Questions about vaccum of thermal sensors

[doxuya]

1. Do cooled detectors also need vaccum packages like uncooled ones?
2. How long can a sensor remain usable vaccum level?
3. What will happen when a sensor lost its vaccum? And how to determine the current vaccum level of a sensor?
4. Can a sensor be serviced after losing its vaccum?

[Spirit532]

1. Yes
2. Decades, if cared for. Slightly reduced performance due to leakage(He and such).
3. Convection makes the image unusable, sometimes the sensor can be damaged by the inrush of air. There's no practical way to measure the current pressure.
4. Yes, usually. Not every time.

[Ben321]

3. Convection makes the image unusable, sometimes the sensor can be damaged by the inrush of air. There's no practical way to measure the current pressure.

I may be mistaken about this, but I believe that the small size of the vacuum cell means that even if air leaked in, it is too small of a volume for convection currents to form. Do you know why 2-pane windows are good insulators? It's because the air-filled gap between the 2 panes is too small for convection currents to form, and the main mechanism of heat transfer between the panes is conduction. Conduction is the slowest method of heat transfer, and it's faster through a solid than a gas. Therefore air, when convection is not present, is a VERY GOOD thermal insulator. My personal opinion is companies that make thermal imagers just use vacuum packaging for the sensors so that it's a "fancier technology", so that they can justify selling their devices at an insanely high price (aka price gouging). But that's just my opinion.

If somebody here has facts regarding this, and could disprove my theory, please let me know.

[bap2703]

[quote author=Ben321 My personal opinion is companies that make thermal imagers just use vacuum packaging for the sensors so that it's a "fancier technology", so that they can justify selling their devices at an insanely high price (aka price gouging). But that's just my opinion.
[/quote]

A funny one !

I guess at cryo temperature you only have a fraction of a watt of cooling, hence a 'good insulator' at your 'standard daily conditions and level of power' isn't good enough anymore.

Water condensation might also be an issue albeit coming only after performance degradation.

[bugi]

I may be mistaken about this, but I believe that the small size of the vacuum cell means that even if air leaked in, it is too small of a volume for convection currents to form. Do you know why 2-pane windows are good insulators? It's because the air-filled gap between the 2 panes is too small for convection currents to form, and the main mechanism of heat transfer between the panes is conduction. Conduction is the slowest method of heat transfer, and it's faster through a solid than a gas. Therefore air, when convection is not present, is a VERY GOOD thermal insulator.

While conduction might be slow, it is still not the only form left; conduction+radiation is still faster than radiation alone. Hmm.. these IR cameras try to measure external radiation, not the temperature of the air inside the sensor, and the effects of the external radiation can be quite small compared to the internal heating. That might be one reason to keep it in vacuum.

Also, while air is indeed good insulator (for normal building purposes), a very thin layer of it makes it quite poor insulator in total (in technical purposes). I have some plans which require really good insulation (planning using those expensive exotic aerogel stuff, comparable or better than air), and I still need several cm of it, not just a millimeter.

Here in Finland we often have 3-layer windows, one gap being smaller, another several cm. I don't know the specifics, but at least in some designs the replacement air into the rooms comes actually flowing through that/those gap (top to bottom and/or bottom to top direction(s)), so, convection would be a minor thing compared to that flow.

Also, air makes often a mess of nice materials through oxidation and what not. This could be controlled by using e.g. dry nitrogen or such, but this still needs sealing (so no price benefit) and will still have heat conduction and will still affect the response of those tiny pixels.

My personal opinion is companies that make thermal imagers just use vacuum packaging for the sensors so that it's a "fancier technology", so that they can justify selling their devices at an insanely high price (aka price gouging). But that's just my opinion.

If I'm not completely mistaken, I think it is quite well understood that making a product in a more expensive way and thus also more expensive to sell will reduce its sales, and the total profits (while the profit % could get higher, especially for luxury-type products). It is almost always better to try the make the product as cheap as possible (yet trying to sell it at highest profit). Also, as long as there are multiple manufacturers of the tech, someone would sooner or later drop the artificially expensive production method in order to try and get more profit, and soon competitors would have to follow, driving the prices lower in general. Unless there is a cartel between all the manufacturers.

[Fraser]

Ben321,

No vacuum, no thermal image.... simple. Ask any thermal camera designer or experienced service tech.

