OSCAR time off by a couple minutes

[Millstone50]

I thought I was seeing a fluke so I made a point of checking my actual wake times against OSCAR, and OSCAR is apparently timestamping things 3 minutes faster than they're recorded. My therapy ended at 7:53am (by the time I checked the NTP-synced clock). My EDF files on the last segment of my therapy have '070256' at the end, but OSCAR has 7:55:56. When is set my time drift in OSCAR to -3 minutes, all my events more accurately line up in Fitbit as well. Do we know why this might be happening?

[Crimson Nape]

Your file timestamp is almost 50 minutes off from the other two reporting times.

[Millstone50]

Your file timestamp is almost 50 minutes off from the other two reporting times.

You're right. I've been looking at so much time data this AM the numbers have started to bleed together. I'll look for some other data.

It could just be the device's clock is off, but I have no way to check. I could probably insert the SD card in, start therapy then stop it right away and see what it is against a real clock.

Edit: OK, I did that. the timestamp on the files at 9:26am UTC-5 was '092920' and date modified 2:29pm (which must be UTC). So the device's clock is fast. Why isn't it setting itself correctly? 

[ScottZZZ]

So the device's clock is fast. Why isn't it setting itself correctly? 

You will have to ask Resmed.  Also, if you search the forum you'll find several threads that have already discussed this in more detail. 

We don't know how (or if) the clock is maintained/synced.  It doesn't appear to use cell tower time, as it differs from cell phones on same service.  It may be getting synced when connecting to the MyAir server, or perhaps only when the DME sends Rx data to the machine.  Or it may be set at the factory and never updated.  We simply don't know.

Based on comparisons I did with an IPCam and a Usleep oximeter (both of which are synced to NTP), I determined my AS 11 was 39 seconds behind.  Months later, it is still 39 seconds behind -- so either the clock crystal is very accurate, or it is being periodically synced to some non-NTP source.  I corrected for that 39s with OSCAR's "CPAP Clock Drift" setting, and all my data aligns well.

[Millstone50]

You will have to ask Resmed.  Also, if you search the forum you'll find several threads that have already discussed this in more detail. 

We don't know how (or if) the clock is maintained/synced.  It doesn't appear to use cell tower time, as it differs from cell phones on same service.  It may be getting synced when connecting to the MyAir server, or perhaps only when the DME sends Rx data to the machine.  Or it may be set at the factory and never updated.  We simply don't know.

Based on comparisons I did with an IPCam and a Usleep oximeter (both of which are synced to NTP), I determined my AS 11 was 39 seconds behind.  Months later, it is still 39 seconds behind -- so either the clock crystal is very accurate, or it is being periodically synced to some non-NTP source.  I corrected for that 39s with OSCAR's "CPAP Clock Drift" setting, and all my data aligns well.

I'll keep my OSCAR offset as well. I found the other threads. I also reached out to my DME and ResMed (both of which will end up being futile).

I just can't wrap my head around it. This device absolutely can and should be synchronized to an authoritative time source.

[ScottZZZ]

I also reached out to my DME and ResMed (both of which will end up being futile).

I strongly, strongly suspect that your expectations are 100% correct.  

[Millstone50]

Official response, "The time on Air11 devices cannot be changed. These devices are connected to the time zone."

Whatever that means.