Version 7 (modified by mggr, 16 years ago) (diff)



Most of the useful information is currently on the main ARSF website -


You can find a thesis covering the underlying principles of the Specim instruments (Eagle & Hawk) at


Some details on Hawk's detector from Jukka @ Specim (email 25/July/2008):

The detector type on Hawk is MCT (Mercury-Cadmium-Telluride),
which means that it's not a CCD that is typically used for VNIR
spectral range.

The manufacturer's specification for that MCT detector bad pixel
is the following; if the pixel peak responsivity differs more
than +/- 30% from average responsivity, it is considered to be
bad. According to specification less than 2% of pixels should
be bad. In practice that is from 0.3 to 1.0 %.

The detector is sealed in a vacuum housing. This vacuum "leaks" 
very very slowly. As a result of this new bad pixels may appear.
Eventually the vacuum has to be recovered using a special method. 
According to manufacturer it may take from 6 months to several 
years before this must be done. So far we've done this only for
a single sensor, which had been used already for several years
before that.

By they way, there are no similar bad pixels on CCD used on Eagle.

... response to query from Ben re: frequency of maintenance:

There is no particular way or time to say when that operation 
is due. You will get a hint from additional bad pixels. Also
if the detector temperature is no longer as low as it used
to be that could be an indication of impurity in the vacuum.

You don't want to try that operation periodically. It requires
sending the system here and disassembly of some parts. After 
that that the system must be reassembled and calibrated.

I'm not expecting your system do be due for that operation yet.
And I understand you are recalibrating the system at the moment.
This means that new calibration file will "recover" most of the
partially appearing bad pixels if any.

Notes of interest

Eagle uses a 1024x1024 CCD, but only the middle ~504 spectral pixels are used in normal operation (representing the range from ~390nm to ~950nm) and may be spectrally binned by 2 (giving 252 bands).

However, when used with a band file for "calibration mode", the instrument returns the full data width (i.e. all 1024 pixels). This must be accounted for in any calibration code. (source: phone chat with Bill, 6/Nov/2007, much amended 30/Jul/2007 following corrections to Mike's interpretation from Bill)

Attachments (2)

Download all attachments as: .zip