S9 flow skew bug... Anyone else seeing this?

General Discussion on any topic relating to CPAP and/or Sleep Apnea.
User avatar
Nord
Posts: 565
Joined: Sun Mar 07, 2010 4:30 pm
Location: GTA Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by Nord » Thu Apr 15, 2010 7:22 am

Thanks alterego...

I have the same software/ firmware as you (and everybody else that checked).
I wonder if there could be any hardware differences... memory, cache, writers... guess we won't know that one.
Glad fiberfan is collecting the data now as she has already had significant skewing.

BTW: I do try to avoid the nap over noontime knowing that it might be an issue... but also a test that I am a more than willing participant

Thanks again for you help alterego...

Nord

User avatar
alterego61
Posts: 83
Joined: Wed Aug 26, 2009 10:11 pm
Location: Ontario, Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by alterego61 » Thu Apr 15, 2010 7:29 am

I received DreamDiver's files too, so will examine those this evening. I have to go earn a living now!

_________________
Mask: Mirage Liberty™ Full Face CPAP Mask with Nasal Pillows With Headgear
Humidifier: S9™ Series H5i™ Heated Humidifier with Climate Control
Additional Comments: APAP 5-7, EPR 1, ClimateLine Hose, ClimateControl Auto 27C/80F, ResScan 3.10 / Win7 64, 16GB SanDisk Ultra Class 4 (15MB/s) SD Card

User avatar
Nord
Posts: 565
Joined: Sun Mar 07, 2010 4:30 pm
Location: GTA Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by Nord » Thu Apr 15, 2010 8:51 am

Just as an aside... I just noticed today...

The Clock on my S9 is now running some 21 seconds slow from my computer clock after synching it to the computer less than 2 weeks ago.

Nord

User avatar
dave21
Posts: 720
Joined: Tue Mar 09, 2010 1:05 am
Location: United Kingdom

Re: S9 flow skew bug... Anyone else seeing this?

Post by dave21 » Thu Apr 15, 2010 11:09 am

I wouldn't worry about the clock running slow, most clocks do run slow, it tends to be down to the battery and the quartz. Computer clocks frequently run slow but Windows on most machines now update over the internet through an NTP time server to keep them in-sync. The S9 and most if not all CPAP's don't have this facility. I did post about this on another thread, by now with all the radio airway clocks you see in our cars now you'd think they could put something into CPAP machines, but I guess the more technology that gets crammed in the more price it goes up.

Thanks
Dave

_________________
Machine
Additional Comments: Running AirSense 10 AutoSet CPAP, previously S9 AutoSet and S8 AutoSet Spirit
Image

User avatar
alterego61
Posts: 83
Joined: Wed Aug 26, 2009 10:11 pm
Location: Ontario, Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by alterego61 » Thu Apr 15, 2010 3:07 pm

OK, I started looking at DreamDiver's files. I'll post results this evening as I produce them. I will leave interpretation and conclusions until I've had time to think things through a bit. I first looked at the files dated 17 March and 20 March.

Summary of Results

- There were significant time lags between the timestamps in the header data of the files and the timestamps embedded in the filenames, ranging from 27 seconds to 122 seconds. Compare this to my own (unskewed) files, which show a consistent 1 second lag for all files between the timestamps in the headers and in the filenames.
- Over the 3 day interval the time lags in two of the files (PLD and SAD) increased, but the other two (BRP and EVE) actually decreased
- The differential lags, or skews, between the timestamps, appear in ResScan graphs exactly as they do in the files, for example on the detailed graph for the night of 19/20 March the flow data graph line begins at exactly 04:40:02 and the pressure chart line begins at exactly 04:41:00


Data from files Named 17 March 21:24:02

Start Time from Summary Usage Bar: 21:24

Detailed Data File Header Timestamps

PLD (Pressure/Leak): 21.24.29 (27 second lag from filename timestamp)
SAD (Oximetry): 21.24.55 (53 second lag from filename timestamp)
BRP (Flow): 21.25.21 (79 second lag from filename timestamp)
EVE (Events): 21.25.42 (100 second lag from filename timestamp)


