Dreamstation and 1 hour sessions?

I've found that occasionally, the Dreamstation will split up the night into 1 hour sessions. I'm seeing this in both EncorePro and SleepyHead, so I know that it's the machine and not the software that's viewing it. Anyone know why this happens? See below for the Sleepyhead view:

It seems that some DreamStation's do not accurately record the sessions sometimes. I have seen 3 machines on here with that problem. Here is a link where Pugsy and Linuxman discussed his machine on Jan 19. viewtopic.php?f=1&t=115002&st=0&sk=t&sd=a&start=30

Do you have auto/off enabled? Although the times given don't sem to be long enough for the machine to shut off and turn on again, I wonder if the machine is somehow sensing a momentary cessation of breathing (or "thinks" it is), ans splitting the session.

I'm not sure if this is right, but if auto/off is enabled, try disabling it (so we can rule this in or out as the problem). Note that on DreamStation machines, you can still leave auto/on enabled, even if you disable auto/off.

Not just the DreamStation. I know of one PR S1 560 that did it...it was the first to be having the multiple session thing.

OkyDoky wrote:It seems that some DreamStation's do not accurately record the sessions sometimes

jds2001 wrote:Anyone know why this happens?

Never could figure out the why. I suspect a bug in the machine's software but couldn't prove it.
I call it the "multiple session bug" but that's just something I came up with for lack of any known cause.
Your number of sessions is very small compared to some of the others I have seen. I saw some sessions lasting just 5 or 6 minutes and 30 to 50 sessions a night.

If it were me I would be alerting whomever you got the machine from. It's not normal and that always makes me worry about a machine when it doesn't act as expected or normal.

D.H. wrote:Do you have auto/off enabled? Although the times given don't sem to be long enough for the machine to shut off and turn on again, I wonder if the machine is somehow sensing a momentary cessation of breathing (or "thinks" it is), ans splitting the session.

I'm not sure if this is right, but if auto/off is enabled, try disabling it (so we can rule this in or out as the problem). Note that on DreamStation machines, you can still leave auto/on enabled, even if you disable auto/off.

Been there...done that and it's NOT related to any auto off or auto on feature.
Note the times of the sessions...one session ends and another starts immediately. Respironics machines don't turn off for 60 seconds after it thinks it doesn't have a human attached.
But we already experimented with auto off way back when I saw the first PR S1 with the multiple session problem. It made zero difference.

Pugsy wrote:

D.H. wrote:Do you have auto/off enabled? Although the times given don't sem to be long enough for the machine to shut off and turn on again, I wonder if the machine is somehow sensing a momentary cessation of breathing (or "thinks" it is), ans splitting the session.

I'm not sure if this is right, but if auto/off is enabled, try disabling it (so we can rule this in or out as the problem). Note that on DreamStation machines, you can still leave auto/on enabled, even if you disable auto/off.

Been there...done that and it's NOT related to any auto off or auto on feature.
Note the times of the sessions...one session ends and another starts immediately. Respironics machines don't turn off for 60 seconds after it thinks it doesn't have a human attached.
But we already experimented with auto off way back when I saw the first PR S1 with the multiple session problem. It made zero difference.

Thanks Pugsy, but I still think that the original poster needs to rule this out (or in).

BTW, I have a Dream Station and have never seen this happen.

D.H. wrote:BTW, I have a Dream Station and have never seen this happen.

And your point is?

I have owned 7 different Respironics machines and never seen this happen.
But I have seen reports where it is obviously happening. Just because you or I haven't seen something doesn't mean a whole lot.

D.H. wrote: I still think that the original poster needs to rule this out (or in).

Won't hurt and won't cost anything but it's a waste of time.
Have you ever taken the mask off with auto off set to on and timed how long it takes it to shut itself off?
If you haven't...try it. You might learn something.

Pugsy wrote:

. wrote:BTW, I have a Dream Station and have never seen this happen.

And your point is?

I have owned 7 different Respironics machines and never seen this happen.
But I have seen reports where it is obviously happening. Just because you or I haven't seen something doesn't mean a whole lot.

. wrote: I still think that the original poster needs to rule this out (or in).

If you haven't...try it. You might learn something.

ROFL.
not bloody likely.

D.H. wrote: Thanks Pugsy, but I still think that the original poster needs to rule this out (or in).

BTW, I have a Dream Station and have never seen this happen.

No auto on (or off) for me. Manual button pushes all the time. And this doesn't happen every night, but has for the last 2 now. And I think it's really interesting that the sessions are EXACTLY 1 hour (and a few seconds), and not some weird interval more or less than that.

As mentioned, I've seen the multi-session behavior with my Dreamstation. Interestingly, my machine stopped doing this about a week ago, and I'm now getting single sessions per night. I'm looking back through my notes to see if I can find any correlation to anything. The only thing I can really see that's been changing is my use of heated humidity, but it's not a clean correlation. As I've achieved better control over mouth breathing (with or without a FFM) I've been able to dial back the humidity, and I now run either an empty chamber (FFM or regular nasal mask), or passover humidity with pillows. Heated tube is running at the lowest level in either case, and I *think* that predates the time where it started doing single sessions. I unfortunately did not record my heated tube settings.

I guess it's possible that there is some sort of power stability issue when the humidity chamber heater turns on and off to keep target temperature. Just conjecture at this point, but I think possible. Also, not only does correlation "not causality make" but with so little data this could well just be random chance.

I'm going to do some more controlled experimentation over the next few days to see if I can recreate, and at least establish a stronger correlation with something.

As we've now ruled out auto/off, I suggest calling the DME/or and the manufacturer.

y'all should compare firmware versions, maybe there's a fixed bug?

My person opinion about a machine that doesn't behave as expected...like if it had been an issue with the auto off feature...is that for the money we pay for these things we shouldn't have to accept something that isn't behaving properly.
Even if we had zeroed down on it being the auto off feature...it should work as designed and we shouldn't have to put up with less than original design.

Now it might be related to software version but remember I saw a PR S1 560 that had the problem too.

Could there be a correlation between some sort of setting variations...possibly but again to me that isn't as it was designed and we shouldn't have to "live with" less than optimal design. It just bugs me too much and that means I don't trust it and eventually that leads me to maybe question the data. In the back of my mind there's always the question "what if this little bug isn't the only little bug and any other bugs could be impacting the recorded results or something else".

But that's just me and you guys are free to do as you wish about accepting "less than original design".
Now if we had spotted massive leaks causing the machine to think no one is attached and auto off wanted to turn it off...that would be a plausible explanation but I don't see anything near large leak on the report above.

Pugsy wrote: But that's just me and you guys are free to do as you wish about accepting "less than original design".
Now if we had spotted massive leaks causing the machine to think no one is attached and auto off wanted to turn it off...that would be a plausible explanation but I don't see anything near large leak on the report above.

The problem with this is that there is no mechanism by which to lodge a complaint that would be at all likely to resolve the issue, at least for me personally, but for most as well. If I go to my DME and say "this machine has a funny, intermittent, behavior with regard to session instances", they will look at me like I'm crazy, I assure you. In my case, they only look at rough summary data. Now that I've met compliance, they're out of even that game and told me they'll not be calling back (which is fine - they're useless). If my doc was more involved *maybe* he might be in a position to ask something of Philips, but I doubt it. To file such a claim as an "outsider" you're going to have to show something that is truly anomalous, and likely, detrimental to therapy. We have no such evidence. Now, a very inquisitive person with access to the Resmed software, in a distributor relationship, and my SD card...yeah, maybe they would have some ability to file a bug with Philips. For the reasons stated below, I think even that would be very unlikely to result in any action on this.

Pugsy wrote:My person opinion about a machine that doesn't behave as expected...like if it had been an issue with the auto off feature...is that for the money we pay for these things we shouldn't have to accept something that isn't behaving properly.
Even if we had zeroed down on it being the auto off feature...it should work as designed and we shouldn't have to put up with less than original design.

Some more context on software and hardware development I think will help here. I work professionally in the very type of embedded hardware and software that's used in devices like this (but with a non-medical focus). All software is buggy. Very, very, buggy. Systems that are of the same rough technical complexity as a cpap machine (that would be fairly low complexity, by the way) would typically have tens to hundreds of bugs found during the test cycle in a given release. Those bugs would be triaged and categorized by impact and priority to fix. Only the ones that would directly impact product function or safety would be even considered for fixes. Among those that will be fixed many will only get workarounds that address the issue, even without root cause understood. Some moderate number of the most severe issues will be root caused and successfully corrected. This is just the way it is. Software is extremely complicated and very error prone,and human resources are limited. If everything was fixed, CPAP machines would cost a lot more than they do. As an industry at least the top two companies seem to be doing just fine with regard to core product function. The fact that we see this apparent bug doesn't mean anything at all with regard to the overall product quality. There are many, many similar bugs in all of these products - both known and unknown, visible and invisible. The operating system on your computer has thousands or tens of thousands of bugs that go unresolved - mostly harmless (apologies to Douglas Adams ).