Various problems arising during operating a sensor can be dealt with on site. Please go through the items thoroughly. You may also consult the sensor's instruction manual.
Types of re-centering failures with STS-2 und STS-2.5
- No failure, just impatience. Please hold on waiting for the green status LED to light up. Wait several minutes.
- Case tilt >0.5° (STS-2.5: FAULT LED on the host box shows one short blink). Turn the feet screws in order to align the bubble of the bubble level and try re-centering again. Tighten the feet screws after aligning.
- The bubble is perfectly aligned; “case tilt >0.5°” is still indicated: Most probably the bubble level is damaged. Please see Repair a sensor
- The seismometer has just been deployed and is now enduring the heaviest phase of temperature equilibration. Temperature gradients within the seismometer provoke large pendulum position drifts. So, in case of a long re-centering time, it is possible that a once properly centered pendulum has drifted to a non-centered state before the re-centering of the other pendulums has finished, and, eventually, the centering process will run into a time-out. Please wait until the temperature has equilibrated enough.
After several autozero cycles one sensor component constantly remains out of range.
See directly above. In case of no success: Assert FAST RESPONSE (via serial interface) and SIGSW (via serial interface or connector pin) and consult the UVWPOS/UVWRAW output (UVW for the appropriate sensor component). If the voltage is still out of range (±10 V), the most probable cause is pendulum sticking or defective centering function. Please contact Streckeisen GmbH.
Notice: the STS-2 has no serial interface, use the connector pins instead.
Offset on the signal outputs
After deployment the instrument adapts slowly to the ambient temperature. Moreover, mechanical settling effects may follow a long storage time, especially when having been stored in the locked state. The drift period lasts for a few hours at the minimum, but may extend to more than one day, especially when the whole site has to equilibrate, too. The characteristic of a high drift is a high "offset" on the output signals, preferentially on the Z output signal. "Offset" in this context means a non-zero voltage remaining constant during many step-response settling periods (one settling period is about 180 s or 540 s for a 360 s instrument). Of course, an "offset" also arises, when the vault temperature is continually changing. If this "offset" exceeds some Millivolts at the differential outputs and has not remarkably diminished after some hours, the temperature drift is very critical. A persistently elevated noise level will be the consequence.
Note: A broad-band signal looks rather different from a short-period one. Any long- period (>120 s or 360 s) change of the internal or external force on the boom deflects the signal proportional to the derivative of this force. Therefore, a constant increase of the force, as it is approximately expressed by a constant temperature drift, converts to a constant “offset” on the output. Normally, short period instruments do not show this behavior or it is concealed by noise.
Deployment related spurious signal phenomena
During some time after deployment, spurious signals can occur, because of mechanical settling effects within the seismometer case. Internal mechanical settling can produce a pendulum force variation that eventually expresses as a voltage equivalent at the outputs. In most cases the pattern of it is distinguishable from the one of a genuine signal, but sometimes not. Provided that the ambient temperature is stable, these distortions disappear after a few days at the latest. Prevent the instrument from any temperature drift higher than a few tenths of a degree Celsius per day! Otherwise, a once distortion-free instrument can resume distortion activity. Distortions of assured internal origin that have remained constant over weeks and months indicate a defective instrument. But keep in mind that, in most cases, distortions are of external origin: Electromagnetic interferences, high temperature drift, pier settling, grains of some non-inert material below the feet, etc. Please contact Streckeisen GmbH in case of uncertainty.
STS-2.5: After deployment, leveling, and autozero one or more position signals remain out of range, indicated by the blue LED’s “NOT CENTERED” on the host box.
Repeat autozero, until all these LED’s are extinguished.
The unfiltered vertical output signal (Z) exhibits several millivolts offset which does not vanish neither on autozero nor after several hours of operation.
See above Offset on the signal outputs.
How can I distinguish between interferences (pings, excursions, “time-marks”) of internal (produced by the instrument itself) or external origin?
Import the suspicious data range into an adequate time-amplitude representation program capable of displaying multiple channels synchronously, filtering and maths, manual zooming, manual gain setting, scrolling through time, locked cursor or locked grid lines. The data range should comprise at least several hours, in order to be able to get an impression of the signal background, where the suspicious interferences are embedded. Then proceed as follows:
- Apply an appropriate band-pass filter to the output signals and, if possible, reduce the sample rate. Proposal: High-pass corner at 1mHz, low-pass corner at 33 mHz, sampling rate set at 100 mHz.
- Transform the output signals into the approximate raw U, V, and W signals by using the formula in the appropriate manual.
- On the display screen present the Z output together with the calculated raw U, V, W outputs. Setting the temporal zoom factor so as to display about 20’000 s on one screen (2000 samples per signal at 100 mHz sampling rate) has proven most effective. Set the gain so as the noise background to be visible. Glitches now are easily recognized.
- Place the cursor to the location of a glitch on Z. Examine U, V, and W amplitudes at the cursor position. If the same glitch is found only on one of the raw signals, the glitch is most probably of internal origin. The raw signal, where the glitch has been found, indicates the affected pendulum or rather its feedback system. Inversely, if the same glitch is found on two or three of the raw signals (Do not care about amplitudes and amplitude ratios!), the glitch is most probably of external origin.
How can I identify signal excursions caused by electro-magnetic interferences? How can I prevent or, at least, mitigate the effects?
Most electro-magnetic interferences can be attributed to currents that are switched on and off in the vicinity of the seismometer (e.g. current through power lines of on-off-controlled battery charger and/or heating device) giving rise to stepwise changes of the magnetic field. Any magnetic field (no matter if DC or AC) exerts a force on the boom. A characteristic property consists in the fact that, in most cases, pulses of the same origin have identical height, and every positive pulse is compensated by a negative one. If it is not possible to deactivate the responsible source or to move it so as to enhance the distance between seismometer and current conducting line(s), the seismometer must be coated with an appropriate magnetic shielding.
STS-2.5: In case of a damaged bubble level: Is it still possible to level the seismometer?
Yes, it is possible when accessing the seismometer μC via serial interface. Doing so, the internal tilt sensor can be read out, enabling to level by observing the X and Y tilt values instead of the bubble level.
An STS-2 can NOT be centered without a working bubble level!