[ANSS-netops] FW: strange Reftek130 timing jumps -1 sec, then + 1 sec hours later [USGS]
David Oppenheimer
oppen at usgs.gov
Mon Aug 4 17:24:44 GMT 2008
If you operate Reftek 130's, please read from the bottom up. Because this
problem will take months to address, our solution is to continue waveform
acquisition but not to time arrivals. This will allow us to use the data for
ShakeMaps etc should a big quake occur.
-David
-------------------------------------------------------
David Oppenheimer office:650.329.4792
U.S. Geological Survey fax: 650.329.4732
345 Middlefield Road.-MS 977 email: oppen at usgs.gov
Menlo Park, CA 94025
_____
From: Ian Billings [mailto:I.Billings at reftek.com]
Sent: Monday, August 04, 2008 10:09 AM
To: oppen at usgs.gov; Customer Support
Cc: NCSS
Subject: RE: strange Reftek130 timing jumps -1 sec, then + 1 sec hours later
[USGS]
David,
You will be able to do the upgrade yourselves as it will be a simple
replacement of the 130-GPS/01 unit for the 130-GPS new Garmin unit.
I will have sales send you a quote today.
Thanks,
Ian billings.
Customer Service Manager
Refraction Technology Inc
1600 10th Street Plano TX 75074 USA
Ph 214 440-1265 or 214 440-1289 (direct)
www.reftek.com ftp.reftek.com
"Want to become an expert with REF TEK equipment? Attend our REF TEK
University. Follow the link on our web site www.reftek.com"
_____
From: David Oppenheimer [mailto:oppen at usgs.gov]
Sent: Monday, August 04, 2008 12:06 PM
To: Ian Billings; Customer Support
Cc: 'NCSS'
Subject: RE: strange Reftek130 timing jumps -1 sec, then + 1 sec hours later
[USGS]
Hi Ian,
I'll ok for the purchase of the clock replacements without input from Lynn
and others since they are out on vacation this week.
Can we do the upgrade or must we return all the instruments to Reftek? If
the latter, will there be additional charges?
The amount of this purchase exceeds our credit card limit, so I'll have to
do a purchase order. Can you send me an official quote?
-David
-------------------------------------------------------
David Oppenheimer office:650.329.4792
U.S. Geological Survey fax: 650.329.4732
345 Middlefield Road.-MS 977 email: oppen at usgs.gov
Menlo Park, CA 94025
_____
From: Ian Billings [mailto:I.Billings at reftek.com]
Sent: Monday, August 04, 2008 6:16 AM
To: Lynn Dietz; Customer Support
Cc: oppen at usgs.gov; croker at usgs.gov; kohler at usgs.gov; macbeth at usgs.gov
Subject: RE: strange Reftek130 timing jumps -1 sec, then + 1 sec hours later
[USGS]
Lynn,
The GPS clocks, SiRF, are the problem here. The 130 will try to keep
correcting for the time jumps but can only go off what the clock is telling
it.
The only real solution here is to replace all the SiRF clocks, 36, with
Garmin clocks. It has been 2yrs since we were told by SiRF that there will
be no upgrade to the firmware to solve the problems you are seeing therefore
the only thing we can do is offer the Garmin GPS at our cost of $170 each.
Please let me know what you decide or if you need to discuss this timing
problem further.
Best regards,
Ian Billings.
_____
From: Lynn Dietz [mailto:dietz at usgs.gov]
Sent: Fri 8/1/2008 5:16 PM
To: Ian Billings; Customer Support
Cc: oppen at usgs.gov; croker at usgs.gov; kohler at usgs.gov; macbeth at usgs.gov
Subject: RE: strange Reftek130 timing jumps -1 sec, then + 1 sec hours later
[USGS]
After getting RTPD to write to its archive again, I compiled a
spreadsheet (attached) lising all of our Reftek130 stations,
their firmware, GPS clock type and whether they shifted
time -1s on 2008/07/31.
Note that we have 56 Refteks deployed:
36 are using SiRF GPS and they all shifted by -1 s
17 are using Garmin GPS and none had a time shift
3 are offline or have no GPS signal
After the initial -1 s shift, most units shifted back +1 sec on
7/31 1900ish UTC (a few did not), then -1 s on 8/1 0100ish UTC.
It's now 8/1 2200 UTC, and there has not been a +1s shift back,
so it appears that all of our SiRF Refteks are running 1 second off.
Earthquake arrival residuals for a unit at Geysers seem to confirm
this (-1 s for GDXB). I haven't checked residuals for other stations.
