Pyroelectric energy detector modeling?

[Chris Carlen]

Greetings:

Does anyone know how to model a pyroelectric energy detector? I mean
the type that is used to detect the pulse energy of lasers, where a
short deposit of energy on the surface of a pyroelectric crystal disc
creates a temperature change which produces an output signal.

Typically, there is a parallel RC network connected to the crystal, from
which the signal is taken:


+--------+--------+------>
| | |
| | |
----- \ -----
PPP / R ----- C
----- \ |
| | |
| | |
+--------+--------+------>

where "PPP" is the pyroelectric crystal sandwiched between two thin
metal plates, with one plate bonded to the grounded heatsink substrate,
and the other plate exposed, but coated with black paint.

I'm hoping I don't have to delve into thermal modeling here.

Are any of these guesses as to the nature of the device's energy to
electrical signal conversion process correct:

1. The device can be represented as a series RC network, where the
charge on the capacitance of the device is proportional to the energy
delivered? (simple)

2. The device can be represented as a series RC network, with rather
low and probably insignificant R, but the capacitance has a charge that
is related to the thermal gradient across the crystal? (not simple,
since it involves modeling the spatial heat distribution on one axis in
time, with the resulting charge being the integral of that).

3. Something else?


What I am ultimately trying to do is slow down a detector. They are
typically specified with maximum pulse durations, which are about one
tenth or less of the RC time constant of the extra network shown. This
ensures that the peak voltage of the response will be proportional to
the energy. However, if the pulse width of the energy source is too
slow, then the relationship between energy and peak voltage is no longer
linear.

We want to adapt a sensor to slower pulse widths, by making the time
constant of the unit longer, so longer pulses will appear as true
impulses relative to the new time constant. But we wish to understand
what is going on in more detail, since we may only be able to compute
the resulting change in calibration. We don't have a variable pulse
width energy source handy to recalibrate empirically.


Thanks for comments.

Good day!



--
____________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
[email protected]

 

[John Popelish]

Greetings:

Does anyone know how to model a pyroelectric energy detector? I mean
the type that is used to detect the pulse energy of lasers, where a
short deposit of energy on the surface of a pyroelectric crystal disc
creates a temperature change which produces an output signal.

Typically, there is a parallel RC network connected to the crystal, from
which the signal is taken:

+--------+--------+------>
| | |
| | |
----- \ -----
PPP / R ----- C
----- \ |
| | |
| | |
+--------+--------+------>

where "PPP" is the pyroelectric crystal sandwiched between two thin
metal plates, with one plate bonded to the grounded heatsink substrate,
and the other plate exposed, but coated with black paint.

I'm hoping I don't have to delve into thermal modeling here.

Are any of these guesses as to the nature of the device's energy to
electrical signal conversion process correct:

1. The device can be represented as a series RC network, where the
charge on the capacitance of the device is proportional to the energy
delivered? (simple)

2. The device can be represented as a series RC network, with rather
low and probably insignificant R, but the capacitance has a charge that
is related to the thermal gradient across the crystal? (not simple,
since it involves modeling the spatial heat distribution on one axis in
time, with the resulting charge being the integral of that).

3. Something else?

What I am ultimately trying to do is slow down a detector. They are
typically specified with maximum pulse durations, which are about one
tenth or less of the RC time constant of the extra network shown. This
ensures that the peak voltage of the response will be proportional to
the energy. However, if the pulse width of the energy source is too
slow, then the relationship between energy and peak voltage is no longer
linear.

We want to adapt a sensor to slower pulse widths, by making the time
constant of the unit longer, so longer pulses will appear as true
impulses relative to the new time constant. But we wish to understand
what is going on in more detail, since we may only be able to compute
the resulting change in calibration. We don't have a variable pulse
width energy source handy to recalibrate empirically.

Thanks for comments.

Good day!

Consider replacing the resistor with a cmos analog switch. The
capacitor will integrate the charge out of the detector, you take a
voltage reading, and then discharge the cap right before the next
test.

 

[Scott Stephens]

Greetings:

Does anyone know how to model a pyroelectric energy detector?

I found an interesting site yesterday with stuff on pvdf sensors:
http://www.ehag.ch/HTML-Files/Components/film_sensors.htm

But if the sensor is LiNbO, things can be more interesting =)
http://arxiv.org/ftp/physics/papers/0209/0209079.pdf
http://arxiv.org/ftp/physics/papers/0309/0309049.pdf
http://www.amptek.com/coolx.html

I mean

the type that is used to detect the pulse energy of lasers, where a
short deposit of energy on the surface of a pyroelectric crystal disc
creates a temperature change which produces an output signal.

I'm hoping I don't have to delve into thermal modeling here.

It's probably much worse then mere thermal modeling =)

3. Something else?

I think heat transports charge.

What I am ultimately trying to do is slow down a detector. They are
typically specified with maximum pulse durations, which are about one
tenth or less of the RC time constant of the extra network shown. This
ensures that the peak voltage of the response will be proportional to
the energy. However, if the pulse width of the energy source is too
slow, then the relationship between energy and peak voltage is no longer
linear.

The first article on x-ray generation mentions that charge can persist
for 30 hours, in a high vacuum, assuming no electons fly away.

Thanks for comments.

Tell me what you find, please =)

--
Scott

**********************************

DIY Piezo-Gyro, PCB Drill Bot & More Soon!

http://home.comcast.net/~scottxs/

**********************************

 

[Chris Carlen]

Consider replacing the resistor with a cmos analog switch. The
capacitor will integrate the charge out of the detector, you take a
voltage reading, and then discharge the cap right before the next
test.

That's a dandy idea! What remains to be seen is whether the total
charge remains linear with absorbed energy at longer pulse widths. I
suspect that at the pulse width approaches the thermal time constant of
about 400ms, then the thing stops responding. I hope it will still be
linear to within 1-2% with 4ms pulses. I will test it with an AOM
modulated argon laser.

Good day!




--
____________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
[email protected]