Any idea where to get a cheap SWIR or MWIR camera?

[Ben321]

These are less common because they have a narrower range of applications, and this also means they tend to be more expensive. However, there are some interesting phenomenon that happen in SWIR and MWIR wavelengths, which can't be observed in LWIR or in NIR (what you get when you modify digital cameras for IR use). SWIR is just a bit longer than NIR, and MWIR is between SWIR and LWIR. Each wavelength range requires its own type of sensor and its own type of optics.

I've heard that MWIR can be used for some thermal imaging applications, but also because it uses a shorter wavelength than LWIR it also behaves slightly more like visible light (more imaging based on radiation reflected from objects than emitted by those objects), which provides a unique blend of reflected and emitted radiation imaging. This produces some really interesting looking images, if you are into it for artistic photography. For thermal imaging, the purpose of thermal imaging, objects must be significantly hotter than with LWIR, because the increased temperature is needed to shift the emitted wavelengths down into the MWIR region. I also believe some natural phenomenon emit light in this wavelengths range, requiring an MWIR camera to see these phenomenon. For example, there is something called air-glow, which is actually the air behaving in a phosphorescent manner. During the daylight, air is excited by UV light from the sun, and at night it emits in the MWIR region. I've heard that some military night vision equipment even uses this air-glow as its illumination source, so even though most night vision scopes used by the military are either VIS+NIR (image intensifiers) or LWIR (thermal imagers), there are actually some interesting pieces of nightvision technology that the military uses now that are actually MWIR viewers. On top of that, gas-detection cameras often use MWIR wavelengths, because most gases that need detection have strong absorption in the MWIR part of the spectrum. So that could be an interesting phenomenon to investigate as well.

As for SWIR, it also has some very interesting applications. It behaves almost identically to visible light, with most imaging being based on reflected radiation, not emitted radiation, and the optics can be made of ordinary glass. Only the sensor material must be different than visible light sensors. I also have heard that it is also possible to thermal image objects at very high temperatures (only slightly cooler than with NIR imaging). This means that soldering irons can be seen easily in SWIR. I've also read that the wavelengths of IR emitted by the above mentioned air-glow include some of the SWIR wavelengths as well. Other natural phenomenon are probably also better seen in SWIR, and looking at the world with an SWIR camera, I mean just taking one outside and pointing it at various things, could be quite an interesting experience.

Of course, both SWIR and MWIR cameras are quite expensive (even more than LWIR cameras from my understanding). So is there any really good source of cheap SWIR or MWIR cameras I could buy? I'd even accept used units if they had cosmetic damage. If it looks ugly but still works, that's fine. Also, even if the lens itself had some scratches (as long as they were near the edge of the lens, not the middle) I'd be fine with that. And even if the sensor had more hot/dead pixels than would be acceptable for its original application (such that the company that was using it ended up disposing of it), I'd be willing to buy it, if the these pixels didn't make the image look too ugly. Most of these bad pixels could be removed from the image later in Photoshop anyway. I'm looking for used/surplus, VERY CHEAP (Price < $1000, and yes that is very cheap for this type of camera) SWIR or MWIR cameras. If you can point me to an online store that is dedicated to selling used SWIR or MWIR devices at DIRT CHEAP prices, please let me know. Ebay is NOT good for this, because such devices are rare on eBay, and the ones that are sold still go for well over $1000.

[eKretz]

I just posted this on another thread, but might interest you as well:

https://www.researchgate.net/publication/282328406_Comparison_of_medium_and_long_wave_infrared_imaging_for_ocean_based_sensing

[Fraser]

Most older MWIR cameras use a cooled detector. The ones that used a Sterling Cooler are a risky purchase as the life of the cooler is limited. Some will already be dead, others may be almost at EOL. Peltier stack cooled MWIR detectors did exist and may be your best option. Not as good as a Liquid Nitrogen or Sterling  cooled detector but OK. Look for the AGEMA 470 series 'shoulder cams'. They are often to be found cheaply on eBay at around $500. One sold in the UK for £10 ! Make sure the lens is included though.

