[News] RECALL THREAD-- IMPORTANT PHILIPS DREAMSTATION & SYSTEM ONE USERS

[qwerty42]

Warning!!! Warning!! No one here has designed a cpap machine or really knows how all the parts interrelate. Who knows if they're even functioning effectively after the hacks?

True, I have not designed one, but I am confident that I could. Engineering is my day job, and the stuff that I work with (satellites and instruments for space) is quite a bit more complicated than these CPAP machines. Mechanically, they are very simple devices: a blower motor, a flow sensor (actually two closely-spaced pressure sensors being used to measure flow rate), an outflow pressure sensor, and a software-controlled closed feedback control loop. I'm sure the companies who make them want everyone to believe they are unmatchable voodoo magic, but they could easily be reverse-engineered and copied by an unscrupulous company wishing to do so.

To someone who is familiar with sensors, instrumentation, control loops, fluid and airflow physics, etc. and integrating them into small devices, it's pretty straightforward to look at the insides of these CPAP machines and recognize the fundamentals of how they work. It's a bit like learning a language which may seem foreign to those who don't speak it, but is more easily interpreted by someone who is fluent.

I don't know exactly how the therapy algorithms are coded in these machines (nobody does except those who have designed and written the code), but that isn't particularly relevant to the foam removal. The things that matter are:

• Are the sensors that drive the machine still getting the same values as before?
  
• Are they still operating correctly, accurately, and measuring the same specific things they did prior to the foam removal?
  
• Have those measurements been impacted in any significant way by the absence of the foam, and if it changes the flow characteristics into the blower, does the system have sufficient control range and authority to compensate, and operate as if nothing has changed?
  

If all of those things are still working as before, the control software doesn't care. It's blind to it. The mechanics and instrumentation are one "black box" which sends data to the software, which is another "black box," which processes it and drives the blower based on it. As long as both black boxes are getting the same inputs and outputs and responding the same way to them, the machine doesn't care if it looks like a pink elephant with purple walruses inside.

A few pages back (here and an important clarification here) I posted images of the DreamStation and pointed out the sensors it uses to provide control, along with part numbers (taken from my own machine's circuit board) and descriptions of each. From the fundamentals of how the DreamStation works and how it acquires its flow & pressure measurements, I concluded that removing the foam would be unlikely to have any impact on its functionality. The only sensor that could be affected at all by the foam removal is the one it uses to measure flow rate, and this would only be if the void from removing the foam creates flow changes (e.g. turbulence or recirculation) across the restriction grating where flow rate is measured. The sensor that measures and regulates mask pressure is *after* the blower, and is unaffected by the foam's presence or absence.

I also followed up on this after removing the foam in my own machine, and compared the measured data for the nominal leak rates of my mask pre- and post-foam removal, which were identical -- meaning the machine was still measuring the same value for the same flow rate as it had done before. So, we have confirmation both from the theory as well as some measured data that removing the foam (and making sure the removal holes are sealed) doesn't affect the DreamStation's ability to work as designed.

Disclaimer: if the foam removal is a hack job and leaves stuff hanging out into the flow path, or doesn't seal the foam removal cutouts, or cuts into any of the pressure measurement ports, or isn't reassembled correctly, etc. -- then yes, it could affect how the machine works. But cleanly and carefully removing the foam, and leaving it just as it was before except with no foam, appears to have no impact on its operation. I am not recommending anyone remove their foam; this isn't medical advice; you now have the information to make your own informed decision and understand the risks, if any.

[WakeUpTime]

Laboratory analysis found that as the foam degraded it tended to stick to nearby surfaces as well as itself.
This reduces the risk of respirable particles entering the breathing circuit.

Many of the early reports to Philips had black or dark grey particles ending up in users' masks.  If they made it to users' masks, they've made it to users' lungs.  Lab tests don't often mimic real life experiences.  There's also a difference between a short-term lab test and a multi-year practical real-life machine use.  In real life, conditions vary considerably.  Machines go on planes, and sometimes even go camping.  There are even machines that might be on their 2nd or 3rd user now.  I personally don't find it comforting or credible that the lab test suggests that particles likely will get stuck in the machine - and certainly not ALL of the particles, and not ALL of the time.

Brits aren't issuing a recall yet.

It's very political in the U.K. at the moment, with many multinational corporations leaving the U.K. because of Britain's departure from the E.U.  Philips just closed another building in the U.K that let go another 400 employees.  Britain is pleading with corporations to keep the U.K. factories, warehouses and offices.  There are still another 2,000+ Philips employees working in the U.K.  (Hence, internal pressure against a strong U.K. gov't response.)

Also, some countries don't have an "official recall"; yet Philips has told them they will replace all the defective machines in their country.

[hegel]

Well, the medical device regulatory agency has run tests and hasn't issued a recall. They write that health risks are minimal.

No one here has run tests or really has hands on expertise in the field.  One can always find generic reasons to discount test results or agencies but really, that's mere assertion and guesswork. Aliens are real and scientists who deny it are in on the cover up, etc.

Again, I wouldn't use a recalled machine. But people here love the sky is falling narrative. And people are being urged to take radical action on their machine and introduce foreign substances (cardboard, cotton, some home cooking material made from plastic, who knows what) in an effort to protect themselves. I hope it doesn't turn out badly for anyone.

A poster earlier took apart his years old machine and found the foam to be good as new. He was reassured. I think the report from the UK is possible good news. some people just hate good news.

I have a background is psychology. This has a whiff of self generating hysteria.

I wonder how many here have found black material in their masks over the years? Are there any threads asking about this? so far, no one has said they have.

[racprops]

qwerty42 Thanks for confirming my less scientific results.

I came to the same conclusions based on my years of working with so many devices.

Really like how you called out the part numbers of the sensors, I have lost a machine or two from what I was told was failing pressure/flow sensors.

Are they available on the open market? I figure they are just about the only parts likely to wear out or fail, other than the fan itself.

You are much more informed than I so question could a failing flow sensor cause a failure to imprint any data to the system's SD Card?? I have a 950 ASV that will no longer do reports otherwise if seemly operating correctly, it only has some 3 or 4 K run hours on it.

I asked a local service shop to check it as I was then planning on reselling it which I told them and they reported a total failure...and of course, Phillips no longer will supply a mainboard.

Rich

[WakeUpTime]

Perhaps you're right, and Philips decided to spend $600M - $1B for just the rare slight possibility that someone could ingest a speck of foam.  Philips was just looking out for Philips device end users' best interests - almost like family.

I have a background is psychology. This has a whiff of self generating hysteria.

Philips:

"Discontinue use of affected units"
(immediately)

[racprops]

hegel, Do you own and drive a late model car?? Are you aware of the airbag recall of the past couple of decades??

If your car had such a recall would you still own it and drive it??

The Dreamstation has only one component that is bad and thus subject to the recall. Like the Airbags, no one is throwing away cars because of them.

If Phillips sends out a foamless chamber I would gladly use it. From what I have seen of the new white flaking new form, I will NOT accept a chamber or a Dreamstation II, not because Phillips MAY have done me harm, but because I do not trust the new foam.

As they have not done so I feel the system is AOK with the foam removed by whatever means, by them or by myself, or by a service tech.

It will take a lot of research and record-keeping to avoid every manufacture that has made a defective product that MIGHT cause harm, heck it will be a very long list for just the ones that really have made products that harmed or killed people.

This is my real-world take on all of this.

Rich

[qwerty42]

qwerty42 Thanks for confirming my less scientific results.

...I have lost a machine or two from what I was told was failing pressure/flow sensors....Are they available on the open market?

...could a failing flow sensor cause a failure to imprint any data to the system's SD Card?? I have a 950 ASV that will no longer do reports otherwise if seemly operating correctly

Hi Rich, sure thing, and thank you for originally making me aware that the DreamStation had those sensors upstream of the blower!

The two differential pressure sensors on that board (for flow and mask pressure) look very similar to a couple of available parts from the same mfgs, but their part numbers appear to be OEM specific. Even though they're clearly derivatives of the parts we could purchase, they might not be the same in terms of output (e.g. different sensitivity; different voltage response vs. pressure; maybe a tighter calibration spec; that sort of thing). Unfortunately none of those potential differences are easy to determine on our own, without specialized equipment. So, you might get lucky and find the consumer equivalent works fine, but you could just as easily find the opposite.

I can't say for sure on your last question, but if a failing flow sensor *did* do that, it'd be because of a check they wrote into the software. The sensor failing on its own shouldn't have anything to do with SD card writing, but if the software has an error-checking scheme, it might refuse to do certain things like writing data to the card if it detects a fault like a bad sensor. That said, I think it's more likely that either the SD card itself is faulty (the first thing I'd check -- I'm guessing you already have too), or the SD card reader contacts are dirty, or the SD card circuit has somehow failed.

[racprops]

"I can't say for sure on your last question, but if a failing flow sensor *did* do that, it'd be because of a check they wrote into the software. The sensor failing on its own shouldn't have anything to do with SD card writing, but if the software has an error-checking scheme, it might refuse to do certain things like writing data to the card if it detects a fault like a bad sensor. That said, I think it's more likely that either the SD card itself is faulty (the first thing I'd check -- I'm guessing you already have too), or the SD card reader contacts are dirty, or the SD card circuit has somehow failed. "

Thanks for the thanks...I tested a different SD Card and both cards are still readable in my PC, and I cleaned the contacts with a spray contact cleaner, so it is either as you suggested a fault check (I believe a Tech say something like that..) or sabotage by the repair station....

I have a junk unit that seems to be working...I might see if I can unsolder the double sensor and replace the one on my downed 950 ASV unit.

I have an isolated solder iron station.

Rich

[Phaleronic]

The db difference between a machine with foam and a machine without foam is minimal.. Which begs the risk/benefit question regarding this company's ethics.

https://www.youtube.com/watch?v=iaYhrXAp1QQ&t=122s

[racprops]

I saw what I remember as a government requirement specs sheet, on it was a listing for noise = +/- some DB.

This requirement may be so low they HAD to add the foam to meet it, at least without more costly engineering, say sound baffles etc.

Rich