Status, September 13th, 2005

Signal Sampling Settings proposed for DBS Red (Dewar 7)

Cicada	CDS Gain	Signal	System Gain	Noise(e)	Read Out	Comment
GAIN		Sample(us)	(e/adu)		Time(s)
1	4.75	2	0.54	4	~7	Hi-gain, No benefit
2*	2	2	1.24	3.85	~7	Default for DBSR
3	1	2	2.54	5.5	~7	Lo-Gain, for Full Well
2	2	3	0.82	3.6	~9.5	Useful
2#	2	4	0.61	3	~12	Hi-gain, lowest noise

Note (1) - The timinig file (# = fast-hig.lod) for the lowest noise option is a different
file to that for standard operation (* = fast.lod).

For convenience - medm.lod == fast-hig.lod
This allows the observer to load this file from Cicada

(2) - although the settings are the same for DBS Red as DBS Blue, the system gain
1.24e/adu is different due to a slight change in the voltage applied to the on-chip
FET amplifiers's Output Drain(VOD). On DBS Blue this was 22V and on DBS Red this is 21V.

This was modified due to the presence of a small amount of output glow seen in dark
frames and believed to be associated with the output amplifier. Reducing the OD voltage
by 1 volt removed this but changed the output FET gain slightly.

This will have no impact in the operation of the CCD.

The image below shows a 2000 second dark frame with the amplifier glow evident at
the top left hand (read-out) corner. Note this CCD has only 1 output amplifier.
There is also 2000 second's worth of cosmic ray events as well, ~180e events.

This is about correct for this device (accepted rate is 1-2 events/cm2/minute).

2000 second dark frame illustrating amplifier glow.

DBSR will be transported to Siding Spring for observational use on
October 6th, utilising 2us sampling, Gain 1.24e/adu,
R/O noise 3.85e, R/O time ~7seconds.

Pixel full well maximum is 100ke/pixel.
The gain setting used here will thus reach
80ke/pixel

Status, August 24th, 2005

Read-noise is 3.5e and the settings in all other ways will be the same as
for the DBSB head...

Spectral Response for DBSR, E2V 19-1 CCD.

The artefact mentioned in earlier reports - a small ring seen on the CCD surface, has
been investigated by E2V and the CCD cleaned. Although this has reduced the intensity
of the appearance of this ring, it can still be seen at short wavelengths as a disk of
somewhat lower responsivity (~8%) and as a ring at longer wavlelengths. The following
pictures were taken on the Test system here, and show flatish field exposures
at 350 and 700nm

Flat field at 350nm showing
disk of lower response for DBSR CCD.

Flat field at 700nm showing partial
ring structure for DBSR CCD.

The details regarding the appearance of this artefact were described by E2V -
see the status report below, relayed to me by E2V.

This basically closes out this matter and the head will be delivered in its present state.

The old DBSR head will be returned shortly after delivery of the new DBSR head.
The current plan is to press this into use as a 2nd test dewar for WiFeS CCD
quantification. This will require some modification to the head at the front.

This will then complete the SII refit of all the old cameras with newer, faster
lower noise E2V detectors.

Status, August 4th, 2005

The DBS Red CCD has just been delivered to RSAA and is now
ready for installation into the DBS camera.

I will then be able to confirm the operational status of the
device after the cleaning process at E2V. As it has been operated
by them in their imaging camera test system, I am not expecting
any problems.

I plan to install this in the DBS dewar next week and start the
characterisation process. As this device is identical to the
DBS Blue CCD - I do not see any problem in attempting to get this
to SSO relatively quickly.

E.T.A. for observing is now estimated to be by the 2nd or
3rd week in Sept.

Status, August 4th, 2005

DBS CCD On its way back.

E2V have now cleaned the CCD and the effect of the artefact seen in the
images below has been significantly reduced.

E2V CCD Status report

This from Paul Jorden at E2V..
'...We have now re-tested the 42-10. This is the basic story:

I think the chip always had various minor 'etch' marks, which
are usual with these types of back-thinned chips. Such marks can
be visible under bright light & magnification. They can cause minor
modulation of the response, but still within the PRNU spec.

The faint depression of response seen as a circular patch (centred on
~850, 270) is still present in the 400nm test image. However the cleaning
seems to have made it less visible by eye, and the circular ring at 650nm is
hardly detectable now...'

If this device arrives by the end of next week - I would hope to get it
in the DBS Red camera and to Siding Spring by the end of August...

Status, July 25th, 2005

'...We have now re-tested the 42-10. This is the basic story.

I think the chip always had various minor 'etch' marks, which are usual
with these types of back-thinned devices. Such marks can be visible
under bright light & magnification. They can cause minor modulation of
the response, but still within the PRNU specification.

The faint depression of response seen as a circular patch (centred on
~850, 270) is still present in the 400nm test image. However the cleaning
seems to have made it less visible by eye, and the circular ring at 650nm is
hardly detectable now.

We are therefore arranging to return it promptly, after completing some
internal paperwork...', Paul Jorden, E2V

As I mentioned last week - as soon as this device re-appears at RSAA
I will install it in the science camera - test it and get it to Siding Spring
as soon as possible after that.
I am aiming for mid-August, based on the CCD arriving this week (25/7/05).