One on eBay at the moment....

https://www.ebay.com/itm/FLIR-Agema-Thermovision-470-IR-Lens-Infrared-Thermal-Camera/142795007166?epid=12019382501&hash=item213f3ee4be

Another possibility on eBay.....

https://www.ebay.com/itm/Agema-Thermovision-470-Infrared-Systems-Thermal-Camera-w-Case-SEE-DESCRIPTION/123330016898?hash=item1cb70ab682:g:hPUAAOSwik9bgECT

I bought a MWIR Raytheon Amber Radiance 1 Sterling cooled camera very cheaply on eBay and it's cooler is in great condition. There are bargains to be had....... but not from dealers ! Dealers in such technology know it's true value. American Infrared are a reseller if used cameras but they know their value so no bargains there.

http://www.americaninfrared.com/home.asp

It is worth noting that SW and MW are sometimes considered the same when it comes to camera descriptions.

EBay and local government equipment auctions are still your best option.

Fraser

[Ben321]

Most older MWIR cameras use a cooled detector. The ones that used a Sterling Cooler are a risky purchase as the life of the cooler is limited. Some will already be dead, others may be almost at EOL. Peltier stack cooled MWIR detectors did exist and may be your best option. Not as good as a Liquid Nitrogen or Sterling  cooled detector but OK. Look for the AGEMA 470 series 'shoulder cams'. They are often to be found cheaply on eBay at around $500. One sold in the UK for £10 ! Make sure the lens is included though.

You say "Peltier stack cooled MWIR detectors did exist", and I'm noticing the word "did". Does that mean they don't make them anymore? If so, why not? They would be the least prone to failure (unlike the mechanically cooled ones), and they would not require replenishing a supply of cold liquefied gas (unlike the liquid nitrogen cooled ones). In fact, I had assumed that if there was such a thing as a peltier cooled MWIR camera, that it would be the newest type of MWIR camera to have been made, with the Sterling and LN2 cooled MWIR cameras being much older technology (from 1990s and earlier).

And as for SWIR, what exactly is the border between NIR and SWIR? And what is a typical silicon CCD or CMOS imaging sensor's response curve? How far into the IR spectrum does a silicon sensor actually go? At what wavelength in the IR spectrum does it drop below 0.1% sensitivity? And can that be improved (to see deeper into the IR spectrum) by using back-illuminated sensors instead of front-illuminated sensors? How about if you were to use a small piece of wood like a toothpick, to scrape off the color filter array (as well as the microlens array and the optical anti-alias filter), so that the silicon was being directly exposed to the IR light? Would that also improve its ability to see deeper into the IR spectrum? And if so, by how much? Has anybody even done such a study, and documented the results?

[mahony]

A typical monochrom CCD/CMOS Silicon chip w/o any filter will happly see up to ~1µm - see this plot:
https://www.google.de/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&cad=rja&uact=8&ved=2ahUKEwiw1afi_YzdAhUOLlAKHd2uAF8QjRx6BAgBEAU&url=http%3A%2F%2Fwww.flong.com%2Fblog%2F2010%2Fa-brief-note-on-infrared-filters-87-vs-87c%2F&psig=AOvVaw1DGTbIvTcC2-nTfhNRoeZR&ust=1535450697308891

In my experience NIR is usually up to 1.1 µm and mostly covered by silicaon whereas the SWIR Region uses InGaAs or similar expensive detector for roughly 1.5 to somewhere around 2.7µm. Then MWIR covers the range up to 5.1µm with deep cooled (most expensive) InSb or MCT detectors. LWIR being from 7...15µm roughly depending on detector (MCT, Bolometer) and atmospheric window.

[CatalinaWOW]

Border between NIR and SWIR depends on who is talking and the era of the discussion.  Most common for NIR seems to cut off in the 1.2 to 1.4 micron range.  SWIR usually is considered to stop somewhere around 3.0 microns.  This lack of precision means you really should definitely what you mean when you use these terms.  Some of the imprecision comes from military users of these systems trying to be coy about their precise operating bands.

