CCD-world: Looking for CCD...

[Paul.Jorden at eev.com]

The following was posted to CCD-world:

Scott, et al,
 >
 > I have an upcoming project and I am looking for a linear CCD with 5-10 um 

 > pixels, a 25-30 mm width, and a pixel height of at least 200 um.
 > I need decent red and NIR sensitivity over the region 800-1050 nm (at
 > least 5% at 1000 nm, preferably closer to 10).  Back thinned,
 > back illuminated would be even nicer.
 >
 > This is for a spectroscopic application, whichis why I do not need an area CCD.
 >  If that is my only option, I can still use a subwindow and bin it.  However, all the
 > CCD's I can find that meet my needs are 1024 elements or so and all
 > have 24-25 um pixel width.  Most of the high pixel count linear arrays with small
 > pixels havesquare pixels, so the height is very small.  And
 > these long, thin arrays do not have good IR sensitivity.
 >
I know that Don Groom has gven you info on the LBL developments..

Although Marconi do not make exactly what you want, we do make high-resistivity, and backside devices.  (The 30-11 is a 'spectroscopic' standard high-res product- but has larger pixels).
The 42-10 device has a 13 um pixel, with 2k*512 format.  We make other devices in the same series, with high-res and thinned.
The 42-10 could be supplied as high-res, thinned if it was any good to you.
[QE should be 30-40% at 1000nm]

For information, I believe that Marconi devices would have the following RT QE at 1000nm:
frontside,     standard  ~10%
backside, standard  ~20%
high-res, thinned        ~40%
Exact value depends on temperature, AR coating (if thinned), device thickness etc.

I've forwarded your details to a colleague who will reply directly to you about options if you want to discuss further.  This message is also for general circulation/ information.

I would make the same point as Don about loss of resolution in the red.  If you want a high QE, then you need lots of silicon thickness  (because the absorption coefficient
is ~100-200 um at 1000nm wavelength).  In this case the silicon thickness >> pixel size, and so spatial resoltion is more defined by chip thickness, than by pixel pitch!

'Industrial' CCDs sometimes attenuate the red response deliberately in order to try and maintain spatial resolution.
With 'scientific' CCDs there is a trade off between spatial resoltion and QE.
For example, you can add a reflecting 'mirror' on the non-illuminated side of the CCD.
This gives a double-pass, and increases far-red response, but only at the expense of degraded spatial resolution.
[see, for example,  http://adcam.pha.jhu.edu/detectors/halo/]

Regards,
Paul
_____________________________________________________________
Dr Paul Jorden,  CCD Sensors, Marconi Applied Technologies (ex EEV)
Waterhouse Lane, Chelmsford, Essex, CM1 2QU, UK
Tel:      44 (0) 1245 -453458 (direct), -493493 (switchboard)
Fax:      44 (0) 1245 453224
http://www.marconitech.com/products/ccd/ccdprod.html

- -- CCD-world -- --
CCD-world is fully moderated. Send posts to CCD-world at astro.ku.dk
Standard replies will go to the list; address personal replies manually.
For more information, please go to:  http://www.not.iac.es/CCD-world/