Projets

Association

Dossiers tech.

Connexion

Rechercher Web Pages

The Artemis CCD Camera

The Artemis CCD Camera was born from the hands of Steve Chamblers, Jon Grove and Arthur Edwards.

All informations about the camera are available from the Artemis website : Artemis CCD

It principal features are:

• Sony ExView multi CCD support up to 1.4Mpixels
• 16 bit ADC
• Peltier Cooled
• Extremely low dark current
• USB 1.1 Transfer mode
• High quality machined aluminium body
• Very high sensitivity
• Embended control port for modified webcams (no parallel port needed)
• High quality and very usefull software package
• Very good quality/price ratio

Please note that the Artemis is also sold by Atik under the Atik 16 and Atik 16HS denominations.


Global aspect and body
The body of the Artemis is made of a high quality aluminium machined body:



No translation in english available yet for this section

Connectors and Kodak screw hole

No translation in english available yet for this section

Assembling the Artemis Camera

It is quite easy for anyone who have good experience of electronics assembling to build up the Camera, thanks to a very well done and detailed documentation provided by Artemis and available from their website.

The only real point of difficulty resides and the CCD sensor mounting : it is necessary to adjust the lenght of the CCD support and solder it at a very precise position.

The position of CCD sensor MUST be exactly parallel to the plan of the camera body, knowing that the peltier module is mounted under the CCD sensor and needs thermal paste. If the CCD sensor is not correctly positionned the dramatic consequences will be :

• Bad cooling of the sensor by the Peltier
• Focal plan and CCD sensor plan will not be the same -> blur pictures

So this is the crucial point of the Artemis assembling!!!



Different CCD Sensors
The Artemis camera support the following Sony Exview sensors:

Models with C means that the sensor features a Bayer matrix, allowing "One shot Color", but implies loss of light.

• ART-254 - 1/3" - 510x492 - 9,6x7,5µ
• ART-254C - 1/3" - 510x492 - 9,6x7,5µ
• ART-255 - 1/3" - 500x582 - 9,8x6,3µ
• ART-255C - 1/3" - 500x582 - 9,8x6,3µ
• ART-424 - 1/3" - 659x494 - 7,4µ
• ART-424C - 1/3" - 659x494 - 7,4µ
• ART-429 - 1/2" - 752x582 - 8,6x8,3µ
• ART-429C - 1/2" - 752x582 - 8,6x8,3µ
• ART-285 - 2/3" - 1392x1040 - 6,45µ
• ART-285C - 2/3" - 1392x1040 - 6,45µ

One great and unique feature of the Artemis is the abilitiy to support any of these sensors only by updating his firmware, and placing the new sensor on the PCB.

This allow to first buy a ART-254 (very cheap) and to switch to a ART-285 (most expensive) when you win a million $ at the loto...

You can find the datasheets of all these sensors in their "AL" version (Black and white) in the Artemis Docs. menu of the website.

Sony CCD Sensors size

In my own case, I choosed the ICX285AL which is the biggest black & white sensor available at this time for the Artemis.

This sensor, as all the Sony sensors availables have the particularity to be quite small, with tiny photosites (the electronic element that "see" a pixel) of 6.45µm*6.45µm. This implies smaller fields than cameras with Kodak KAF Sensors as the size of the active part of the Sony chip is only 2/3 inch, and it is the biggest one in the Artemis available range.

Sony CCD Sensors sensitivity

The Sony Exview SuperHAD sensors seems to have pretty good sensitivity and have been used to build famous CCD Cameras like the Starlight SXV series previously to the Artemis.

Pixels Size

The relative small size of the pixels have now proved trough the experience of existing CCD Cameras like the Starlight SXV-H9 (ICX285 Sensor, just like the Artemis) and now the Artemis series that they have nothing less than larger sensors in terms of sensitivity. The technologies used by Sony in their development are completely up to date (thanks to the Digital cameras mass market?) and their sensors have very good sensitivity in wavelenghts that were not really covered by competitors like the blue part of the spectrum.

SuperHAD

Their sensitivity in the infrared range is also very good as these sensors are designed for low-light/night security cameras.

The SuperHAD (Super Hole Accumulation Diode) Design allow to optimize the sensitive surface by using extended microlenses to collect the light even in the usual "dead zones" of a CCD sensor. More information about this design on Sony's website: Sony Super HAD

These sensors use an antiblooming technology that prevent streaks on the final picture caused by "electrons overflow" on pictures where too much light has been collected.

Antiblooming feature is achieved by a particular structure of the CCD Gate (Photosite element) that implements a "drain" to collect overflow electrons and avoid them to run and fill the others photosites.

This implies a loss of sensitive area of about 25-30%, and this is partially compensated by the use of microlenses that will collect the light even in these "dead zones".

Antiblooming allow sensors to evacuate 100x to 1000x their maximum electron capacity (well depth) before starting to overflow.

You can find interesting elements about antiblooming and CCD on the Apogees Website : Apogee Website

Quantum efficiency

Quantum efficiency is the ability to produce electrons when photons hits the sensor.

Despite their relative pixel's small size and the use of antiblooming gates, you can see on the graph below that Sony sensors and ICX285 in particular have a very good quantum efficiency compared to common actual sensors, probably because of their more recent development.

