Quote:
Originally Posted by RickS
Now, if you could get a sensor with Sony ICX-694 QE and read noise and a 52mm diagonal then I've be an instant convert.
|
There are a few out there that are even better than the ICX-694, the deep depletion eXcelon (Princeton Instruments) being one of them. The Eagle V with the E2V chip being another. From memory, the diagonal on both of these cameras is 38mm.
The catch being the that the retail cost for the Eagle is ~$80K.
Not withstanding the extraordinary cost, I was curious to see if there was a case for cameras such as these amongst amateurs so I did some note pad calculations to that end. The results may surprise you...
First up, the quantum efficiency of three chips of interest:
Wavelength
__ eXcelon QE
__ ICX-694 QE
__ 16803 QE
___ (adjusted QE)
400nm_________82%________65%_______41%_________(36%)
450nm_________92%________72%_______45%_________(39%)
500nm_________90%________76%_______55%_________(48%)
550nm_________90%________76%_______60%_________(52%)
600nm_________90%________72%_______56%_________(49%)
650nm_________92%________66%_______46%_________(40%)
700nm_________93%________53%_______42%_________(37%)
I don't think too many people would argue that the CHART32 telescope represents a benchmark for amateur imaging instruments so I will use it for comparison. It is an 80cm cassegrain made by ASA with a camera using the 16803 chip.
Now, considering the relative QE of the choices we have above, what size telescope would you need to operate with these chips to achieve the same signal? fwiw) I added a fourth column (above) to reflect the slightly lower transmission of the Keller cassegrain as a function of its coatings. The assumption was enhanced coatings for the secondary (94%) and Al/SiO2 for the primary. The smaller theoretical telescopes with the eXcelon and Sony chips were assumed to have Max 'R'EAL from Spectrum Coatings, FL. (which wouldn't be an option for a 32" telescope)
Anyway here are the apertures derived for telescopes with equivalent signal as a function of wavelength and detector: The assumption is that the focal planes of each telescope are scaled to achieve the same angular resolution per pixel.
Wavelength
____eXcelon
____ICX-694
_____16803
_____________RC or DK
___Newtonian
___CHART32
400nm_________21.2"______24.1"________32"
450nm_________20.8"______23.8"________32"
500nm_________23.4"______25.7"________32"
550nm_________24.3"______26.8"________32"
600nm_________23.6"______26.7"________32"
650nm_________21.1"______25.2"________32"
700nm_________20.2"______27.0"________32"
It is probably fair to say that An eXcelon chipped camera on a 22" cassegrain or an ICX-694 on a 26" Newtonian would be functionally equivalent to the CHART32 with the 16803. When you include the mounting, housing and logistics of each choice, the 16803/ASA cassegrain combination comes in dead last if the metric by which we evaluate these systems is a function of signal per dollar. It's only advantage being the number of its pixels. It is also probably fair to say that the QE advantage of the eXcelon alone isn't enough to make it a viable choice over the ICX-694 in the context of instruments even the most ambitious amateurs are likely to deploy, with the assumption being that all three systems are operated such that they are shot noise limited.
But what about when camera noise defines the system S/N ratio?
Let's now compare the above chips in terms of exposure required at common emission line wavelengths to achieve an equivalent S/N assuming system noise is defined by manufactures spec for readout noise alone.
Relative exposure to achieve same S/N
CCD chip
_______R.O. Noise
____Hbeta_____Oiii_____Halpha_____Sii
16803 Proline
_____10e-1
_______57_______50_______81_______86
ICX-694 QHY
______5e-1
________7________7_______10_______11
ICX-694 Microline
__ 3e-1
_______2.5_______2.4______3.5_______4
eXcelon
__________2.3e-1
_______1________1________1_______ 1
Conclusion:
Assuming that you employ a filter with enough finesse to achieve the condition where the readout noise of the camera is dominant (A fabry perot interferometer [etalon] with appropriate blocking filter would be such an example) then the eXcelon chipped cameras (can) become the most cost effective option, depending on the size of the telescope.