[ANSS-netops] reftek unity or x32?

[Philip Crotwell]

I think we have mostly observed it with C2, but am not positive we
haven't seen it occasionally with CO. CO should have a max compression
ratio of a bit below 4, and reftek packets are 1024 bytes, so it
should be unlikely to get more than the 1008 samples in a packet. And
our seismometers are unknown passive borehole instruments, so very low
output voltage, so you will likely have less trouble. But...

Anyway, I guess it is just one of those things that keeps this job
"interesting", right?

Philip

On Mon, Mar 21, 2011 at 12:02 PM,  <mwithers at memphis.edu> wrote:
>
> So you see this in both CO and CO2?  Did you try the 32 data format?
> We're currently using CO with unity gain on the broadbands and haven't
> noticed gaps.  Though it sounds like it might be prudent to fix the
> code or switch to 32 if we have the bandwidth.
>
> Mitch
>
> Center for Earthquake Research and Information (CERI)
> University of Memphis                Ph: 901-678-4940
> Memphis, TN 38152                   Fax: 901-678-4734
>
>
> On Mon, 21 Mar 2011, Philip Crotwell wrote:
>
>> Hi Mitch,
>>
>> We have a couple of stations with reftek 130s at 32x gain, but the
>> reason has nothing to do with trying to make the data "better". Rather
>> there is a bug in the reftek to earthworm module such that if you are
>> using compression (CO or C2 in the reftek) and the data is relatively
>> quiet (low count range), then the reftek can compress more the data
>> for transmission in each packet (>1008 samples) than earthworm allows
>> in its internal tracebuf (1008 samples == 4032 bytes) packets. Hence
>> earthworm tosses those packets and you have lots of little gaps.
>>
>> You might not have this problem with an S13, but if you see messages
>> from reftek2ew about packet too large, then you might need to ether
>> back off on the compression, or bump up the gain so that it doesn't
>> compress as well. The other odd thing about this is that the error
>> message from reftek2ew tells you that the packet is too big, but
>> doesn't tell you which station it came from, so finding the guilty
>> party can be a big challenging. The message (actually reported by the
>> wave server) looks like:
>>
>> 20080702_UTC_01:50:12 wave_serverV: tport msg[6288] i255 m50 t19 too
>> big. Max is 4096
>>
>> It would be nice if reftek2ew would recognize that the packet was too
>> big and split it into two, but my understanding is that ain't the way
>> it works now.
>>
>> Good luck,
>> Philip
>>
>> On Mon, Mar 21, 2011 at 11:02 AM, Mark E. Meremonte <meremonte at usgs.gov>
>> wrote:
>>>
>>> Hi Mitch,
>>>
>>> I discovered that the S13 is about 30 times more sensitive than the L22.
>>> Thus I would use unity.
>>>
>>> Mark
>>>
>>> At 01:10 3/18/2011, Mark E. Meremonte wrote:
>>>
>>> In Haiti we used L22s with 32x and .8 damping with good success.
>>> However,
>>> the S13s are more sensitive.  Croker makes good point that local s/n
>>> conditions may help the decision.
>>>
>>> Mark
>>>
>>>
>>> At 16:49 3/17/2011, mwithers at memphis.edu wrote:
>>>
>>> We're upgrading some of our analog 3-c S-13 sites to digital telemetry
>>> and Reftek RT130 DAS's.  We're trying to decide whether to run at
>>> unity or x32 gain.  Being lazy and indecisive, my immediate question
>>> is what do other people do?
>>>
>>> Mitch
>>>
>>> Center for Earthquake Research and Information (CERI)
>>> University of Memphis                Ph: 901-678-4940
>>> Memphis, TN 38152                   Fax: 901-678-4734
>>>
>>> _______________________________________________
>>> ANSS-netops mailing list
>>> ANSS-netops at geohazards.usgs.gov
>>> https://geohazards.usgs.gov/mailman/listinfo/anss-netops
>>>
>>> *********************************************************
>>> Mark E. Meremonte        Geophysicist    USGS/ANSS/NEIC
>>>
>>> U.S. Geological Survey        Work: 303-273-8670   Cell: 303-478-5766
>>> MS 966, Box 25046          Email: meremonte at usgs.gov
>>> Denver, CO  80225             Web: http://geohazards.cr.usgs.gov
>>>
>>> Ship:  U.S. Geological Survey, 1711 Illinois St., Golden, CO 80401
>>> **********************************************************
>>> _______________________________________________
>>> ANSS-netops mailing list
>>> ANSS-netops at geohazards.usgs.gov
>>> https://geohazards.usgs.gov/mailman/listinfo/anss-netops
>>>
>>> *********************************************************
>>> Mark E. Meremonte        Geophysicist    USGS/ANSS/NEIC
>>>
>>> U.S. Geological Survey        Work: 303-273-8670   Cell: 303-478-5766
>>> MS 966, Box 25046          Email: meremonte at usgs.gov
>>> Denver, CO  80225             Web: http://geohazards.cr.usgs.gov
>>>
>>> Ship:  U.S. Geological Survey, 1711 Illinois St., Golden, CO 80401
>>> **********************************************************
>>>
>>> _______________________________________________
>>> ANSS-netops mailing list
>>> ANSS-netops at geohazards.usgs.gov
>>> https://geohazards.usgs.gov/mailman/listinfo/anss-netops
>>>
>>>
>