Pan-STARRS bandpass filters

The desiderata for the Pan-STARRS broadband filters (other than the Solar System filter) are:

• Number of filters must be small because of filter and filter changer costs
• Pass bands must be broad to cover the wavelength range of silicon detectors with a small number of filters
• Filter pass bands must not overlap
• Pass bands must be narrow enough that differential atmospheric refraction and chromatic aberration are not troublesome
• The filter system must be similar to an existing system
• Avoid the variable extinction of the water absorption feature at 930-960 nm

The filter set consists of g, r, i, z and Y filters. The first three have the bandpasses of the corresponding Sloan Digital Sky Survey filters. The z filter has the SDSS shortwave cutoff, but its longwave cutoff avoids water absorption bands beginning at 930 nm. The corresponding SDSS filter has no longwave cutoff, but the longwave response of the SDSS CCDs (SiTE) falls nearly to zero by 1050 nm.

The shortwave cutoff of the Y filter is set by the water absorption bands that end around 960 nm. The longwave cutoff of the Y filter is under discussion. The design presented herein is that proposed by John Tonry with a longwave cutoff at 1030 nm to avoid the worst of the sensitivity of the detector to temperature variations.

Filter designs

Departures from the design performance (shown herein) occur because the thickness of the dielectric layers cannot be controlled precisely. The Solar System filter design came with the following comment that likely applies here as well.

I'm sure we could get to about 85% average transmittance through the passband. With the AR on the back it could be close to 90%.

These designs were provided with the following comment.

Here are the data for the various filters, q2003-150. Please review the designs and let me know if I need to modify the performance of any of the filters. Where you find "spikes" between 300 and 350nm assume that the transmission will be less than .1%. The materials I will use will absorb the light and I just don't include the n-ik value in the index for modeling purposes.

Geza Keller

This is the fallback solar system filter (2003.02.10 Solar System bandpass filter) for comparison purposes. It is design q2003-51, February, 2003.

This is the g filter design transmission. Note the logarithmic transmission scale. Pursuant to Keller's comment at the beginning of this section, the transmission between 300 and 350 nm will be less than 0.1%.

This is the r filter design transmission. Pursuant to Keller's comment at the beginning of this section, the transmission between 300 and 350 nm will be less than 0.1%.

This is the i filter design transmission. Pursuant to Keller's comment at the beginning of this section, the transmission between 300 and 350 nm will be less than 0.1%.

This is the z filter design transmission.

This is the Y filter design transmission.

This is the design transmission of all filters with a linear transmission scale.

Designs are provided by:
Geza L. Keller
QSP Optical Technology, Inc.
1712-F Newport Circle
Santa Ana, California 92705 USA
Tel: 714-557-2299
Fax: 714-557-2170

The filter designs are all broader than the specifications. It is likely that this makes the designs easier and the filters less expensive. The 50% transmission points of the filter designs are given in the following table.

Specifications

The specifications that were provided to QSP follow:

The 50% transmission points are given in the following table.

Date created: 2003.05.23
Last modified: 2003.12.19
Copyright © 2003, Walter A. Siegmund

Return to report index