If you still doubt what they tell you, buy a cheap older thermal camera with the vacuum nipple on the sensors casing and puncture the vacuum nipple. See what happens  Hint, do not expect to be able to use the camera afterwards.

I must say, I love reading your theories. Everything can, and maybe should, be challenged ..... but please do research topics carefully first. Read patents and scientific journals. In this case, maybe start with the original Honeywell patents 

Fraser

[Fraser]

For anyone wondering about how double glazing works and the difference between a void containing air and a good quality vacuum, here is an interesting discussion thread.....

http://www.sciencechatforum.com/viewtopic.php?p=284623

Convection currents are real and a problem in thermal imaging sensors. A hard vacuum is essential in the design. Air is too thermally conductive and remember that the die in a Microbolometer reaches natural thermal equilibrium at around 32C and a cooled die in a Sterling Cooled camera is operated at 77K. Any thermal conduction into the void around the sensor array in the form of convection currents impacts performance. A cooled sensor operates inside a Dewar to increase the efficiency of the system. The coolers cold finger has to cool just the sensor die and not the environs around it. A cooled camera that suffers total loss of vacuum in the dewar around the sensor array and comes up to atmosphere struggles to achieve 77K, suffers frosting and basically does not work.

Fraser

[Max Planck]

Therefore air, when convection is not present, is a VERY GOOD thermal insulator.

My energy bills, especially on cold winters, would disagree 

My personal opinion is companies that make thermal imagers just use vacuum packaging for the sensors so that it's a "fancier technology", so that they can justify selling their devices at an insanely high price (aka price gouging). But that's just my opinion.

If I were you, I would follow Frasier's suggestion of reading some patents. If not, in the article quoted below you will find the results of tests of microbolometer FPA performance under different vacuum levels.

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.873.2431&rep=rep1&type=pdf

Max

[Spirit532]

Convection currents

I may have not been clear, what I meant is that air would, as you say, conduct too much heat. Convection/brownian motion would play a part though, since that's how you reach thermal equilibrium of a gas in a (perfectly) sealed container with (absolutely) no external sources of change (the spherical horse in a vacuum kind).

[Ben321]

I understand how a vacuume being lost would bring big problems for a cooled unit, but for one already operating at room temperature, I don't see how it needs a vacuum. All they would need would be a thermister mounted like a milimeter away from the VOx microbolometer, to measure the air temperature immediately adjacent to the VOx microbolometer. Then you could use DSP processing in a microcontroller to subtract the ambient temperature from the image to get the desired temperature in the image.

[Spirit532]

The air will conduct far more heat than the lens could impart on the sensor from the image.
Unless you were imaging something ridiculously large and hot(thousands of degrees, meters in diameter), you wouldn't really be able to see anything at all.

Again, read the paper linked, or just buy a camera and pop the nipple and connect it to a vacuum pump. See the performance change at varying levels of vacuum.

[bugi]

I understand how a vacuume being lost would bring big problems for a cooled unit, but for one already operating at room temperature, I don't see how it needs a vacuum. All they would need would be a thermister mounted like a milimeter away from the VOx microbolometer, to measure the air temperature immediately adjacent to the VOx microbolometer. Then you could use DSP processing in a microcontroller to subtract the ambient temperature from the image to get the desired temperature in the image.

They don't need thermistor or such; the FPA's have "blind pixels" on the edges (normal pixels but with a cover over them, preventing the external thermal radiation to hit those pixels); their output is already indicating the FPA's own temperature, and used for the appropriate purposes (note, more than one purpose). If I remember right, TE-Q1 SDK also estimates ambient temperature from the FPA temperature by math (i.e. they have characterized the FPA vs. ambient temperature difference). Certainly, a real sensor for that would be better, but it is a hazzle to get it measure accurately the external temperature in a small dongle (read: measuring the outer case temperature is easy).

[Bill W]

I understand how a vacuume being lost would bring big problems for a cooled unit, but for one already operating at room temperature, I don't see how it needs a vacuum. All they would need would be a thermister mounted like a milimeter away from the VOx microbolometer, to measure the air temperature immediately adjacent to the VOx microbolometer. Then you could use DSP processing in a microcontroller to subtract the ambient temperature from the image to get the desired temperature in the image.