Data from files Named 20 March 04:39:26

Start Time from Summary Usage Bar: 04:39

Detailed Data File Header Timestamps

BRP (Flow): 04.40.02 (36 second lag from filename timestamp)
EVE (Events): 04.40.30 (64 second lag from filename timestamp)
PLD (Pressure/Leak): 04.41.00 (94 second lag from filename timestamp)
SAD (Oximetry): 04.41.28 (122 second lag from filename timestamp)

_________________
Mask: Mirage Liberty™ Full Face CPAP Mask with Nasal Pillows With Headgear
Humidifier: S9™ Series H5i™ Heated Humidifier with Climate Control
Additional Comments: APAP 5-7, EPR 1, ClimateLine Hose, ClimateControl Auto 27C/80F, ResScan 3.10 / Win7 64, 16GB SanDisk Ultra Class 4 (15MB/s) SD Card

User avatar
alterego61
Posts: 83
Joined: Wed Aug 26, 2009 10:11 pm
Location: Ontario, Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by alterego61 » Thu Apr 15, 2010 4:16 pm

Comparing the serial numbers on our respective machines, assuming larger serial numbers were manufactured later, this is the order from the data I have received:

Newest: alterego61 - Nord - DreamDiver - fiberfan :Oldest

_________________
Mask: Mirage Liberty™ Full Face CPAP Mask with Nasal Pillows With Headgear
Humidifier: S9™ Series H5i™ Heated Humidifier with Climate Control
Additional Comments: APAP 5-7, EPR 1, ClimateLine Hose, ClimateControl Auto 27C/80F, ResScan 3.10 / Win7 64, 16GB SanDisk Ultra Class 4 (15MB/s) SD Card

User avatar
DreamDiver
Posts: 3082
Joined: Thu Oct 04, 2007 11:19 am

Re: S9 flow skew bug... Anyone else seeing this?

Post by DreamDiver » Thu Apr 15, 2010 7:22 pm

alterego61 wrote:Comparing the serial numbers on our respective machines, assuming larger serial numbers were manufactured later, this is the order from the data I have received:

Newest: alterego61 - Nord - DreamDiver - fiberfan :Oldest
We still haven't finished the preliminary SD card testing yet, I know, but things keep looking less like SD card and more like firmware. Your suggestion of looking at serial numbers is intriguing. If every S9 has the same firmware version, the problem may be that some of us have a flawed EPROM or some other flawed electronic component on the S9's 'motherboard. Looking at serial numbers may be a method to discover this. If we find that those within a certain serial number range are flawed, the indicators could point in that direction.

alterego61 already seems to have the beginnings of this in hand. I know it's yet another direction entirely, and I don't want to volunteer him for something without asking him if he's willing to do it, but if others are willing to PM him with their serial numbers and whether they are seeing skew, he may be able to determine for us whether there is a range of flawed S9's. If there is, the culprit may not be the SD card, but rather some other core component of the S9.

alterego61 - is this an avenue you would be willing to pursue if others are willing to send you their serial number and confirmation (or not) of skew? Preliminarily, it may be a case of first isolating those who have no skew and asking them for serial numbers.

_________________
Mask: ResMed AirFit™ F20 Mask with Headgear + 2 Replacement Cushions
Additional Comments: Pressure: APAP 10.4 | 11.8 | Also Quattro FX FF, Simplus FF

User avatar
Nord
Posts: 565
Joined: Sun Mar 07, 2010 4:30 pm
Location: GTA Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by Nord » Thu Apr 15, 2010 7:57 pm

Something just occurred to me that I hadn't given much thought.

I have always formatting my cards in Windows and I think that was from close to the beginning.
If someone gets a new S9 and doesn't format the card... then the S9 or factory has already done that... correct ???
If they add an instruction file or just some instructions within other files to tell it how to write in a certain way...
And that is not the same as Windows writes it... ???

alterego... are you seeing anything in the formatting system that might match that theory ???

Nord

User avatar
alterego61
Posts: 83
Joined: Wed Aug 26, 2009 10:11 pm
Location: Ontario, Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by alterego61 » Thu Apr 15, 2010 7:58 pm

I'm happy to collect the serial number information if people would like to PM me and let me know whether they are seeing skewed data or not.