Thermoelectric cooling for sensors in these bands never really took off.  The temperatures required force multistage coolers.  One I saw had seven stages.  Lots of problems isolating the stages, lots of power, and by the time you have that many junctions thermal cycling reliability isn't all that wonderful.  When detectors were small and images were formed by scanning it worked and was sort of competitive with other technologies.  The relatively large size of imaging arrays, and dramatic improvements in Stirling cooler life drove thermoelectric out of the market.

[Ben321]

Border between NIR and SWIR depends on who is talking and the era of the discussion.  Most common for NIR seems to cut off in the 1.2 to 1.4 micron range.  SWIR usually is considered to stop somewhere around 3.0 microns.  This lack of precision means you really should definitely what you mean when you use these terms.  Some of the imprecision comes from military users of these systems trying to be coy about their precise operating bands.

Thermoelectric cooling for sensors in these bands never really took off.  The temperatures required force multistage coolers.  One I saw had seven stages.  Lots of problems isolating the stages, lots of power, and by the time you have that many junctions thermal cycling reliability isn't all that wonderful.  When detectors were small and images were formed by scanning it worked and was sort of competitive with other technologies.  The relatively large size of imaging arrays, and dramatic improvements in Stirling cooler life drove thermoelectric out of the market.

After doing some research, it seems that there is some overlap between NIR and SWIR. NIR is basically all IR between 700nm (edge of the visible spectrum) and 1200nm (the approximate cutoff for silicon based sensors). Meanwhile, SWIR seems to be defined as about 900nm up to somewhere about 2200nm or so. Then there is an unnamed (and not used by any sensors I'm aware of) portion of the IR band between the SWIR and MWIR. MWIR tends to be between 3um and 5um. Then after that there is an unnamed (and not used by any sensors I'm aware of) portion of the IR band between 5um and 7um. Then 7um to 14um is typically considered LWIR. Then there's a very large unnamed portion of the IR band (sometimes called FIR, or Far Infrared) between 14um and THz waves (the shortest wavelengths of microwaves).

[CatalinaWOW]

Looking at atmospheric transmission gives some insight into band definitions.  Additional insight will come from looking into the history of materials used for detection.  Lead salt detectors like lead sulphide we're the first step in moving into the infrared and had much influence on defining the shorter wavelength bands.

[Ben321]

Any idea where to get a cheap SWIR camera (basically any camera that can see between 900nm and 1900nm)? These tend to be more expensive than their LWIR couterparts. Even the sensor itself is HUGELY expensive. I once contacted Hamamatsu (I think that was the company) about a 256x256 SWIR imager chip that I saw they were selling on their website (and I mean just the chip, no circuit board or other components). They quoted me a price of something like $10000 for JUST THE IMAGER CHIP. Anybody here know if that is a typical cost for an SWIR imager chip?

My original plan was to plan get around the high cost of buying an SWIR imager, by buying the sensor array chip itself, and then building the device itself as a project. I'd need design a circuit, program a microcontroller, figure out how to interface it to a computer, and of course pay to get somebody to etch a circuit board, and then solder the components. Not to mention carefully drilling holes in the board and mounting a C-Mount lens holder (I already have a glass C-Mount lens) on the board. It would be hard work, and take a while, but it could be done.

However, I filed that plan in the trash bin, the moment I found out that the focal plane array chip itself would cost $10000. I seriously hope that is not a common price for these SWIR focal plane array chips, and that it was only high from that one company because it came from the famous component manufacturer Hamamatsu (sort of like how Sony TVs are always more expensive than Visio TVs, because Sony is such a famous TV manufacturer).

[Vipitis]

The high resolution FLIR Tau Core for SWIR is sold at 15k https://www.oemcameras.com/tau-swir.htm

Boson probably even more. There are some other cores I the market but no price tags.

[CatalinaWOW]

Part of this depends on what you define as cheap.  A decade ago 128x128 InSb cameras were available new from several vendors for $20k or less.  I haven't followed the market but those cameras and their 256x256 descendents should still be available at roughly the same price.