Vacuum is needed or the pixels do not heat up when you focus a hot scene on them

Responsivity with vacuum = 2000 (arbitrary units for a given temperature difference)
Responsivity at air pressure = 5 (arbitrary units for a given temperature difference)

Bill

[Max Planck]

Au contraire 

If the scene is hot, vacuum will not be so critical. It becomes critical at lower scene temperatures. It is all about heat transfer phenomena within the FPA pixel. The scene radiation heats up the pixel membrane, thus changing its resistance. Sensitivity is limited by the fact that heat from the membrane escapes through bolometer support legs, which is unavoidable. Vacuum eliminates pixel heat losses through gas conduction or more precisely it makes it negligible compared to the heat transport through the support legs.

Max   

[Fraser]

@Max Planck,

Whilst I do not disagree with your comment, it does need to be placed in the context of whether a thermal camera Microbolometer array will operate satisfactorily without the vacuum being present. I would argue that the performance in such a circumstance is so poor as to render the camera pretty much useless. If wishing to image only hot items is the only requirement, fine, but most thermal cameras are required to image both low and high temperatures. The vacuum is essential to this capability. Ben321 appears to be asking whether the vacuum is essential or just some weird marketing trick to increase retail prices. I would firmly state that the vacuum is essential to a sensors acceptable performance in almost all common usage cases. It most definitely is not some marketing ‘trick’ as the developers of Microbolometer arrays have been trying to simplify and cost reduce the encapsulation for decades ! There is an ongoing Military project to develop ultra compact, low cost thermal cameras. The microbolometer sensor encapsulation has been redesigned to simplify production and has resulted in what we see in the latest miniature ULIS sensors. The Lepton follows the same principles. The SEEK microbolometer is an even smaller and lower cost solution. The actual vacuum capsule volume has been significantly decreased to a mere vestige of the size found in older encapsulations, but it is still present. The developers just cannot get rid if it. A smaller vacuum chamber is not all good news though. There is greater potential for vacuum poisoning that degrades performance and minute vacuum leakage over time has a greater effect on a small vacuum chamber volume than a larger one.

I think that just about covers things

Fraser

[Bill W]

Au contraire 

If the scene is hot, vacuum will not be so critical. It becomes critical at lower scene temperatures.
Max   

This is true only in the sense that if imaging a furnace in a steelworks you could use a sensor with an NeTD of 10K.  Indeed a camera for those circumstances would probably be adjusted through gain reduction and have that NeTD.

Bill

[Max Planck]

@Max Planck,

Whilst I do not disagree with your comment, it does need to be placed in the context of whether a thermal camera Microbolometer array will operate satisfactorily without the vacuum being present. I would argue that the performance in such a circumstance is so poor as to render the camera pretty much useless. If wishing to image only hot items is the only requirement, fine, but most thermal cameras are required to image both low and high temperatures.

[...]
Fraser

Fraser,
I just commented on the incorrect information given by Bill W.
I also think I commented earlier clearly on the vacuum importance.

If you are interested, the key element here is in fact the thermal conductance of the membrane (pixel)  - which is the sum of the thermal conductance through the supporting legs Glegs and the thermal conductance through the air Ggas (if no vacuum) - and the thermal time constant of the pixel. Given typical geometry of actual microbolometers, if vacuum is lost, Ggas becomes roughly two orders of maignitude bigger than Glegs. You can find a good description in the articles below:

He, X., Karunasiri, G., Mei, T., Zeng, W. J., Neuzil, P., & Sridhar, U. (2000). Performance of microbolometer focal plane arrays under varying pressure. IEEE Electron Device Letters, 21(5), 233–235. doi:10.1109/55.841306
Eriksson, P., Andersson, J. Y., & Stemme, G. (1997). Thermal characterization of surface-micromachined silicon nitride membranes for thermal infrared detectors. Journal of Microelectromechanical Systems, 6(1), 55–61. doi:10.1109/84.557531

As a sidenote, the authors in the first article used a 20 um separation between the membrane and the substrate, thus lowering Ggas about 10 times comparing to typical designs which are using in most cases rougly a 2.5 um gap to create a quarter-wave Fabry-Perot resonator and boost the sensitivity around 10 um. Vacuum still remains the best option.       

[Hyper_Spectral]

Another great read from when I delved into the depths of researching this topic..

doi:10.12693
http://dx.doi.org/10.12693/APhysPolA.127.1295

http://przyrbwn.icm.edu.pl/APP/PDF/127/a127z4p121.pdf