CCD-world: LED with very stable light fux
Bruce Atwood
atwood at mps.ohio-state.edu
Fri Apr 21 10:42:02 CLT 2000
The following was posted to CCD-world:
LED's are very well behaved devices if they are treated with care. (HP
had and application note or article some years ago predicting 1%
stability over 100 years for their hermetic parts.) Over current or
static discharges can damage regions of the junction and permanently
change their characteristics. For typical devices (T1 Red for example)
operated at reasonable currents (10 ma) the -1%/degree C temperature
coefficient (light vs temp at constant current) lets you operate at 1
percent linearity (total flux vs exposure time) for exposures over a
range of 1 micro sec to 1 sec. I have used a constant current source
and a 201HS switch with good luck for 20 (ouch) years. If you want to
do better I would use the forward voltage of the device driven at
constant current to sense the temperature of the device (about -1
mv/degree C). A four wire circuit would be required to get a good
voltage reading with any reasonable cable length. From changes in the
forward voltage generate a small correction to the current to compensate
for changes in junction temperature. All in all a much easier task than
closing the optical loop and can clearly be made to operate faster and
with lower noise than an optical feed back system. If you must do the
optical feed back trick at least use a more or less matched pair and
close couple one of the sources to your detector to get a better signal
level to work with.
Tom Droege wrote:
>
> The following was posted to CCD-world:
>
> I have done some linearity tests on phototubes using a pulsed LED. I still
> have a graph on my Fermilab office wall showing 10E6 linear range. But it
> was done using neutral density filters, not good enough for this purpose.
> I was using the PM to test the electronics I built, not the other way
> round. I do not know the linearity possible, but other tests have
> indicated 1000-10000 to 1. I used an LED driven by an avalanche transistor
> and a short piece of cable. One can make 1-2 ns wide, very high amplitude
> pulses this way. Trigger the avalanche transistor so that it goes at the
> same amplitude. The cable makes sure the same energy is going into the LED
> on every pulse. The same thing can be done using the Tektronix mercury
> relay pulser (if you can find one). If the LED is driven with a 60 or 100
> volt, 2 nanosecond pulse then most of the things that vary are not changing
> much with respect to 100 volts. The LEDs work just fine. You can put 1-5
> amp pulses through them. Another trick is to temperature regulate the LED
> with a small TEC. Put the LED in a block of aluminum and regulate the
> block temperature.
>
> Of course feedback is a good idea. The feedback sensor described below by
> Roger could be used to regulate the HV supply to the cable. A few switches
> to gate the signal into the feedback loop and ... I could make a career
> out of doing this. ;^)
>
> Tom Droege
>
> At 06:43 PM 4/19/00 -0400, you wrote:
> >The following was posted to CCD-world:
> >
> >Hi Reinhold,
> >
> >I had a similar requirement when testing CCD linearity by measuring
> >signal as a function of exposure time with a pulsed LED. The
> >expereiment takes a long time if you can't generate fast stable
> >pulses. I was aiming for 0.05% accuracy of the integrated light in 1
> >ms or greater pulse.
> >
> >I used an integrated photodiode-preamp from Burr Brown (OPT211) to
> >servo control the LED intensity. The photodiode/preamp has much
> >better temperature stability than the LED and is almost free from self
> >heating effects. The LED is quite temprature sensitive, and thus
> >varies in intensity depending on how long it has been on.
> >
> >One has to deliberately add some overshoot at turn-on to compensate for
> >the fact that the turn off time of the LED is very much faster than the
> >turn-on time for the servo controlled LED. To acheve this circuit
> >allowed adjustment of initial LED current. To find the correct setting
> >one compares the output from 1 long pulse with the output in single
> >CCD readout that integrates N pulses each 1/N th of the duration.
> >
> >The circuit is described in "How Linear are typical CCDs?", page 251 of
> >the proceedings of the ESO workshop Optical Detectors For Astronomy,
> >1996. I bet you have a copy.
> >
> >Roger
> >
> >
> >
> >> To: CCD-world <CCD-world at astro.ku.dk>
> >> CC: "Dorn, Reinhold" <reinhold.dorn at eso.org>
> >> Subject: CCD-world: LED with very stable light fux
> >>
> >> The following was posted to CCD-world:
> >>
> >>
> >> Hello CCD-world,
> >>
> >> I am working on a new designed CCD for curvature wavefront
> >> sensing, which measures the light at a location between the
> >> the pupil plane and the image plane. In a real system, there
> >> is a membrane oscillating at a fixed frequency around 2 kHz and
> >> the CCD measures the photons collected during the intrafocal
> >> and extrafocal periods. From these signals we than calculate the
> >> normalized difference to be able reconstruct the wavefont tilts.
> >>
> >> To simulate the optical setup and to test the performance of
> >> the CCD independently of the optical bench, I am locking for
> >> very stable light source, preferred an LED. This LEDs should have
> >> stable rise and fall times in the range of 0 to 5 microseconds and
> >> a very stable flux for a certain amount of time (i.e. 240 microseconds).
> >> Note that we try to move just a couple of electrons with the CCD and
> >> therefore the LED flux should be extremely stable.
> >>
> >> The LED could then be driven by a clock pulse of the CCD controller
> >> with and accuracy of 25 ns.
> >>
> >>
> >> flux
> >> |
> >> | /----\ /----\
> >> |----/ \-- ----/ \------ time
> >>
> >> Does anyone know of a manufacturer or any source which is
> >> producing an LED with those requirements. Any help is welcome.
> >>
> >> Best wishes from Garching,
> >>
> >> Reinhold Dorn
> >>
> >> --------------------------------------------------------------------
> >> Reinhold Dorn e-mail: reinhold.dorn at eso.org
> >> ESO-European Southern Observatory phone : +49-89-32006-547
> >> K.Schwarzschildstr.2,D-85748 Garching fax : +49-89-3202362
> >> http://www.eso.org/odt http://pc002863.hq.eso.org
> >> --------------------------------------------------------------------
> >> - -- 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/
> >>
> >- -- 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/
> >
> - -- 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/
--
Bruce Atwood
Director
Imaging Sciences Laboratory
Astronomy Department
The Ohio State University
140 W. 18th Ave.
Columbus, Ohio 43210-1173
(614) 292 6279 (V)
(614) 292 2928 (F)
- -- 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/
More information about the CCD-world
mailing list