Status, July 20th, 2005

I would then hope to have this device in the DBS Red camera and
up at Siding Spring within 2-3 weeks.

Status, July 18th, 2005

In a conversation with Paul Jorden (at E2V) whilst I was over in the UK a week or
so ago, he confirmed that the strange ring may be either a manufacturing fault
caused during the backside thinning process or a mark on the surface.

They apparently have their own data showing the artefact before delivery to RSAA.

Although the artefact does not put the detector out-of-spec. for science use
Paul has suggested after further investigation they will probably attempt cleaning
and if this isn't successful will replace the device with a new CCD of the same type.

This should mean the detector will arrive back here before the end of July
and hopefully we can then get the camera quantified and up to Siding Spring
as soon as possible after that.

Status, June 16th, 2005

A decision will then be taken as to what course of action is appropriate.

There are 2 possibilities-

• The detector has a manufacturing fault and E2V will replace it
• The ring artefact is not a detector fault but has arisen
  since it left E2V

The delivery of this head to Siding Spring now looks likely not to happen
until at least early August

Status, June 2nd, 2005

Strange Ring artefact is in the CCD AR coating

Close up image taken of detector surface

This looks to me like a fault in the AR coating or some other chip related problem
if this proves to be the situation we ought to attempt to get a new device from E2V.

The observing community can decide whether to proceed with this option of exchange
or not and live with what we have....

Status, June 1st, 2005

Strange Ring artefact seen in Images with Science Detector

The 4 images below illustrate the artefact. Two (images #1 & #4) were taken using
the internal preflash (red) LED light sources, 1 (#3) is an exposure taken on the
test box with a 1cm aperture using a flat-field source of white LED light, taken on
the 27th May and the final exposure (#2) is a 60 second dark frame also taken on the
27th May (last Friday).

The second of the pre-flash (red) LED exposure (#4) was taken yesterday after
cleaning the field-flattener which is situated right on the chip package.

Strange!

The camera will be opened tomorrow and the source of the strange ring investigated.

Indications at the moment favour some mark or defect on the Detector itself!

Pre Flash Image taken on 27th May

1. DBS Preflash FITS data.

60 sec. dark exposure taken on 27th May

Some evidence can be seen that the ring artefact also exisits in this dark exposure.

2. DBS Dark frame FITS data.

5 sec. White LED Flat-field exposure taken on 27th May

3. White LED Flat Field FITS data.

Preflash exposure taken yesterday after cleaning field flatener.

4. DBS Post Clean Preflash FITS data.

If it proves to be a CCD fault, the detector will have to be returned to E2V
for inspection and possible replacement.

This course of action will be investigated after tomorrow's inspection of the
optics and detector.

Status, May 9th, 2005

Science Detector now installed in DBSR

The earlier vacuum problems were solved on Friday after using the Alcatel
leak detector - being demo'd at R.S.A.A. last Thursday to find a very small
end-cap vacuum leak. The image above shows the RT read-out just obtained.
DBSR will be pumped and cooled tomorrow and testing started after WFI leaves
for Siding Spring - early next week.

The DBS Red CCD will now be characterised and as stated below, it is
hoped that delivery to Siding Spring can take place at the beginning of June.

Assembly Of Optical Components and Science CCD in DBS-R

2. Another view, similar to above, showing Field Flattener.

3. Close up view from side showing calibration LEDs.

4. An even closer view of CCD and Field Flattener.

5. View of internal wiring and Temp. components.

6. General overall view of dewar and internal parts.

Status, May 6th, 2005

The head has now been pumped and the Engineering CCD is currently undergoing a
cold test. All indications are that the head is now behaving properly. Therefore
the Science CCD will be installed on Monday during spare time in the final
WFI CCD ouput amplifier change over.

The detector will then be characterised and it is hoped will be ready for
delivery to Siding Spring by the end of this month (May, 2005).

Status, April 28th, 2005

This is being investigated today. The Engineering CCD has been operated at
a lower temperature and all appears ok. The above problem, when sorted out,
will mean a further vacuum and cold test with the Enginnering CCD before
the science array is installed.

This has therefore put back the delivery to Siding Spring - to mid-May at
the earliest.

Status, April 18th, 2005

Test Image from Engineering grade DBS CCD

mounted in newly re-fitted (now) DBSR dewar

The image shows the same features as illustrated in the image taken in the DBS
Blue camera. See the DBS-B pages for a comparison image. As you will see the pixel
defect trailed into a column at Room Temperature is reduced to a cluster of pixel
defects when the CCD is cold, The Enginering DBS-B image showing this at the bottom
right-hand side of the image. The X, 50-pixel, under and overscan (Bias strip),
regions can also be seen in this image.

Status, March 30th, 2005

After this new red camera has been commisioned at SSO, in the 2nd week of May,
the old DBS Red camera will be returned and its operational use re-directed,
possibly for use on the Echelle, utilising the Engineering grade E2V 2kx2k CCD.
This device has more than adequate performance for this use.

with a new detector from E2V - a 0.5kx2k , thinned imager, anti-reflection coated and optimised in the Red.

Documentation Available

Te one we actually use in DBS is a Non Inverted Mode device

E2V 4210 NIMO CCD for DBS.