

As the 'Operating Sequence'(in the service manual) indicates, the Platinum 5's O2 sensor circuit " will wait five minutes for the concentrator to begin producing clinically acceptable oxygen and the oxygen sensor to stabilize", so during this period it seems that the light will not be operating.

With the flow set at 5lpm, the O2 concentration may not be at 73% (even though O2 is being supplied), hence the (RED light) alarm. I'd suggest setting the flow to say 1~2lpm and wait 15~30 minutes to get the O2 concentration (within the internal storage tank) to a high O2 concentration and see how the lights and alarm react.

Cheers

[bulgie]

Jihoon Jo, thanks for such a detailed and information-packed post! Much to digest there.

About waiting for O2 concentration to improve up later — mine went the other way, went from O2 sensor being green for a long time, to later setting off its alarm. I forget now how long I ran it before it freaked out. During this time I wasn't using it, just letting it run at about 5 lpm with no torch or anything else attached.

Since my last post, I have connected it to a torch, and it makes a pretty good flame. Not as big as I tend to like for, say, a fork crown, but I like real big flames so many people would say the flame I got was plenty big.

Since it seems to be working adequately, I don't think I'll do any more along the lines of rebuilding, not for now, I'll just use it.

I have the alarm buzzer disconnected, so that won't be troubling me anymore.

Regards,
Mark

[Ahmdo Molah]

Mark,

Just a few points I forgot to mention in my previous post;

1. Invacare Platinum Series have ceramic zirconia O2 sensors which have a finite lifespan of several years (unlike ultrasonic O2 sensors which last indefinitely). The ceramic zirconia O2 sensors' finite lifespan maybe due to the fact that, "electric current flowing through a metal film resistor heats the disk in excess of 300° C."(572° F Whoa!).
If the O2 sensor is kaput, it might explain the alarms despite otherwise good system condition and O2 supply.

2. The service manual's Section 4 - TROUBLESHOOTING (p.17~24) might help you determine the issues and current state of your unit.

3. It seems you are getting a certain amount of O2, great! If you'd like to check the maximum O2 concentration your unit can put out, you might try setting the flow low(like 1lpm) for 15~20 minutes then light your torch with whatever O2 flow you prefer. If you are getting O2, then your O2 will likely be beefier as the concentration will be higher.*
* If the flow is kept high(maxed out), the O2 concentration will slowly decrease like a percent every several seconds and bottom out according to the flow setting. The DeVilbiss has an internal tank about the size of a mellon and the O2 concentration takes a while(a few minutes) to bottom out. The Invacare Platinum 5 seems to have an O2 reservoir("Product Tank") perhaps about the size of a Foster's beer can(?), so it should produce similar gradual reduction in concentration.

4. The rate of drop in O2 concentration (due to increase in flow) is quicker than rise of concentration by lowering flow setting.

Cheers

[waterlaz]

I have given up on my attempts to fix the DeVilbiss 525. Maybe some day I will try to replace the sieves.
All the years of using a semi-working concentrator made so frustrated that I bought a brand new Nuvo 8, which as the name suggests produces 8 lpm of oxygen.
Oh, the wonders! It has made fillet brazing enjoyable again. On the other hand, I have learned that I don't really need 8 liters of oxygen.
The flame is so much hotter with the new unit (probably due to better oxygen purity) that for most work 3lpm is enough and my torch definitely can't handle more than 5lpm.