Exmoor Sensor and SWIR

[Avalon]

When reading about Sony Exmor sensors I stumbled upon it's 4th generation Exmor sensors claiming to have enhanced infrared sensitivity because of deeper pixel well depth. In fact there are for sale infrared CMOS camera's that can detect in range of 400-1700nm although SWIR sensiivity is low but it's not at least some exotic technology.

[Andy Perrin]

Link? (Also, I’m reserving judgement on whether that can be considered exotic!)

 

I just spent ten minutes searching for this tech and can’t find it. All the EXMOR and EXMOR R literature shows the quantum efficiency dropping like a rock at 1100nm, just like every other CMOS that doesn’t have another material on top (like the germanium in the TriWave, or quantum dots).

[Avalon]

Link? (Also, I’m reserving judgement on whether that can be considered exotic!)

 

I just spent ten minutes searching for this tech and can’t find it. All the EXMOR and EXMOR R literature shows the quantum efficiency dropping like a rock at 1100nm, just like every other CMOS that doesn’t have another material on top (like the germanium in the TriWave, or quantum dots).

 

 

 

 

Here is the quote: " With the first generations of Exmor technology addressing the pixel/sensor noise issue, it only seemed natural that the fourth generation Exmor technology addresses the NIR sensitivity by introducing deeper pixel wells to help capture these longer wavelengths.“

https://www.framos.com/en/news/what-is-sony-s-exmor-technology-anyway"

I see no mention of SWIR spectrum only that sensors have higher quantum efficiency (QE) curves in the NIR range between 800-1200 nm wavelength, sorry if I mistaken you.

However below is link of CMOS based SWIR camera which can capture up to 1700nm although sensitivity curve goes very low. As I understand deeper wells allow to capture more energy because longer waves can penetrate deeper into photodetector material.

 

http://www.ir-viewers.com/product/contour-ir-digital-cmos-camera/

 

There are amazing promises of CMOS sensors using graphene instead of silicon which would have very high and broadband spectrum sensitivity from visible to terahertz spectrum! Interesting quote explaining why: “In graphene, there is no energy gap for converting valence electrons into free electrons. Therefore, in principle, it can absorb light with any wavelength to turn valence electrons into free electrons.“

They would also have better performance, be cheaper and able to operate in room temperature. But we will have to wait for such technology arrive to the market.

 

https://www.graphene-info.com/mitsubishi-electric-developing-graphene-based-super-wideband-image-sensor

 

Meanwhile I had thoughts about modifying existing technology to detect infrared radiation but I do not have complete technical understanding and might be falling into wishful thinking so correct me if I‘m wrong.

 

Image intensifier tubes such as Gen3 used in night vision devices use gallium arsenide photocathodes to capture infrared spectrum down to 930nm what is enough for applications usual in military or hobbyist night vision. But there are various photodetector materials for different spectrums and intensifier tube offers advantage of amplified sensitivity. For example specially manufactured InGaAs photocathodes can operate up to 1700nm (not sure what is manufacturing technique, thicker coating?). Photocathode screen can be carefully cut out and replaced with another able to convert photons into electrons, no other component of tube requires modification. Vacuum pump will be required to suck out air from tube, as well magnetron sputtering (maybe) to coat photocathode window with desired material.

 

Another question is about microbolometers used in thermal camera‘s. Is sensor spectrum response limited to 7.5–14 μm range or it’s possible to do full spectrum conversion as well with thermal camera’s? I wouldn’t be surprised germanium lens and filters might be used to pass only FLIR rays suitable for more popular applications. In microbolometer sensors signal is detected by heating bolometers so in theory any sufficient and well absorptive radiation should be possible to detect?

There are microbolometer based SWIR camera‘s which can detect 1.5–5.1 µm spectrum.

[Andy Perrin]

The second link is interesting, although the sensitivity is too low for photography probably. Many of these cheaper SWIR alternatives are geared toward calibrating lasers, which have such a high intensity that the low sensitivity isn't an issue. Graphene is cool but people are hyping it to the skies. Believe in it when you see the camera on the market.