With some clever filtering, optics cooling and cold shielding you should be able to operate one of the LWIR bolometers as a MWIR sensor.  Haven't really thought it through to see what performance to expect or what the necessary components might cost.

[Ben321]

The high resolution FLIR Tau Core for SWIR is sold at 15k https://www.oemcameras.com/tau-swir.htm

Boson probably even more. There are some other cores I the market but no price tags.

As far as I know, the Tau series, is FLIR's only series that has an SWIR unit. All of FLIR's other series (such as Boson) are only LWIR. Also, I think Boson is their cheapest series (even cheaper than Tau). The only cheaper series than Boson (which is a cheap professional series) is the Lepton series meant only for hobbyists. So if There was an SWIR Boson, I suspect it would be cheaper than an SWIR Tau.

[Ben321]

Part of this depends on what you define as cheap.  A decade ago 128x128 InSb cameras were available new from several vendors for $20k or less.  I haven't followed the market but those cameras and their 256x256 descendents should still be available at roughly the same price.

With some clever filtering, optics cooling and cold shielding you should be able to operate one of the LWIR bolometers as a MWIR sensor.  Haven't really thought it through to see what performance to expect or what the necessary components might cost.

InSb requires deep cryo cooling though. InGaAs can operate at room temperature (or at worst it might require some thermoelectric cooling). I thought the high prices for older sensor technology (like InSb) was the result of requiring an expensive cryo cooler. But it seems that the imager chip itself costs at least $10000, which as far as I'm concerned is price gouging. There's no way that a chip that could fit on the tip of your finger is actually worth as much as a cheap car. That must be at least 100x the manufacturing cost of the chip. Most companies are satisfied with making a profit that is 2 to 10 times the manufacturing cost (depending on how greedy they are), but there is no way in the world that any product should be sold at 100x the manufacturing cost. That is completely unjustified. There is no legitimate business reason to charge 100x the manufacturing cost of a product.

[Fraser]

FJW Find-R-Scope Cameras have been an option for SWIR imaging for some years. I almost purchased one on eBay for around £200 so they do come onto the market cheaply now and then. There are also the Find-R-Scope optical scopes (like old style image intensifiers) but they suffer image distortion due to design. Also note that the cameras do not use a simple glass lens, they use a lens material optimised for the SWIR spectrum, but I forget its name.

https://www.findrscope.com/CategoryProductList.jsp?cat=IR+Camera+1800+%2F+2200nm:+Infrared+Cameras+with+400-2200+nm+sensitivity.

Fraser

[Ben321]

FJW Find-R-Scope Cameras have been an option for SWIR imaging for some years. I almost purchased one on eBay for around £200 so they do come onto the market cheaply now and then. There are also the Find-R-Scope optical scopes (like old style image intensifiers) but they suffer image distortion due to design. Also note that the cameras do not use a simple glass lens, they use a lens material optimised for the SWIR spectrum, but I forget its name.

https://www.findrscope.com/CategoryProductList.jsp?cat=IR+Camera+1800+%2F+2200nm:+Infrared+Cameras+with+400-2200+nm+sensitivity.

Fraser

If I'm not mistaken, SWIR between 1000 and 1800nm (or maybe even 2000nm) can work with normal glass lenses (but possibly with some small amount of absorption of the light by the lens). Not sure how long the wavelength can get before large amounts of absorption force you to use a silicon lens (like for MWIR).

Also, if I'm not mistaken, FindRScope uses an image conversion tube which (much like a light intensifier tube for a conventional nightvision scope) is a vacuum tube that gets hit by photons on one end, which ejects electrons, which hit a phosphor at the other end. The phosphor then emits visible light. And, like all old-school electronics (like radios) that used to have vacuum tubes, you have a filement which heats the cathode. And the heater filiment WILL burn out after a certain period of time (unlike a modern InGaAs solid-state SWIR sensor, which theoretically can last indefinitely because there's no heat to burn out parts of the device). So while the FindRScope may have a cheaper initial cost, over the time you'd use it, you'll need to replace it multiple times (like changing the old incandescent bulbs in your ceiling lights). Spending $1000 every year or 2 to replace the device is NOT cost effective.

Also, I don't think the FindRScope can see much longer wavelengths than about 1300nm or 1500nm (depending on the model). That's too short to see the sky-glow phenomenon, which I believe I read is closer to 1900nm. InGaAs on the other hand, I believe can see the full SWIR spectrum from 900nm to 2200nm (or whatever the official upper limit is). The FindRScope would definitely not be the best option.

[Fraser]

Please check the link I provided !

The Find-R-Scope cameras have used various technologies for the detector.

Wavelength conversion can use the same technology as image intensifiers and does work at SWIR. One of the cameras detailed in that link has a bandwidth of 400nm to 2200nm. You would likely have to add filtering to lower that bandwidth if wanting to look at just SWIR. It was the 400-2200nm camera that sold for around £200.

On the topic of wavelength conversion tubes....

Please research image intensifier tube construction. There is no heater filament as they are cold cathode electrostatic devices. Life is dependant upon the vacuum retention and any abuse that image burns the fragile phosphor coatings.

Fraser

[Ben321]

Please check the link I provided !

The Find-R-Scope cameras have used various technologies for the detector.

Wavelength conversion can use the same technology as image intensifiers and does work at SWIR. One of the cameras detailed in that link has a bandwidth of 400nm to 2200nm. You would likely have to add filtering to lower that bandwidth if wanting to look at just SWIR. It was the 400-2200nm camera that sold for around £200.

On the topic of wavelength conversion tubes....

Please research image intensifier tube construction. There is no heater filament as they are cold cathode electrostatic devices. Life is dependant upon the vacuum retention and any abuse that image burns the fragile phosphor coatings.

Fraser

Ok. I see. I was thinking you were talking about these https://www.findrscope.com/CategoryProductList.jsp?cat=Infrared+Viewers:Standard+Infrared+Viewers

[Vipitis]

This one says Boson SWIR in 640x512 resolution. I suspect that it is more expensive because it's newer. But they don't have a pricetag... So buying a single one might be far more expensive that getting a few thousand.

E: forgot the link: https://www.oemcameras.com/boson-swir.htm

[CatalinaWOW]

Part of this depends on what you define as cheap.  A decade ago 128x128 InSb cameras were available new from several vendors for $20k or less.  I haven't followed the market but those cameras and their 256x256 descendents should still be available at roughly the same price.

With some clever filtering, optics cooling and cold shielding you should be able to operate one of the LWIR bolometers as a MWIR sensor.  Haven't really thought it through to see what performance to expect or what the necessary components might cost.

InSb requires deep cryo cooling though. InGaAs can operate at room temperature (or at worst it might require some thermoelectric cooling). I thought the high prices for older sensor technology (like InSb) was the result of requiring an expensive cryo cooler. But it seems that the imager chip itself costs at least $10000, which as far as I'm concerned is price gouging. There's no way that a chip that could fit on the tip of your finger is actually worth as much as a cheap car. That must be at least 100x the manufacturing cost of the chip. Most companies are satisfied with making a profit that is 2 to 10 times the manufacturing cost (depending on how greedy they are), but there is no way in the world that any product should be sold at 100x the manufacturing cost. That is completely unjustified. There is no legitimate business reason to charge 100x the manufacturing cost of a product.

InSb fpas are far more complex than a silicon imaging sensor.  First problem is sensor array itself.  InSb is no where near as nice as silicon for processing, and there are many orders of magnitude less manufactureing volume.  With that smaller industrial base electronics to read sensor can't be done in InSb so a silicon mating chip is built.  That chip costs what you are expecting.  But then the InSb chip and the silicon chip must be mated.  Several approaches have been tried.  None are cheap or easy.  Finally the hybridized sensor must go in a vacuum Dewar with an optical window and hermitic electrical connections.  Another cost not found in your basic visible sensor.  Costs depend on array size but ten years ago, when I retired such packaged sensors were well under 10,000 and sold at a reasonable mark up.

Those unavoidable costs of MWIR sensors are why there was so much push for bolometer development.  They can be produced with relatively minor modifications to standard silicon processes, with readouts in the same chip.  The cost advantages are why they are used, in spite of performance limitations.

[Ben321]

Part of this depends on what you define as cheap.  A decade ago 128x128 InSb cameras were available new from several vendors for $20k or less.  I haven't followed the market but those cameras and their 256x256 descendents should still be available at roughly the same price.

With some clever filtering, optics cooling and cold shielding you should be able to operate one of the LWIR bolometers as a MWIR sensor.  Haven't really thought it through to see what performance to expect or what the necessary components might cost.

InSb requires deep cryo cooling though. InGaAs can operate at room temperature (or at worst it might require some thermoelectric cooling). I thought the high prices for older sensor technology (like InSb) was the result of requiring an expensive cryo cooler. But it seems that the imager chip itself costs at least $10000, which as far as I'm concerned is price gouging. There's no way that a chip that could fit on the tip of your finger is actually worth as much as a cheap car. That must be at least 100x the manufacturing cost of the chip. Most companies are satisfied with making a profit that is 2 to 10 times the manufacturing cost (depending on how greedy they are), but there is no way in the world that any product should be sold at 100x the manufacturing cost. That is completely unjustified. There is no legitimate business reason to charge 100x the manufacturing cost of a product.

InSb fpas are far more complex than a silicon imaging sensor.  First problem is sensor array itself.  InSb is no where near as nice as silicon for processing, and there are many orders of magnitude less manufactureing volume.  With that smaller industrial base electronics to read sensor can't be done in InSb so a silicon mating chip is built.  That chip costs what you are expecting.  But then the InSb chip and the silicon chip must be mated.  Several approaches have been tried.  None are cheap or easy.  Finally the hybridized sensor must go in a vacuum Dewar with an optical window and hermitic electrical connections.  Another cost not found in your basic visible sensor.  Costs depend on array size but ten years ago, when I retired such packaged sensors were well under 10,000 and sold at a reasonable mark up.

Those unavoidable costs of MWIR sensors are why there was so much push for bolometer development.  They can be produced with relatively minor modifications to standard silicon processes, with readouts in the same chip.  The cost advantages are why they are used, in spite of performance limitations.

I think this was the InGaAs SWIR imager chip I was looking at that cost so much. https://www.hamamatsu.com/us/en/product/type/G13393-0808W/index.html

Note that this isn't an InSb chip. It's InGaAs, so I assume it works differently than the InSb chips (possibly not even needing a readout chip physically attached to it, and possibly not even needing a vacuum container). Now I could see that costing several hundred dollars for just the chip (compared to a silicon CCD chip with the same pixel count costing under $100). But there's no reason for it costing $10000+. No manufacturing process would make a single chip, smaller than the tip of your finger, cost that much. I assume this imager chip has the readout chip underneath it (if it needs it that is) already built into the chip they are selling. And I assume it probably needs no vacuum at all, as I see nothing that resembles a vacuum flask in the picture (though the buyer of the chip may need to purchase their own vacuum flask separately and mount the chip in it). So the price they are selling it at, is basically ROBBERY.

[CatalinaWOW]

Believe what you want to.  It is very difficult to build things in unconventional materials and low volumes.  In one case I was involved in the the yield of a very, very small sensor array was well under one percent for the first half dozen production runs.  To my knowledge it never approached 50%.  Volumes were never over a few thousand a year.  Since the process equipment can't be used for other things you must amortize that cost over a small total production.  InGaAs has a production base that may approach a thousandth of a perecent of silicon, and these arrays, even if priced as you would like would have sales volumes measured in thousands, not millions or tens of millions so you just can't compare the costs to high production silicon.

[CatalinaWOW]

Reading the link you provided does provide many clues to the cost.  In the first line it says it is a hybrid.  So the readout is not included in the InGaAs array.  It then says backside illuminated.  Indium bump bonded to the readout.  Meaning that to avoid other very difficult problems the electrical connections to the diodes are made through the diodes side of the wafer.  But since the whole point of these things is to absorb light and convert it to electricity  InGaAs will be a very poor transmitter of the desired wavelength and the InGaAs layer will have to be extremely thin.  I don't know which process they are using, but all I know of are difficult, which translates to expensive.  One example is to bond the array to the readout and then grind the back surface away until it is thin enough.

[eKretz]

Ben, you have no idea what you're talking about. Your knowledge base regarding manufacturing is obviously lacking. Many items the size of a fingertip cost many thousands of dollars, and justifiably so. You can't only think about the material and labor that it takes to make the item. There's also the cost of the building, insurance, utilities, the machinery needed to make the parts (all of which together can cost MANY MILLIONS of dollars) and many other factors in the cost for any given item that is sold. Low volume means a larger percentage of this cost must be factored into those tiny items.

[Fraser]

Ben321,

Whilst I understand your frustration at the cost of some specialist parts, maybe a little story will help you better understand the situation......


I decide to put some of my hard learnt knowledge into a new imaging sensor that can detect previously unknown "Geek" particles in space. I spend a year of my life designing the imaging sensor and another year searching for Unicorn Tears that are essential to the sensors operation. Boy those Unicorns are hard to find, and as for making them cry.... well that was so hard until I used the Tazer on them !

I harvest enough Unicorn Tears to make a batch of the "Geek" particle sensors. I test them and they work beyond expectations. The Unicorn Tears dielectric constant is far better than expected and the specific gravity is perfect for the sensor.

Pretty soon NASA hear about my "Geek" sensor and they send the "Boys" round to negotiate supply of such for a new Exploration Satellite costing 1 Billion Dollars. Now I know most of the NASA space rated sensors are tens of thousands of Dollars and they do not contain very hard to find Unicorn Tears !

I set my unit price per sensor at $100K or $98K if purchased in a batch of 10 and paid through PayPal Friends and Family. The NASA boys jump at the chance to fly the sensor and buy 10 for experimentation. Harvard University buys 2 and MIT buy 5.

Whilst enjoying the financial returns on my hard work and riding on a lovely white Unicorn, I receive an email on my iPhone asking to buy one if my "Geek" sensors for hobby work. I advise the sender of the $100K cost plus $12 postage and packing with "On yer bike" couriers Inc.

I soon hear that someone on a prominent and respected forum is calling me a crook and a robber because I am asking "too much" for my sensor. Sure enough the person who enquired about buying a sensor is stating that it is too expensive and "Robbery".

So let us analyze the situation....

I design and build a specialist sensor, fill it with the 'secret sauce' (Unicorn Tears) and I get paid a lot of money for it by large companies and institutions. This is the reward of my efforts and all the pain of getting kicked in the nuts by angry Unicorns after I Tazered them for Tears !

Why should I sell the sensor for less to anyone who wants one ? It contains my work and the magical Unicorn Tears that others cannot find. What incentive is there for me to make the sensor available cheaply when the market happily pays $100K per unit and demand heavily outstrips supply of Unicorn tears ?

Now if my design was such a 'rip-off' another party would buy one, reverse engineer it and make a cheaper copy to gain a slice of the pie. The problem they find is that the 'secret sauce' cannot be identified or found. I do a deal with a third party to make the sensors as my shed is getting overcrowded with little projects like higher voltage Tazers for Unicorn Tear extraction. The deal is the Third party build the sensors and I supply the 'secret sauce' in individual serial numbered viles. The sensor is distributed through the Third Parties network and they charge the standard $100K per unit. I get $90K per unit sold and only have responsibility for collecting the Unicorn Tears.

The Third Party sells lots of sensors and the market is still strong. There is no incentive to reduce the price as the users of the sensor are content with what they are getting for their money.

So Ben, why is it Robbery to ask a high price for a sensor that cannot be knocked up on a kitchen table ? We live in a Capitalist World.

Hope you liked the story.

Note: No actual Unicorns were harmed during the making of this story I don't want PETA after me !

Fraser

[Fraser]

And for anyone who does not believe Unicorns And Unicorn Tears Exist,

Here is a picture if some Gin that contains Unicorn Tears !