_________________
Mask: Mirage Liberty™ Full Face CPAP Mask with Nasal Pillows With Headgear
Humidifier: S9™ Series H5i™ Heated Humidifier with Climate Control
Additional Comments: APAP 5-7, EPR 1, ClimateLine Hose, ClimateControl Auto 27C/80F, ResScan 3.10 / Win7 64, 16GB SanDisk Ultra Class 4 (15MB/s) SD Card

User avatar
alterego61
Posts: 83
Joined: Wed Aug 26, 2009 10:11 pm
Location: Ontario, Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by alterego61 » Thu Apr 15, 2010 8:04 pm

@Nord, no, I have not seen anything like that in the files.

_________________
Mask: Mirage Liberty™ Full Face CPAP Mask with Nasal Pillows With Headgear
Humidifier: S9™ Series H5i™ Heated Humidifier with Climate Control
Additional Comments: APAP 5-7, EPR 1, ClimateLine Hose, ClimateControl Auto 27C/80F, ResScan 3.10 / Win7 64, 16GB SanDisk Ultra Class 4 (15MB/s) SD Card

User avatar
alterego61
Posts: 83
Joined: Wed Aug 26, 2009 10:11 pm
Location: Ontario, Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by alterego61 » Thu Apr 15, 2010 8:13 pm

@Nord, to be clear, the file Identifcation.tgt in the root directory of the SD card seems to be the one that "defines" the environment. It is a plain text file containing the following fields:

#VRN
#VPC
#VID
#SRN
#SID
#PNA
#PCD
#PCB
#MID
#FGT
#BID

You can open yours with a plain text editor like Notepad to see what values yours contains. I would be careful to make sure you do not make any changes to it - there is also a file Identification.crc which looks like the checksum file...if you make changes to the .tgt file so it does not match the .crc file it may well create problems.

My SD card started off blank formatted in Windows, so this file must be written to the SD card by the S9 if it sees a blank card.

_________________
Mask: Mirage Liberty™ Full Face CPAP Mask with Nasal Pillows With Headgear
Humidifier: S9™ Series H5i™ Heated Humidifier with Climate Control
Additional Comments: APAP 5-7, EPR 1, ClimateLine Hose, ClimateControl Auto 27C/80F, ResScan 3.10 / Win7 64, 16GB SanDisk Ultra Class 4 (15MB/s) SD Card

User avatar
Nord
Posts: 565
Joined: Sun Mar 07, 2010 4:30 pm
Location: GTA Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by Nord » Thu Apr 15, 2010 8:23 pm

Thanks alterego...

When I think about it... of course it writes that additional files. I already knew that. It was a theory born in a daze.

I'm just overtired... going to bed in a minute with hope for long restful sleep.

Nord

User avatar
DreamDiver
Posts: 3082
Joined: Thu Oct 04, 2007 11:19 am

Re: S9 flow skew bug... Anyone else seeing this?

Post by DreamDiver » Thu Apr 15, 2010 8:34 pm

alterego61 wrote:I'm happy to collect the serial number information if people would like to PM me and let me know whether they are seeing skewed data or not.
Wonderful. It would probably be best to get serials from those who have absolutely no skew first to check preliminarily whether this is an avenue worth persuing. If you see serials interspersed with those known to be in a range of skewed S9's, this is probably not a good avenue.

Anyone not seeing skew?

_________________
Mask: ResMed AirFit™ F20 Mask with Headgear + 2 Replacement Cushions
Additional Comments: Pressure: APAP 10.4 | 11.8 | Also Quattro FX FF, Simplus FF

kennethryan
Posts: 153
Joined: Sat Oct 13, 2007 8:01 pm

Re: S9 flow skew bug... Anyone else seeing this?

Post by kennethryan » Thu Apr 15, 2010 9:18 pm

Nord wrote:Something just occurred to me that I hadn't given much thought.

I have always formatting my cards in Windows and I think that was from close to the beginning.
If someone gets a new S9 and doesn't format the card... then the S9 or factory has already done that... correct ???
If they add an instruction file or just some instructions within other files to tell it how to write in a certain way...
And that is not the same as Windows writes it... ???

alterego... are you seeing anything in the formatting system that might match that theory ???