 

Meanwhile I had thoughts about modifying existing technology to detect infrared radiation but I do not have complete technical understanding and might be falling into wishful thinking so correct me if I‘m wrong.

 

Image intensifier tubes such as Gen3 used in night vision devices use gallium arsenide photocathodes to capture infrared spectrum down to 930nm what is enough for applications usual in military or hobbyist night vision. But there are various photodetector materials for different spectrums and intensifier tube offers advantage of amplified sensitivity. For example specially manufactured InGaAs photocathodes can operate up to 1700nm (not sure what is manufacturing technique, thicker coating?). Photocathode screen can be carefully cut out and replaced with another able to convert photons into electrons, no other component of tube requires modification. Vacuum pump will be required to suck out air from tube, as well magnetron sputtering (maybe) to coat photocathode window with desired material.

 

Another question is about microbolometers used in thermal camera‘s. Is sensor spectrum response limited to 7.5–14 μm range or it’s possible to do full spectrum conversion as well with thermal camera’s? I wouldn’t be surprised germanium lens and filters might be used to pass only FLIR rays suitable for more popular applications. In microbolometer sensors signal is detected by heating bolometers so in theory any sufficient and well absorptive radiation should be possible to detect?

 

There are microbolometer based SWIR camera‘s which can detect 1.5–5.1 µm spectrum.

 

1) No such thing as a Bayer array or an IR blocker on an FLIR camera, so there is no such thing as a "full spectrum conversion" either. The limits to 7.5-14 microns comes from the ATMOSPHERE not the sensor. Air is absorbent outside the window. There are other windows though.

2) Plenty of FLIRs own products work in the MWIR or SWIR so definitely microbolometers work in that range, but the LENS is not designed for that window, and the other wavelengths obviously would have to be filtered out. The lens is one of the most expensive parts! Converting a microbolometer LWIR camera to a SWIR one is probably more expensive than just buying a SWIR cam off the shelf.

3) InGaAs is standard tech for SWIR, I already covered that above. No need to go to great lengths with screens and photocathodes -- it can't be cheaper than just using an InGaAs sensor.

 

Finally, I did not mean the SWIR sticky to be a place for this kind of comment, so I am going to ask Andrea to split this off into a separate thread and lock the sticky now.

[Editor's Note: Done 13 May 2020]

[Stefano]

I waited to post until the topic was split.

 

We already talked about the sensitivity of thermal LWIR cameras, and I know that the atmosphere is transparent only in "windows" and it certainly limits the useful IR wavelengths. Probably you can not photograph the Moon at 6 μm, for example. But does the atmosphere absorb IR so well that even a few meters are enough to block "out-of-band" IR light? Can a LWIR camera see light in a broad range of wavelengths at close range?

[Andy Perrin]

Stefano, I actually have no way to know which wavelengths I can see with the LWIR cameras because I have no filters for them with known characteristics. The only way I could identify whether the atmosphere blocks at (say) 6 microns would be to have a band pass filter for that wavelength and see whether the camera gets anything. But bandpass filters for thermal use would have an interesting issue because the *filter itself* is emitting in the range the camera can see! For instance, suppose the camera can see 6 microns but with low sensitivity. If I put on a 6 micron bandpass filter is at room temperature , then the filter will emit 8-13 micron radiation that will possibly drown out the 6 micron signal. It's analogous to having a visible light filter that fluoresces in UV and trying to use to for UVIVF. If any UV hits the filter, it will wash out the image.

[Stefano]

Also the camera lens can glow in thermal. Shiny objects should have a low emissivity though. Maybe the filter must be "metallic" too. I always wondered why shiny metal glows red-hot when heated past the Draper point. It may be due to the thin layer of oxide forming at the surface, but what about the tungsten filament in tungsten lamps? There is no oxide layer in that protective inert gas atmosphere.

