Archived SH Discussions [Q&A Thread from Sep-2011 thru May 2014]

[SuperSleeper]

Why are they so protective of the equipment? It seems so wrong to drive therapy by insurance compliance......

When Professor Barbara Phillips asked the question "Would you like to see the diagnosis/treatment of OSA alter in any way" she replied:

Get insurance rules and regulations out of the picture! Let clinicians manage patients. Stop requiring a sleep study in order to have CPAP paid for. Stop pretending that a CPAP titration (or an in-lab PSG, for that matter) is some kind of magical test.

[SuperSleeper]

How does the machine know the time? If I unplug it it does not have a clock to reset....
My timings seem a little off - like this morning I know I was awake and off the machine at 7, but the data ends at 720. If it was an hour out I would fix the time zone, or the DST.

I've wondered the same thing. Is the internal clock only resettable with proprietary software/

On the PRS1, there's a special data cable that plugs into the PCB underneath the SDcard slot.. There is software can update the firmware, and reset the clock.. Us mere mortals can't get our hands on it..

Philips simply doesn't care if PRS1 users get dodgy firmware.

I had a little look at the PCB connector behind the SD card cover.. It's only 4 data lines, two of which are obviously power... I don't recommend anyone try connecting anything to them.. But it's probably just a crappy USB or serial interface..

I've got to read the repair manual and see if there is any more clues on it..

[SuperSleeper]

For all you Mac users, either use the Updater or go to the SourceForge website and download the LATEST build of the Beta SleepyHead software.

Neat.. was a silly fencepost overrun error.. I have no idea why it doesn't die the same on the windows build.. Regardless, I'll upload a fix tomorrow for Windows too. :/

Just a heads up for PRS1 users, there is some slightly raw PRS1 Unintentional Leak detection stuff in that mac build..
(it should work ok, but it's still early code)

Hopefully I'll have a better version up once I get this algorithm nutted out properly.. It will be nice having a proper leaks graph for PRS1

[SuperSleeper]

Very interesting. PRS1 records no data for the first 10 seconds of a session.

I noticed that too. It's probably a good thing, as short sessions cause all sorts of nasty problems datawise..

If you can come up with a simple procedure of figuring this out, (for normal people I'll add the clock drift feature in preferences..

Hopefully I can do it in a way that doesn't require recalculations/reimports... just a simple positive or negative time offset.

One thing I noticed back when I was hacking with Python code.. There is a delay of around 10 or so seconds, but I never spent enough time with it trying to work out whether it was at the start or at the end of a session.. (Start makes a little more sense, but I could be wrong)

The absolute best way it know for sure is record a short session, watching the computers clock, and do some rapid breathing/block the mask/anything that will show up in the flow or leak waveforms... record the exact time you do this, and take note of the imported data..

Might need to do it a few times and get an average..

[SuperSleeper]

Zonk suggested that I post the following on this thread.
I'm not getting detailed data when I download the SD card with ResScan or Sleepy Head. The first week there was no detailed data, the second week had 3 days of data the fourth and fifth weeks had a few days of detailed data but all that showed up was flow rate and events. When I download I click on summary and all detailed data. Anyone have any ideas what my problem could be?

Hi the two obvious causes are SD card write protection, and SD card data corruption..

The S9 routinely deletes high detailed data over 7 days (including Flow Waveform, and a high resolution Mask Pressure waveform)
It routinely deletes all graph data older than 30 days,
and all event summary information (in STR.EDF) older than 370 or so days.

SleepyHead 0.9.1 beta and onwards now has an option (enabled by default) to maintain a backup of your ResMed SD data card during import, to counter the routine data deletion issues (and keep the most recent ~370 days STR.edf file).. This data is kept in it's original format (with the option of compressing the EDF files with a standard compression format to save space)

[SuperSleeper]

[quote-jedimark]There is a ton of cool stuff in future than can benefit from being able to sync with a PC's clock..

Maybe one day SleepyHead will have motion sensing IR webcam support.. or movement sensor belt support..

Although most of this stuff really just needs to sync start times..

But anyway, the drift option should be fairly easy to implement.

I just need to get unintentional leaks finished for PRS1, and Oximetry fixed first.. :-/[/quote]

[SuperSleeper]

Do you know about the zeo? This monitors brain activity during sleep. I would love to sync that up with SleepyHead too - except my mask uses my forehead, so I am not buying a zeo...yet.

SleepyHead 9.2 update already has Zeo CSV import..

Adding the drift feature will be important for this, but the Zeo is not all that accurate anyway..

Zeo CSV data is in 30second increments, so impossible to sync accurately, and it only seems to differentiate (roughly) between awake and deep sleep.. the interim stages (REM & Light sleep) are mostly bullocks..

[SuperSleeper]

I've downloaded and installed 0.8.8-3. I'll let you know if the problem recurs.

When I move the mouse pointer across any of the graphs on the Overview page, the screen blinks.

This happened just now when I launched SleepyHead. I closed the program, launched again, and the problem recurred. Closed again, launched for the third time, and problem as disappeared.

Just so you know, Mark, ever since I installed the first beta version a few days ago this problem has disappeared and not returned. (I didn't want to say anything earlier and jinx it, but after a few days of no problems I feel safe in declaring it gone!)

[SuperSleeper]

I've seen some graphs of sleep spindles and stuff from the zeo. I agree that it is not very accurate - that is why I am not buying it yet - but I think there is a way to get at data that is less agregated

I think both the Zeo and (accelerometer) movement based smartphone apps are worth investigating for differentiating between full sleep and wake cycles..

But there is no (cheap) way of determining REM as of yet..

A small accelerometer placed near the eyes would help stage REM properly with other frontal EEG signals. Would be awesome if Zeo tried this with their next generation headset.

Best idea I can come up without resorting to serious electronics hacking is the circuit board out of a cheap nintendo Wii nunchuck, and an arduino development board or similar. I think this is too big and uncomfortable to wear on the head.. but could have other uses (I was thinking movement belt for the chest)

[SuperSleeper]

Hi I recently acquired a Sandman Auto that is nearly new ,does anyone know a place where I could download the Silverlining Software 1.6?

Does your device have an SD card or a USB stick? or does it have one of those (evil) SmartCards?

Edit: I looked it up.. Does yours take that key shaped thing?