WHT oscillations

Small oscillations in effective pointing can be expected at any of the characteristic frequencies of the telescope and its control system, e.g. those frequencies associated with the mains electricity (50 Hz), oil pumps (70 Hz), gear trains, time service, mirror-support system and telescope structure (natural frequency about 4 Hz).

Oscillations of frequency N Hz will in general only be cancelled if autoguiding at a frequency >~ 10 N Hz, and with suitable loop gain. For most instruments, the autoguiding loop typically lasts >~ 3 sec (>~ 1 sec exposure + 2 sec overheads), i.e. the loop frequency is <~ 0.3 Hz, so small oscillations with frequency < 0.03 Hz should adequately be removed by autoguiding. (NB for the AO system NAOMI, now decommissioned, the guide frequency was much higher, several 10s of Hz.)

In practice, oscillations are rarely a problem for observers, but there are occasional reports of small oscillations of various frequencies. The notes below relate to an ongoing investigation of 1-arcsec oscillations with period ~ 1 minute at the Cassegrain focus, and an investigation of rapid oscillations observed at the GRACE Nasmyth focus.

Information about the large-amplitude azimuth oscillations which blighted observing in late 2013 and early 2014 can be found on a separate page. These oscillations were due to a failed electronic component, now repaired.

Cassgegrain oscillations

In April 2009, Danny Steeghs (Warwick) obtained with WHT ISIS a series of short (2-sec, autoguided) spectra of 2 objects placed symmetrically on the ISIS slit. In the figure below, the top left plot shows:

1. A slow drift in the position (along the slit) of one of these two objects, which might be ascribed to flexure between ISIS and the autoguider.
2. An oscillation of full amplitude ~ 1 arcsec (5 pixels) and period ~ 67 sec (see blow-up in middle plot in top row, and power spectrum at right):

On 2009 Jun 29, Miguel Santander obtained a series of 10-sec J-band images of stars using LIRIS (0.25 arcsec/pixel)mounted at the Cassegrain focus, with the telescope autoguiding (in the optical). The star centroid positions (measured by Pablo Rodriguez) are plotted below (plot width ~ 0.008 days = 700 sec):

	Star 1 (1418-06)	Star 2 (1440+40)	Star 3 (1415+29)
X
Y

There is no evidence here for oscillations. However, there are surprisingly large drifts in position, up to 0.5 arcsec over a few minutes. The origin of these drifts is unknown.

Nasmyth oscillations

Rapid image oscillations, with amplitude ~ 1 arcsec, were sometimes seen (on < 10% of nights) when observing with the adaptive-optics suite NAOMI. 'Rapid' means too fast for the AO system to correct, so probably faster than 5 Hz. The oscillations were known to originate outside the Nasmyth (GRACE) optical bench hosting NAOMI (because the oscillations changed orientation when the derotator was rotated).

There was no evidence, from observations at other focal stations, that the telescope suffered such oscillations, so suspicion fell on the Nasmyth (tertiary) flat.

To test this, and to check for any dependence on telescope position in azimuth and elevation, an accelerometer was installed on the support structure of Nasmyth flat in 2016, and measurements of oscillations in left-right and tip-tilt directions were taken by Karl Kolle, at a series of 18 positions in azimuth, and 12 in elevation.

The resulting power spectra showed the expected peaks from the mains electricity (50 Hz) and the oil pumps (70 Hz) but at all frequencies the amplitude of the movement was < 10 nm, orders of magnitude too small to explain the oscillations seen at Nasmyth. There were no significant differences in the spectra taken at different telescope positions.

There is therefore no evidence that vibration of the Nasmyth mirror was responsible for the oscillations seen at the GRACE Nasmyth focus, and the source of these remains a mystery.