This fact is proved by the reduced acquisition times I could see using my Artemis camera.

Well Depth

Interesting information about well depth can be found on the Apogee's website: Apogee - Well depth

The estimated well depth of each sensor is described in the technical details of each camera on Artemis's website : Artemis CCD


A small Well Depth

The well depth criteria is a capacity that caracterize the amount of electrons each Gate (photosite element) is able to collect before being in a state of overflow.

For the ICX285, this criteria is estimed to 32000 electrons. This is quite little keeping in mind that a common Kodak sensor have a well depth of 85000 electrons.

Why this difference? This is explained mainly by two things :

• Well Depth is determined mainly by the sensitive area and the CCD gate's structure
• Size of photosites : ICX285's photosites are small : 6.45µ->41µm^2, whereas common kodak sensors have at less 9µ->81µm^2, that is to say a gain of 100% of sensitive area!
• The Kodak sensors used in common astrophotography applications don't have antiblooming gates, so this save an other 25% of sensitive area

Consequences of a reduced well depth

There is absolutely no impact in regard of the quantum efficiency (the ability to produce electrons when photons hits the sensor), in opposite of the reduced sensitive area.

In fact the main consequence is a problem of saturation and dynamics of the picture. For example if you want to shoot M42, the great Orion Nebula, you will sature quite quickly the part of the sensor which "see" the bright heart of the nebula, but there won't be blooming because of the sensor's structure. In an other hand you will need a long pose shoot to reveal the faint structures of the nebula.

In such bright AND faint objects it will be necessary to work with multiple shoots and add them later.

In an other hand, if you use for example 2x2 binning, you will combine 4 pixels in one "virtual" one. The resulting pixel will be 4 times more sensitive, with a well depth multiplicated by 4, that is to say 32000*4=128000 electrons.

Conclusion : in the "real life" this fact is not really representative of the usage of the ICX285. It allows up to 60 minutes and much more acquisition time, saturation is as with all CCD devices a question of acquisition time to choose acurately. In faint with very bright areas objects with M42, the problem with be the same with a big part of the CCD Astro Cameras available today.

Analog to digital conversion

The Analog to Digital Conversion (ADC) is accomplished by a 16bit ADS8322 element from Texas Instruments.

His work on the CCD Camera is crucial : it reads the number of electrons capture by each photosite, ie the number of photons captured by the sensor, and communicate the result to the Artemis main processor.

According to the features of the sensor there is absolutely no need to use a converter of more than 16 bits: the estimated well depth is 32000 electrons, and a 16 bit Analog to Digital converter is able to count 65535 different values.

In the Artemis, the work ADC Converter is to "read" the amount of electrons corresponding to a virtual number of photons, ie the brightness of each photosite of the sensor, which will result in a pixel value.

The resulting FIT file of an acquisition with the Artemis CCD Camera (or any other one) is no more than a gigantic table of 1360*1024 cells containing each one a value from 0 (absolute black) to 65535 (absolute white).

In real life, the value range would be more 100-47000 due to precharge and other complex things in the camera, but this is the way things are done in the heart of the artemis.

Dark Signal and noise

I have been really surprised by the very low amount of dark signal produced by my Artemis 285.

The dark signal is determined by two main factors:

• Sensor design
• Temperature, as temperature is energy and this make electrons to move -> Thermal noise

The Artemis features one single peltier module. It carries out about 20 degrees (celcius) off the CCD sensor, making it to be about 20 degrees below ambient temperature. This component consumes a lot of energy (5V, 2.5Amp).

When common CCD Cameras use two or three peltier modules (to obtain 20 to 60 degrees below ambient temperature), and implies the use of huge power supplies, the Artemis only feature one.

The reason is very simple : the dark signal of the ICX285 appears to be very low : 300 adu/47000 on a 5min shoot.

After substracting a reference dark to a dark shoot, the remaining noise is of 100 adu/47000 on a 5min shoot : the ICX285 have a really low noise level!

I didn't notice any bright satured pixels on my sensor.

You can find a sample dark on the "Artemis Docs" section of the website, but be carefull: some pixels appears to be bright but are not "dead pixels" : they only have +200adu/47000 than the others, this is completely unnoticeable in real conditions.

Software Support

Today, the camera come with the following pieces of sofware:

• Artemis windows drivers
• Artemis Astroart/Maxim DL drivers
• Artemis Capture and Diagnostic
• Artemis SDK for those who wants to develop their own software

I didn't try the Artemis with Maxim, but it is a real pleasure with AstroArt.

Artemis Capture is also a very friendly and efficient software. It is frightening in helping to find the focus or making image adjustments.

Artemis accessories

No translation in english available yet for this section

Conclusion

The Artemis Camera is in my point of a view one of the best CCD Camera available today for it's price regarding it overall performances.

It remains in my mind a powerfull, very sensitive and reliable toy that I will use as often as possible.

Artemis or Atik?

As the two cameras are exactly the same, except for the color, I think that if you feel to achieve the assembling of the camera then choose the Artemis, if not let's go for an Atik!

An online tour of the possibilities of the camera can be seen on the Artemis CCD galery: