Echelle Mode

Filters for Hydra Echelle Mode

Note! Resolutions calculated, not yet measured.

The Hydra echelle mode uses a coarse (316 l/mm) echelle grating, blazed at 63 degrees, giving an effective blaze of roughly 56120A. It operates in what are (for an echelle) very low orders (5-15). This grating is used in order#=56120/lambda, e.g. 10th order near 5600A.

When used with the large (300 microns) fibers, the SiTe 2K x 4K, and the 400mm f.l. Bench Schmidt camera, the projected size of the fiber on the CCD will be about 7 pixels. The dispersion in A/pixel is approximately the wavelength divided by 105,000, e.g. .053A/pixel at 5600A in 10th order. The resolution will be 105,000/7 or approximately 15K. When used with the 200microns slit the projected fiber size will be 4.5 pixels and the resolution about 23K. With the 100 microns slit the projected slit size will be 2.5 pixels and the resolution about 40K. There are two filters for each order, centered at orders N+25 and N+75 where N is the order number. Each filter is made as wide as possible without permitting significant amounts (maximum .2%, average much less) of any undesired order to pass through.

If the filters were perfect, this would mean that a filter centered at order N+25 would cover from (N-1)+75 to N+75 and the filter centered at N+75 would cover from N+25 to (N+1)+25. In practice the filters cannot be made this wide because they have finite "skirts" in their bandpass. Nevertheless, there will always be a filter that covers from the overlap zone below the blaze to well beyond the blaze of every order and another which covers from well below the blaze to beyond the overlap zone of every order.

The free spectral range is larger than the size of the chip. This means that there is always an optimum filter for every wavelength range.

The filters become less efficient and the skirts wider as the wavelength becomes shorter.

The current set of filters was made by cementing two 1/2" x 2" sections together. The interference filter layer does not extend quite to the edge of the original filters, so that light from more than one order passes through the cemented filter in a strip about 2mm wide on each side of the junction. Consequently, some fibers near the junction cannot be used. The bad fibers can be identified by taking an exposure with the quartz lamp. The flux increases in contaminated fibers near the center and decreases at the very center where light is blocked by the cement. About 8-10 fibers in the center of the field are typically contaminated. Fibers #203, 23, 172, 155, 90, 186, 197, 32 and 84 were reported as being compromised in one filter configuration, listed in order from most to least contaminated. The location of the problem varies to some extent between filters.


Filters for Hydra Echelle Mode


Filter Order Center
#3 6.75 8259 1058 84% .gif
#4 7.25 7689 907 87% .gif
#5 7.75 7193 784 80% .gif
#6 8.25 6757 684 79% .gif
#7 8.75 6371 601 78% .gif
#8 9.25 6027 531 75% .gif
#9 9.75 5718 472 80% .gif
#10 10.25 5439 422 78% .gif
#11 10.75 5186 379 75% .gif
#12 11.25 4955 341 75% .gif
#13 11.75 4745 309 75% .gif
#14 12.25 4551 280 70% .gif
#16 13.25 4207 233 47% .gif
#17 13.75 4054 214 44% .gif
#18 14.25 3912 196 35% .gif
#19 14.75 3780 181 40% .gif
#20 15.25 6356 167 50% .gif


Filters #2/8920A and #15/4372A are unavailable until further notice.

Observe that filter #18 does not have a flat top, but has a single central peak at the design wavelength.