[Andy Perrin]

always wondered why shiny metal glows red-hot when heated past the Draper point. It may be due to the thin layer of oxide forming at the surface, but what about the tungsten filament in tungsten lamps? There is no oxide layer in that protective inert gas atmosphere.

 

Stefano - even with a low emissivity for metals, you still see the blackbody radiation. It's just, with emissivity = 0.15 then you only see 15% of the radiation that you otherwise would. It's DIMMER but that doesn't mean nothing gets out at all. Oxidized layers are not required.

 

Also, while the germanium lens is shiny in visible light, it is not in LWIR. It's quite transparent. There are more reflection losses due to the high refractive index of Ge than you have with glass in visible, though.

 

Camera lens glow is an issue but not such a big one because the signal (transmission) to noise (lens glow) is very much in favor of the signal. When I said it was a problem for the bandpass filters, that's strictly because I imagine the out-of-band signal at 6 microns are going to be much weaker than the in-band signal that the camera is designed for.

[Stefano]

Still about tungsten lamps, they could be a way to see if LWIR cameras can see only LWIR. The filament emits abundant blackbody radiation in a very broad wavelength range. The glass/quartz envelope should act as a very rough 4 μm shortpass filter. Can you see the filament inside with a thermal camera?

[Andy Perrin]

 

 

It's hard to tell if it was just warming up the side of the bulb, though. I don't really think we're seeing the filament.

[Stefano]

Well, if it warms up the top because of hot "air" flow, the bottom should be cooler.

 

Maybe the glass envelope is too thin to completely block LWIR.

[SteveE]

With the thermal imager in my FLIR DM285 meter, I cannot see the filament in an 1133 high intensity lamp bulb. It has a clear envelope about an inch in diamater, with a central filament. In the thermal imager, the envelope is opaque red, reading about 200 F with the default emissivity of.95

[Stefano]

I am asking this question because, as Avalon said, there is no reason to have a limited range. Microbolometers can be heated by any form of EM radiation, potentially from Gamma rays to radio waves. And germanium and ZnSe lenses are transparent down to SWIR and green-blue respectively. There is no filter above the sensor, unlike normal cameras (some MWIR cameras have a filter to limit the range to 3-5 μm). I can't understand why this cameras can only see from 7-8 μm to 14 μm. The only thing I can think of is that the absorptive material on the microbolometers is designed to absorb only LWIR.

[Andy Perrin]

I already explained that the atmosphere absorbs outside the 7-14 micron window. I'm not lying, you can see for yourself.

 

There are clearly other windows in there and quite possibly the microbolometer would work for them. (It is not true that it will work for gamma down to radio waves though! Wavelengths larger than the absorption element ("pixel") will obviously not be absorbed at all, so no radio waves, unless you redesign the bolometer itself to have an element big enough to catch them. But higher frequencies would possibly be absorbed.)

 

(from http://www.astronomy.ohio-state.edu/~pogge/Ast161/Unit5/atmos.html, replacing earlier graphic since this one is better)

[SteveE]

Here is the 1133 HI-Z bulb with my FLIR DM285 thermal imager

The right of this image (where the white is on the bulb) is actually up.

 

[Stefano]

Here is the 1133 HI-Z bulb with my FLIR DM285

 

If I understood correctly, the red bulb is the actual glass envelope.

[Stefano]

I already explained that the atmosphere absorbs outside the 7-14 micron window. I'm not lying, you can see for yourself.

Yes, I know. I don't doubt that the atmosphere will absorb at certain bands in the long range, I am referring to short paths. I don't think the absorption is so strong that at 1 meter or less you still have those windows, but I can be wrong.

[SteveE]

If I understood correctly, the red bulb is the actual glass envelope.

You are correct, it is visually completely clear with the filament painfully visible.

[Stefano]

Here you can see a sort of black haze in the distant buildings.

 

 

I don't know if it is a MWIR or a LWIR camera. I should search for the camera model.

[SteveE]

Sorry, double posted instead of editing

[Andy Perrin]

By the way, as a matter of general interest, FLIR has published how they calculate things. Here are the relevant pages from their manual: