In a message dated 27/05/01, mlamb1@......... writes:

> Is there really....any technical advantage to using two side by
> side photocells (misc variety in a differential bridge), as opposed
> to using just one alone?  I'd think one cell would be simpler
> mechanically but I'm not sure if it would react the same
> circuit wise or as well (?) as two cells.
> 
> Its kind of relevant, as I'am on a large bright LED and single
> CDS (photoresistor) sensor project at the moment...which could
> be changed over time. 
> 
Dear Meredith,
       
       You might like to consider some of the following design ideas:-

       Precision Si photocells have closely matching characteristics in terms 
of sensitivity, wavelength response, change of sensitivity with illumination, 
change of sensitivity with temperature and reverse leakage current. The 
response is fast, free of drift and the photo current is linear with 
illumination. It is not possible to buy CdS cells matched to anything like 
this precision. The response is very much slower and may show drift over some 
seconds. The CdS response is very non linear with changes in illumination 
level and any two cells may be expected to have significantly different 
response curves. They are also quite temperature sensitive.

       If you use a single photo cell, it's output level will change with 
changes in illumination, temperature etc. as well as with any movements of 
the slit. If you use two matched cells in a bridge, the centre zero position 
is effectively constant since the signals are subtracted and only the offset 
(signal) scaling is effected. This is a great advantage for instruments like 
seismometers which have a central balance point. Seismometers need a sensor 
with both very high sensitivity and very high stability. 

       The light output of a LED varies quite strongly with it's temperature. 
The light has a considerably higher intrinsic noise than a filament lamp. 
LEDs are 'fast' devices and will faithfully copy any electrical noise from 
the power supply into the emitted light. The light output does change with 
age. Providing a good heat sink to reduce temperature changes only helps a 
bit. 

       If you run a filament lamp from a stabilised power source of 2/3 to 
3/4 it's rated voltage, the filament life becomes virtually infinite. The 
light intensity is high and the sensitivity to ambient temperature changes is 
small. The slightly yellower light does not matter, since a Si cell has a 
high sensitivity into the near Infra Red. To vary the light output you have 
to change the filament temperature and the 'thermal inertia' tends to damp 
rapid fluctuations such as noise in the power supply. You may need to take 
precautions to cool the glass envelope and the metal glass seals and to avoid 
the heat causing convective air currents.

       Regards,

       Chris Chapman
In a message dated 27/05/01, mlamb1@......... writes:



Is the
re really....any technical advantage to using two side by

side photocells (misc variety in a differential bridge), as opposed

to using just one alone?  I'd think one cell would be simpler

mechanically but I'm not sure if it would react the same

circuit wise or as well (?) as two cells.



Its kind of relevant, as I'am on a large bright LED and single

CDS (photoresistor) sensor project at the moment...which could

be changed over time. 



Dear Meredith,

       

       You might like to consider some of the following design ideas:-



       Precision Si photocells have closely matching characteristics in terms 

of sensitivity, wavelength response, change of sensitivity with illumination, 

change of sensitivity with temperature and reverse leakage current. The 

response is fast, free of drift and the photo current is linear with 

illumination. It is not possible to buy CdS cells matched to anything like 

this precision. The response is very much slower and may show drift over some 

seconds. The CdS response is very non linear with changes in illumination 

level and any two cells may be expected to have significantly different 

response curves. They are also quite temperature sensitive.



       If you use a single photo cell, it's output level will change with 

changes in illumination, temperature etc. as well as with any movements of 

the slit. If you use two matched cells in a bridge, the centre zero position 

is effectively constant since the signals are subtracted and only the offset 

(signal) scaling is effected. This is a great advantage for instruments like 

seismometers which have a central balance point. Seismometers need a sensor 

with both very high sensitivity and very high stability. 



       The light output of a LED varies quite strongly with it's temperature. 

The light has a considerably higher intrinsic noise than a filament lamp. 

LEDs are 'fast' devices and will faithfully copy any electrical noise from 

the power supply into the emitted light. The light output does change with 

age. Providing a good heat sink to reduce temperature changes only helps a 

bit. 



       If you run a filament lamp from a stabilised power source of 2/3 to 

3/4 it's rated voltage, the filament life becomes virtually infinite. The 

light intensity is high and the sensitivity to ambient temperature changes is 

small. The slightly yellower light does not matter, since a Si cell has a 

high sensitivity into the near Infra Red. To vary the light output you have 

to change the filament temperature and the 'thermal inertia' tends to damp 

rapid fluctuations such as noise in the power supply. You may need to take 

precautions to cool the glass envelope and the metal glass seals and to avoid 

the heat causing convective air currents.



       Regards,



       Chris Chapman