Nord
Argh, while I was looking around at the menus on my new machine I accidentally turned it on for a few
second. So the SD card filesystem report is not *quite* a virgin Resmed card...

Here's the info I glean looking at the card on my Linux box:

Partitioned: 1 partition, type 0xB (W95 FAT32)

Label: No volume label

dosfsck output:

ryan@khym:~$ sudo dosfsck -nv /dev/sdc1
dosfsck 3.0.3 (18 May 2009)
dosfsck 3.0.3, 18 May 2009, FAT32, LFN
Checking we can access the last sector of the filesystem
Boot sector contents:
System ID "MSWIN4.1"
Media byte 0xf8 (hard disk)
512 bytes per logical sector
4096 bytes per cluster
32 reserved sectors
First FAT starts at byte 16384 (sector 32)
2 FATs, 32 bit entries
995840 bytes per FAT (= 1945 sectors)
Root directory start at cluster 2 (arbitrary size)
Data area starts at byte 2008064 (sector 3922)
248950 data clusters (1019699200 bytes)
63 sectors/track, 255 heads
247 hidden sectors
1995529 sectors total
FATs differ but appear to be intact. Using first FAT.
Checking for unused clusters.
Checking free cluster summary.
Free cluster summary uninitialized (should be 248936)
Auto-setting.
Leaving file system unchanged.
/dev/sdc1: 6 files, 14/248950 clusters



Curious that the two FATs differ. I don't have a sector editor handy, or I'd investigate this further.
A quick glance of a hex dump seems to show one FAT marks a bunch of single-cluster files while
the second has a few file chains (of course I could be dumping the wrong area; I'm not real
experienced at disk hacking).

So one thing to check: when you format the card on a PC make sure it gets formatted as FAT32
not FAT16, and with long filename capability (VFAT). I've had problems with embedded systems
getting it wrong in the other direction.

_________________
MaskHumidifier
ken

User avatar
alterego61
Posts: 83
Joined: Wed Aug 26, 2009 10:11 pm
Location: Ontario, Canada

Re: S9 flow skew bug... Anyone else seeing this?

Post by alterego61 » Thu Apr 15, 2010 10:36 pm

I have not had as much time as I would have hoped tonight to look at data, but here is an interesting set of files from Nord (sorry this doesn't format too well without a table):

Timestamps
File Type Sampling Freq Sample Interval Header Rescan Start/Finish
20100415_045646_SAD.edf Oximetry 1 Hz 1 sec 04.57.12 (no data) (no data)
20100415_045646_PLD.edf Pressure/Leak 0.5 Hz 2 sec 04.57.05 04:57:04 05:06:02
20100415_045645_BRP.edf Flow 25 Hz 0.04 sec 04.56.53 04:56:53 05:06:53
20100415_045645_EVE.edf Event/Annotation Event Driven Event Driven 04.56.59 n/a n/a

Notes:

- two of the files (SAD,PLD) have names with embedded timestamps which imply a 1 sec lag in time of creation
- it looks like ResScan made allowance for the 1 sec lag in file creation by adjusting the start time of the graph for Pressure/Leak back by 1 sec
- the ResScan graph for Flow Data lasts exactly 10 minutes
- the pressure/leak data graph runs 2 secs short of 9 minutes
- sample interval for pressure/leak data is once every 2 secs
- there is skew of 12 secs in the timestamps in the file headers of the BRP (flow) and PLD (pressure/leak) files; the PLD file lags
- from earlier testing, the BRP and PLD files are only written to SD card once a minute

I have no time left this evening to compare other files or draw conclusions from this in detail, but this has set me thinking. I hope I can make sense of it tomorrow. At a minimum maybe it will help ME sleep better tonight!

_________________
Mask: Mirage Liberty™ Full Face CPAP Mask with Nasal Pillows With Headgear
Humidifier: S9™ Series H5i™ Heated Humidifier with Climate Control
Additional Comments: APAP 5-7, EPR 1, ClimateLine Hose, ClimateControl Auto 27C/80F, ResScan 3.10 / Win7 64, 16GB SanDisk Ultra Class 4 (15MB/s) SD Card