3 ms·

I used to work on infrared imagers long ago in the late 80s, so I still sometimes notice obscure technical details about IR. Many modern anti-tank weapons were

I used to work on infrared imagers long ago in the late 80s, so I still sometimes notice obscure technical details about IR. Many modern anti-tank weapons were designed back in the 90s and then battle-tested over the next decades. So like aircraft and spacecraft, a lot of the electronics is quite antiquated by modern standards. For example: The Javelin shoulder-fired missile that we are currently reading about in Ukraine was put into service in the late 90s. It finds its way to the target using an infrared focal-plane array with a resolution of 64x64 [1]. Part of the reason that the resolution is so small: A room-temperature object will have peak E&M radiation around 8-12 microns, and mercury-cadmium-telluride imagers were optimal for that wavelength range in the 90s. However, processing of the materials was pretty difficult back then, so 64x64 was pretty much all you could manage while achieving a usable yield. Eventually the current generation will be replaced, and that will presumably include a serious upgrade to the IR resolution. It's hard to imagine that won't affect the usefulness of tanks in the coming decades. [1] https://en.wikipedia.org/wiki/FGM-148_Javelin#Focal_plane_array_(FPA) https://en.wikipedia.org/wiki/FGM-148_Javelin#Focal_plane_ar...

tyingq 5y ago
It does seem like high resolution would make a big difference. If you look at pictures of tanks that are pretty fully covered with reactive armor, there are still lots of gaps of varying sizes for things like hatches, view ports, exhaust, between reactive armor modules, and so on. So a high res focal plane could direct the weapon into one of many sweet spots.
- chipsa 5y ago
  I think the point of higher resolution isn't to necessarily hit the sweet spots, but rather to better improve countermeasure rejection.
livueta 5y ago
Sorry if this is a dumb question: Reading your link, it sounds like the HgCdTe-based FPAs are primarily an electrochemical thing, akin to the microchannel plate approach of traditional analog image intensification. My impression of more modern IR-spectrum imaging (Teledyne/BAE/L3 stuff) was that it is basically a bolometer sensor behind vanadium/germanium-doped glass, then somehow hooked up to a cmos sensor. So, modern sensors are still partially based in an electrochemical process, just a different one that's more amenable to improvements in manufacturing process or something? I'm thinking of some of the new stuff that can deliver 1920x1080 60hz output - that'd definitely help with targeting.
- jpmattia 5y ago
  > it sounds like the HgCdTe-based FPAs are primarily an electrochemical thing HgCdTe is a semiconductor like Si or GaAs, but it has a much smaller bandgap. > My impression of more modern IR-spectrum imaging (Teledyne/BAE/L3 stuff) was that it is basically a bolometer sensor Bolometers rely on heating. HgCdTe (& CMOS imagers) rely on photon generation of electron-hole pairs, which is an altogether different mechanism. > behind vanadium/germanium-doped glass The most common reason for putting Ge in front of an IR imager is to filter out more energetic photons (ie, visible light), since those will also be detected by most IR imagers. > I'm thinking of some of the new stuff that can deliver 1920x1080 60hz output - that'd definitely help with targeting. Yes. State of the art is UHD, eg https://www.teledyneimaging.com/en/aerospace-and-defense/products/sensors-overview/infrared-hgcdte-mct/ https://www.teledyneimaging.com/en/aerospace-and-defense/pro... The issue will be the product cycle, which for the military is extremely long.
  - livueta 5y ago
    I appreciate the reply, this cleared up some misconceptions.
tlavoie 5y ago
Is this also why civilian IR devices are typically slow and/or very low resolution, or alternately very expensive? My fire department thermal imaging cameras are a few thousand CAD. A smaller, palm-sized one from FLIR is several hundred dollars, but the frame rate is very low. I have a Seek model that plugs into my phone, and while very inexpensive, it's definitely at the tail end of capability. (Still fun though.) In comparison, I have had a tour of a medical transport helicopter, and got to try the pilot's NVG goggles. It's light intensification rather than IR, but an impressive upgrade from the Starlight scope I got to use in the 80s.
- jpmattia 5y ago
  > Is this also why civilian IR devices are typically slow and/or very low resolution, or alternately very expensive? I haven't followed the professional/consumer market in IR, but as a general statement: The materials systems that you need for IR imaging are different than the CMOS imagers used for visible light (ie, silicon). The IR materials don't have anywhere near the economic or manufacturing scale, so I'd expect that you'll pay more for IR imagers, or will have to accept lower resolution/frame rates.

tyingq 5y ago

It does seem like high resolution would make a big difference. If you look at pictures of tanks that are pretty fully covered with reactive armor, there are still lots of gaps of varying sizes for things like hatches, view ports, exhaust, between reactive armor modules, and so on. So a high res focal plane could direct the weapon into one of many sweet spots.

chipsa 5y ago
I think the point of higher resolution isn't to necessarily hit the sweet spots, but rather to better improve countermeasure rejection.

livueta 5y ago

Sorry if this is a dumb question: Reading your link, it sounds like the HgCdTe-based FPAs are primarily an electrochemical thing, akin to the microchannel plate approach of traditional analog image intensification. My impression of more modern IR-spectrum imaging (Teledyne/BAE/L3 stuff) was that it is basically a bolometer sensor behind vanadium/germanium-doped glass, then somehow hooked up to a cmos sensor. So, modern sensors are still partially based in an electrochemical process, just a different one that's more amenable to improvements in manufacturing process or something? I'm thinking of some of the new stuff that can deliver 1920x1080 60hz output - that'd definitely help with targeting.

jpmattia 5y ago
> it sounds like the HgCdTe-based FPAs are primarily an electrochemical thing HgCdTe is a semiconductor like Si or GaAs, but it has a much smaller bandgap. > My impression of more modern IR-spectrum imaging (Teledyne/BAE/L3 stuff) was that it is basically a bolometer sensor Bolometers rely on heating. HgCdTe (& CMOS imagers) rely on photon generation of electron-hole pairs, which is an altogether different mechanism. > behind vanadium/germanium-doped glass The most common reason for putting Ge in front of an IR imager is to filter out more energetic photons (ie, visible light), since those will also be detected by most IR imagers. > I'm thinking of some of the new stuff that can deliver 1920x1080 60hz output - that'd definitely help with targeting. Yes. State of the art is UHD, eg https://www.teledyneimaging.com/en/aerospace-and-defense/products/sensors-overview/infrared-hgcdte-mct/ https://www.teledyneimaging.com/en/aerospace-and-defense/pro... The issue will be the product cycle, which for the military is extremely long.
- livueta 5y ago
  I appreciate the reply, this cleared up some misconceptions.

tlavoie 5y ago

Is this also why civilian IR devices are typically slow and/or very low resolution, or alternately very expensive? My fire department thermal imaging cameras are a few thousand CAD. A smaller, palm-sized one from FLIR is several hundred dollars, but the frame rate is very low. I have a Seek model that plugs into my phone, and while very inexpensive, it's definitely at the tail end of capability. (Still fun though.) In comparison, I have had a tour of a medical transport helicopter, and got to try the pilot's NVG goggles. It's light intensification rather than IR, but an impressive upgrade from the Starlight scope I got to use in the 80s.

jpmattia 5y ago
> Is this also why civilian IR devices are typically slow and/or very low resolution, or alternately very expensive? I haven't followed the professional/consumer market in IR, but as a general statement: The materials systems that you need for IR imaging are different than the CMOS imagers used for visible light (ie, silicon). The IR materials don't have anywhere near the economic or manufacturing scale, so I'd expect that you'll pay more for IR imagers, or will have to accept lower resolution/frame rates.