What mechanism causes Cds photoresistor (LDR) latency?

[sjb741]

Here it states

It takes usually about 10 ms for the resistance to drop completely when light is applied after total darkness, while it can take up to 1 second for the resistance to rise back to the starting value after the complete removal of light.

 

[D.A.(Tony)Stewart]

LDR's have extremely high sensitivity to changing resistance with light but also have high capacitance and very high resistance at low currents resulting a low dV/dt=I/C or large RC=T

They also have high tolerance errors unlike Photo Diodes (PD) which modulate current with light power ~ 0.5mA/mW and are very low capacitance , even lower with reverse bias. Inverting amplifiers make these practical and more accurate. Photo transistors are more sensitive but now introduce the higher 50% tolerance errors of hFE making them useful as switches but not linear sensors.

 

[dick_freebird]

Material is probably engineered for high carrier lifetime to maximize sensitivity. But the CdS material probably has to be passivated for long term stability and the interface between active material and overcoat tends to be a place where charge can be trapped (long or short term or a continuum of trap-depths hence lifetimes) and modulate the surface conductivity by local field.

 

[D.A.(Tony)Stewart]

Yes when you convert a magnetic material to disable rust into an inert oxide = insulation=dielectric, you are adding capacitance and resistance.

 

[sjb741]

Thanks - sounds like the trapped charge is the cause of the asymmetry. Is it roughly true then that trapped charge plays no part in conduction until the mobile charges are used up. Then they diffuse back into the bulk once the bulk carrier concentartion is sufficiently low?

 

[D.A.(Tony)Stewart]

- comparing the tolerance errors and latency of LDR's to an inexpensive PD , OA and 1% resistor gain, why would you want to consider LDR's?

 

[dick_freebird]

Thanks - sounds like the trapped charge is the cause of the asymmetry. Is it roughly true then that trapped charge plays no part in conduction until the mobile charges are used up. Then they diffuse back into the bulk once the bulk carrier concentartion is sufficiently low?

Well, it's just my theory.

You have a bulk lifetime that (I expect) is normal semiconductor physics. But at the interface and in close-in regions of the bulk-insulator is some population of traps which populate on the "hot" carriers kicked off by photons and de-populate by tunneling at a slower pace and there may be multiple trap types (depends on how many permutations of crystal and interface disorder there are, besides the desired ones).

 

[D.A.(Tony)Stewart]

There are probably a lot of particle physics explanations for LDR's higher sensitivity to latency, hysteresis, temperature changes, narrower spectral efficacy and capacitance increases with better tempco stable materials and also contains more toxic materials, although Gallium and Arsenic might top the list in gaseous state for fabrication of diodes. (One GaAs process engineer once told me a drop converted to a gas could wipe out a stadium.)

 

[timof]

Impurity photodetectors (photoresistors) have a lot of device physics effects that can make transient response slow and complex - not only carrier lifetime (i.e. generation-recombination), also - carrier transport through the "bulk" of the photoconductor, carrier injection from the injection contact, recharging of traps or other "volumes" (e.g. quantum wells, impurities, etc.) capturing and releasing the carriers, interplay between these effects, etc.

 

[c_mitra]

One GaAs process engineer once told me a drop converted to a gas could wipe out a stadium.

That must be a real bucket size drop- may cost you a fortune- or a nano stadium.

Arsenic is toxic but it has been also used as medicine. Ehrlich used arsenic to treat Syphilis.

 

[D.A.(Tony)Stewart]

It was 1 of 500 different organic arsenic compounds that Ehrlich found, but others had found other double-edge sword compounds, giving it the name king of poisons and poison of kings. (and queens). He was referring to the pure gas measured in [ppb] in semi-foundries, rather than oxides or other compounds,

 

[crutschow]

why would you want to consider LDR's?

If you want a resistance that varies with light, which photo diodes cannot do.
Photo diodes and photo transistors vary their current, not their resistance (which is quite high).

 

[D.A.(Tony)Stewart]

If you want a resistance that varies with light, which photo diodes cannot do.
Photo diodes and photo transistors vary their current, not their resistance (which is quite high).

The same effect occurs in common emitters used as active loads. " Current controlled Resistance with a constant reverse Vbc voltage" Effectively identical to PD's except controlled by light current and similar to LDR's. So "cannot do" is false.

 

[D.A.(Tony)Stewart]

When you bias an LDR with fixed resistance it changes inversely with footcandles below ft-c.

When you reverse bias a photodiode pulled high and measure current by using a large resistance to detect voltage. One could also plot the PD as the V/I=R changes from the current or voltage ratio with a fixed R reference.

REF https://components101.com/sites/default/files/component_datasheet/LDR Datasheet.pdf

Opinion
-

LDRs or photoresistors have a long response time. They may take several seconds to change conductivity after exposure to light.
Photodiodes, on the other hand, have an almost instant response.
Although an LDR is tuned to the visible spectrum of light, photodiodes are sensitive to both visible and infra-red lights.

The differences are due to chemistry and size of the array.

The similarities with all semiconductors is that the capacitance inherent to all dielectrics is greater with the amount of conductance. Normally diode capacitance is rated at 0V and declines with inverse voltage such the bulk resistance as a saturated diode , Rs or transistor or FET and capacitance product RC=T like f=1/2T is a constant of bandwidth for a family of similar products and construction.

