Polaris State of Health
The Polaris Libra seismographs record waveform and State of Health (SOH) data. The SOH data are normally recorded once per minute and include essential information such as battery volts, temperature measurements, satellite transmission information at a remote station. The SOH data for all Polaris stations are available on Polaris website in near real-time. Every morning at 7:00 AM EST SOH data for the past 24 hours, and every Monday morning at 7:10 AM EST SOH data for the past 7 days are retrieved, plotted and copied to the website.
A new feature is added to the SOH graphs on August 2002. The station links are color-coded to identify their Polaris Carina Hubs (1 through 4). As well links are created for graphs of the TDMA table for each Carina Hub. These Hub graphs may be slightly out of date as they need to be manually extracted after each major change to the tables.
The SOH data provide a quick visual quality check on Polaris recording. For stations running on solar panel, the battery volt is an important parameter to monitor and the SOH plots quickly indicate if the batteries at the site are fully charged.
The temperature measurements of the Libra seismograph also provide valuable data. For example, the controller and satellite modem temperatures are good indicators of temperature inside the Libra instrument while the SSPB temperature is a good indicator of air temperature at the satellite antenna.
The Polaris Libra seismographs are currently configured with a TDMA frame of 5 seconds. This means that every station gets, at least, one broadcast window every 5 seconds to transmit its data to the central Hub. The number of "good" data packets is therefore expected to be around 12 per minute. The "good" and "bad" packet SOH data provide a quick quality check on the state of transmission from each site.
A few typical SOH plots will be reviewed here to identify key quality controls. For example, this daily plot for station YMBN which is powered by solar panels shows the battery volt rising to more than 15 VDC at midday time and the charge controller remains in bulk charge mode keeping the voltage high for about one hour before switching to a float charge at about 14.5 VDC. This level of charge continues until the late afternoon hours when the battery volts begins to decline and finally settles to just below 13 VDC for the evening. The battery volt for a sunny day like this remains well above the Low Volt Disconnect (LVD) level which is set to 11.5 VDC.
The weekly plots are good indicators for longer term trends of the battery condition. For example, this weekly plot for station ACKN which is powered by solar panels shows consistent and high charge levels for almost all seven days in late October. Weekly plot for same station about a month later shows much less charging due to the fact that the daily amount of sunshine has declined substantially and winter-like conditions such as rain, clouds and snow also contribute to a reduction in charge level.
The Polaris SOH plot, for example for station LDGN shows that while the SSPB (air) temperature has decreased from about -10 C to -30 C during this week in November, the modem temperature inside the battery and instrument vault has declined only from 10 C to about 5 C due to a well insulated vault.
The "Packets" graph on this LDGN SOH plot shows perfect satellite transmission for this particular day since the number of good packets has stayed fixed at 12, as expected, and the number of bad packets has stayed at zero. In fact, the NWT stations show this good transmission quality fairly consistently. For example, this weekly plot for station YMBN shows only a few cases of higher than 12 packets per minute. Same station shows a more noisier transmission during another week. Note that the large sudden drops of the number of good packets appear to be related to a counter resetting and do not indicate problems with transmission.
The SOH plots can help determine breaks in data recording. For example, this SOH plot for station TYNO shows that station went down for about 2 hours (due to scheduled maintenance) around midday on December 2, 2001.
The SOH plots can help identify problems with the hardware. For example, this weekly plot for station ACKN shows a sudden drop in SSPB temperature around October 31. Daily plot for same stations shows that the problem started shortly after 06:00 UT on October 31. This drop in temperature appears to be due to a malfunction in the temperature controller inside the Libra seismograph and does not indicate a real temperature change. This problem does not appear to affect the data transmission adversely.