I am leaving for vacation tomorrow and I'm not sure if anyone else
will be able to follow this while I'm gone.
Lynn Dietz
At 01:05 PM 8/1/2008, Lynn Dietz wrote:
>I found the SOH from a datalogger (NC.NLH, DAS 9581) that did the
>timeshift both 7/31 and 8/1. The SOH from 7/31 is attached.
>
>20080731_UTC_01:10:04 reftek2ew(9): Error: Starttime of tracebuf
>earlier than last timestamp: (NLH HNZ NC --) 1217466588.710 -
>1217466597.620, 892, 100.0, lp 1217466589.700 (9581 S0 C0)
>20080801_UTC_01:05:00 reftek2ew(9): Error: Starttime of tracebuf
>earlier than last timestamp: (NLH HNZ NC --) 1217552685.430 -
>1217552694.340, 892, 100.0, lp 1217552686.420 (9581 S0 C0)
>
>Lynn
>
>
>At 12:20 PM 8/1/2008, Ian Billings wrote:
>>Thanks Lynn,
>>
>>Yes we would be very interested in any SOH files you may have in the
>>future. My guess is the stations behaving like this have the SiRF
>>Trimble type, 130-GPS/01, which has suffered in the past with second
>>jumps due to its firmware. There is no fix to this firmware
>>unfortunately.
>>
>>SOH information shows "GPS TYPE NEMA" for a SiRF type module and "GPS
>>TYPE NEMA Garmin" for the Garmin type which has regular firmware
>>updates, currently at 4.30
>>
>>Later 130 code 2.8.8 checks for time discrepancies and corrects them if
>>they are constant for more than 3hrs.
>>
>>SOH files will definitely help us with this problem.
>>
>>Ian.
>>Customer Service Manager
>>Refraction Technology Inc
>>1600 10th Street Plano TX 75074 USA
>>Ph 214 440-1265 or 214 440-1289 (direct)
>>www.reftek.com ftp.reftek.com
>>"Want to become an expert with REF TEK equipment? Attend our REF TEK
>>University. Follow the link on our web site www.reftek.com"
>>-----Original Message-----
>>From: Lynn Dietz [mailto:dietz at usgs.gov]
>>Sent: Friday, August 01, 2008 1:50 PM
>>To: Customer Support
>>Cc: oppen at usgs.gov
>>Subject: strange Reftek130 timing jumps -1 sec, then + 1 sec hours later
>>
>>For the last 2 days, I've noticed some strange coincidental timing
>>shifts on a large number of our Reftek130s. We have many Reftek
>>acquisition systems; I've checked two of them and about 15 of 26
>>units have displayed this behavior. I expect that many more DASes
>>on the other acquisition systems are behaving the same way.
>>
>>On 2008/07/30 everything is normal.
>>Starting near 2008/07/31 0100 UTC, many units shifted time by -1 sec.
>>Then around 2008/07/31 1900 UTC, they shifted time back +1 sec.
>>Again around 2008/08/01 0100 UTC, they shifted -1 sec again.
>>It's almost 1900 UTC, we'll see if they shift back.
>>
>>The timeshifts were noted in my reftek2ew and Earthworm gaplist
>>logs. I've attached a file containing semi-sorted log info for
>>one channel of each misbehaving Reftek130 datalogger.
>>
>>Unfortunately, RTPD had stopped writing to its archive, so I don't
>>have SOH for any of the units. The dataloggers don't appear to have
>>a backup of SOH either (only trigger files). I just restarted the
>>archive, so if the jumps happen again, I'll be able to send you SOH.
>>
>>I don't know if it's related, but the GPS almanac was updated on
>>2008/07/28 1600ish UTC to include information about the up-coming
>>leap second, and we are seeing -1 sec time issues on some other
>>vendor's dataloggers since that time.
>>
>>Lynn Dietz
>>
>>_______________________________________________________
>>Lynn Dietz email: dietz at usgs.gov
>>U.S. Geological Survey phone: 650-329-5520
>>345 Middlefield Rd - MS 977 fax: 650-329-4732
>>Menlo Park, CA 94025
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geohazards.usgs.gov/pipermail/anss-netops/attachments/20080804/1c85ff09/attachment-0001.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: NC.reftek.timeshift.xls
Type: application/vnd.ms-excel
Size: 18944 bytes
Desc: not available
Url : http://geohazards.usgs.gov/pipermail/anss-netops/attachments/20080804/1c85ff09/attachment-0001.xls
More information about the ANSS-netops
mailing list