- So as the LDR is a much larger junction it also is capable of lower resistance but inherently higher capacitance which affects the turn off time when R is high. Photodiodes have similar properties of slower turn-off than turn-on times but many orders of magnitude faster but also lower currents.

The biggest advantage of PD's is very low error tolerance of mA/mW of optical power often in the 500 uA/uW of optical power or 0.5mA/mW. unlike photo-transistors, PT which vary in sensitivity over 300% also just like opto-couplers, due to the variances of hFE from process, temperature and saturation and far less due to LED variances.

The biggest advantage of LDR's is the simplicity of use with a wide dynamic range of sensitivity for ON to OFF control or dimming. This could be expressed as mA/mW but usually shown as a linear inverse plot on a log-log scale of R vs lux of optical power. But you could never expect all to have the same sensitivity threshold for resistance vs optical power yet they do provide over 3 decades of mA/mW with more gain from size.

Does anyone agree? disagree?

 

[c_mitra]

He was referring to the pure gas measured in [ppb] in semi-foundries, rather than oxides or other compounds,

The safe limits are described in https://www.cdc.gov/niosh/idlh/7440382.html.

Just quoting from the same page:

Revised IDLH: 5 mg As/m3

Basis for revised IDLH: The revised IDLH for inorganic arsenic compounds is 5 mg As/m3 based on acute inhalation toxicity data in animals [Flury 1921; Spector 1955]. This may be a conservative value due to the lack of relevant acute toxicity data for workers. [Note: NIOSH recommends as part of its carcinogen policy that the "most protective" respirators be worn for inorganic arsenic compounds at concentrations above 0.002 mg As/m3. OSHA currently requires in 29 CFR 1919.1018 that workers be provided with and required to wear and use the "most protective" respirators in concentrations exceeding 20 mg As/m3 (i.e., 2,000 x the PEL).]

I believe the oxide to be the most potent poison; it is also absorbed via the skin. It is cumulative like Pb and Hg (excreted poorly).

There is enough evidence of Arsenic as a carcinogen at low concentration but that may take years.

 

[dick_freebird]

In the fab you would likely be exposed to arsine gas
if anything. But arsine is pyrophoric so would become
atomized arsenic oxide on contact with atmosphere.

All of the process gasses are nasty to some degree,
most of them highly flammable (SiH4, PH5, AsH3, BH3)
and many self-igniting.

Arsine - Wikipedia

en.wikipedia.org

Arsenic concentrations in silicon ICs are very low, shot
mostly at underlayers (where low thermal diffusion
lets dopant stay put through process thermal cycles; if
you didn't care about that, you'd have used phosphorus)
and encapsulated by overglass. Usually ~ 1 - 100ppm
and tough to get at anything past the surface unless
you are consuming the silicon, atomically fine.

Now take an acetylene torch to a GaAs RFIC, you'd
get a whole 'nother level of stank.

 

[c_mitra]

In the fab you would likely be exposed to arsine gas
if anything. But arsine is pyrophoric so would become
atomized arsenic oxide on contact with atmosphere.

All of the process gasses are nasty to some degree,
most of them highly flammable (SiH4, PH5, AsH3, BH3)
and many self-igniting.

Arsine - Wikipedia

en.wikipedia.org

Arsenic concentrations in silicon ICs are very low, shot
mostly at underlayers (where low thermal diffusion
lets dopant stay put through process thermal cycles; if
you didn't care about that, you'd have used phosphorus)
and encapsulated by overglass. Usually ~ 1 - 100ppm
and tough to get at anything past the surface unless
you are consuming the silicon, atomically fine.

Now take an acetylene torch to a GaAs RFIC, you'd
get a whole 'nother level of stank.

You are right that both PH3 and AsH3 stink like hell. But nothing can beat the arsenic compound called cacodyl oxide (https://en.wikipedia.org/wiki/Cacodyl_oxide) in stink!

A very similar sulfur compound (CH3)2.CS (thioacetone) is reported to be the stinkiest of all. They bind to the smell receptors so well that you will continue to smell them for hours after you are exposed.

However, the oxide does not smell; it has no disagreeable taste either (choice item for poison). Arsenic is widely distributed and is present in water in trace quantities in may areas.

If you burn a GaAs based IC in a C2H2 torch, I do not expect smell any stinking by-product (you will get oxides of gallium and arsenic).

 

[D.A.(Tony)Stewart]

The lowest country safety limit is Japan. This might correlate with my "3rd hand info" on the toxicity by an engineer who installed a GaAs foundry for Kopin's surface-sliced AM-LCD chips on the east coast with gas leak detectors for fuming GaAs furnaces used for vapor deposition.

Arsine :

Japan

Occupational exposure limit 0.01 ppm (0.032 mg/m3)

Now did anyone wish to comment on my correlation of LDR's and PD's? or transistor common emitter "active load" circuit model being a base-current-controlled conductor (Y=1/Rce)

Contrasting specs:

The equivalent circuit of a common emitter (CE) current sink being an hFE* Ib controlled Ic = Vce/Rce
where Rce = Vce / ΔIc . This term is commonly used by Diodes Inc to correlate Vce(sat)/Ic of their low Rce transistors to low RdsOn FETs and is similar to a photo diode being ΔI/ mW of optical power current correlating to Light Controlled Resistors (LDR's) but differing greatly in size.