detpars

• detname: The most important entry in detpars is the chip identification. This should have been set automatically when you ran obsinit at the beginning of your run and specified which chip you were going to use. (See B.) However, it is a good idea to check this, as the chip designation controls what microcode is sent to the CCD controller, and you will be quite surprised at the results if the wrong microcode is loaded. (``detcap" specifies the pathname of where this name is used to find the microcode.) In principle, you will be warned when you take an observation if the chip name does not agree with the name returned from the chip itself. The current chip designations are basically the lower-case version of what is shown on the observing schedule, e.g., ``t5hc'', ``t1ka'', ``t2ka", ``t2kb", ``s2ka'' , ``f3kb'' , ``gcam" ``cryo", or ``ti5".

• firstcol, lastcol, firstrow, lastrow, colbin, rowbin: These parameters describe the ``active" area of the chip that you wish to use, and whether adjacent pixels should be binned or not. For most applications, the chip will be formatted to full size, and no binning used, but in the example shown the astronomer is using the T2KB chip for spectroscopy, and only needed to read out a thousand rows near the center. Note that all the Kitt Peak chips get an extra 32 columns of overscan in order to monitor the electronic pedestal level, and hence in the example shown the images will be (2048+32 = 2080) by 1001 pixels. In the case that you do choose to bin the data, the first and last columns and rows refer to that of the unbinned pixels.

• to preflash or not to preflash: You will only ever have to preflash if you are using a TI chip, in which case you will need to specify the number of seconds of preflash you need---typically 3 to 5. This should be adjusted (along with the ``intensity" digit-switches on the CCD electronics box) until you get 50-100 electrons per unbinned pixel for the TI chip (200-400 electrons if you are binned ).

• gain: All our CCDs other than the TI chips have adjustable gain settings. The gain parameter in detpars is an ordinal number that allows you to change the number of electrons per ADU (Analog to Digital Unit, the numbers you see on the screen). The correspondence between these ordinal settings that you enter in detpars, and the corresponding number of electrons per ADU, can be found typing ccdinfo as long as the proper CCD identification is in detpars. This can also be found by using dcapinfo (for instance, dcapinfo t1ka).

Many of the parameters that the user is likely to change are displayed on the ``status" line that appears at the beginning of each integration: the chip name, the format and binning factors, and the gain.

Digression: How To Choose A Gain

  ``How do I choose a gain setting?" is one of the most asked questions. All of our chips (other than TI5) have gains which are adjustable by the software. Why would you want to adjust the gain? The Tektronix chips all have a full-well capacity of >100,000 , and some as high as 250,000 , before there is any detectable deviation from linearity. However, the A/D converters are limited to 16-bits, and cannot output data that is greater than 65,535. Thus to make full use of the dynamic range you would like the gain factor to be about 4. Why not simply set it to 4? The reason is that most of our chips also have very low read-noise, and thus if the gain is greater than the read-noise, you will be undersampling the read-noise---in effect, increasing the read-noise to the level of the gain simply because of digitization noise. (You can't very well recover a read-noise of 3.5 if each data unit is equivalent to 4.) So like most things in life, there are trade-offs. Up-to-date values of the read-noise and linearity limits for all of our chips can be found in the CCD Characteristics Manual in each dome.

This was more of a problem before the new 16-bit A/D converters were implemented; now, most users will be happy by just accepting the default gain settings (see Sec. Q).

If you are attempting to do 1% stellar photometry of stars in a cluster, you are probably interested in covering as large a magnitude range as possible, and furthermore, your noise is going to be primarily photon-noise, not read-noise. Go for the largest value of per ADU as you can without exceeding the linearity of the particular chip. Generally, this will be the default gain, which is also the gain that will give you the least amount of horizontal bleeding from very saturated stars.

If you are doing surface brightness studies of objects through narrow-band filters, and the read-noise is significant but the dynamic range of your objects is limited, you may wish to stay with the largest gain number (smallest number of per ADU). Similarly in some very low-signal spectroscopic applications you are limited by the read-noise.

Note: Some of our CCDs show some very low-level ``streaking" to the right of the most heavily exposed stars. This problem is significant with S2KA and T2KB. The electronics has been adjusted to minimize this problem for the default gain setting. If you are concerned about the effects of very saturated stars on your direct imaging data, you would do well to stay with the default gain settings.

Digression: The Power of ccdinfo

  Once your detpars is setup, you can verify all of these important parameters by simply typing the command

ccdinfo

You will get a response that resembles that of Figure 3.

If you ever wish to get information about all the (major) chips available, one can do a dcapinfo show=list. Information about a particular chip can be found by typing dcapinfo t1ka, for example.