Mµ Thermal Imager - real or fake?

[mikeselectricstuff]

Ooh look what I found typing that microbolometer part number into google
 Flir Microbolometer cost analysys report
 Full version only $3700! Anyone got an un-blur filter...?

Another article here
From the above

The microbolometer is encapsulated in a hermetic housing consisting of a ceramic substrate HTCC, a metal box and an IR window in silicon. The microbolometers operate under a high vacuum, 10-4mbar, 100 times lower than for a MEMS gyroscope. The final test integrates an expensive four-day leak test to guarantee package tightness.

First time I've seen a getter used in a semiconductor device!

[firewalker]

And still no trace of that cheap sensor.

Alexander.

[EEVblog]

2nd page:
http://www.bluetoad.com/print.php?pages=27&issue_id=132414&ref=1

[mikeselectricstuff]

Looking through these articles it seems there isn't much to be saved from reducing the resolution - looks like a large part of the cost is the packaging & test. The die photo shows the sensing part is maybe a quarter of total die area - by the time you scale the peripheral stuff, a low-res version might maybe halve die area, but not significantly affect any of the other costs.

[mikeselectricstuff]

Good find guys 

I have wanted to know more about the FLIR microbolometers but my contacts in the industry clam up when asked anything about them....NDA's.

I am surprised about the sensor resolution revelation. I would have thought production cost and failure rate would have been lower on lower resolution sensors which would have been an incentive to make sensors matched to units. Obviously this is not the case.

From my understanding of the DoD restrictions on the technology, it applied to the SENSOR and not the rest of the unit, yet manufacturers consider all technical info to be covered under the restrictions. In either case, the fact that the sensor is a hi res 320x240 type interests me. Unless it is hardware crippled to not operate at 30Hz, I would have expected it to definitely be liable to the ITAR restrictions as a sensor, forgetting whether the rest of the camera electronics are 30Hz or 9Hz. Dual use regulations apply to the harvesting of parts as well.

I would imagine regs cover (possibly differenty) component parts and finished equipment. 
Once you get into the technicalities of how fast a sensor can be driven I suspect things get rather muddy. It could be argued that the sensor is useless without drive electronics, and they don't document the interface. There's also the issue of running stuff out of spec. If a sensor is designed and specified for 9FPS, chances are it _can_ be run faster. 
Same as a gun barrel being able to withstand more powerful ammo than it was designed for. 
This may also be connected to restricions on supplying information as it may help someone re-purpose things. I'm sure the sensor in a FLIR i3 could make a pretty good heat-seeking guidance system.

the whole dual-use stuff is nonsense as pretty much anything _could_ be used for mischief...

[mikeselectricstuff]

...
I have actually considered the idea of doing a kickstarter to buy something exotic for a teardown, but the ITAR stuff would make me slightly nervous of doing it with this.

Wait, don't those other cameras you disassembled also fall under ITAR? What were their resolutions?

The FireFLIR probably does. It's 320x240 at I think 30fps, definitely more than 9fps. A KS campaign would attract more publicity, and possible interest from the manufacturer.
In terms of revealing info, those teardown reports contain way more info than what I could glean by a full teardown. I wonder if they (attempt to) restrict distribution of those...?

[IanB]

As far as I know, ITAR is a US regulation that covers technology developed in the USA and exported from the USA.

So the simplest way to avoid falling under these regulations is to independently develop equivalent functionality outside the borders of the US.

This is something that will eventually take care of itself as the USA is rapidly abandoning its technical lead in the race to outsource everything to the rest of the world.

A second problem with control and secrecy over most "sensitive" technology is that the lawmakers fail to understand that anyone with sufficient education can deduce and manufacture identical things without reference to the secret documents. The secrets of the universe are really not something you can hide in a safe. They are sitting there in plain sight for anyone who cares to look.

[mikeselectricstuff]

From what is said it appears the 9Hz refresh is a software configuration matter and if set can be exported under lesser restrictions than full frame rate units without the 9Hz limitation imposed.
All very odd as any software limitation can often be reversed.

It says "Factory set" - OK it almost certainly is a software setting, but it could also be something like a different FPGA config file, which is for all intents and purposes different hardware. If they wanted to get really serious they could put a fuse on the sensor die that was blown to either limit readout rate (e.g. lowpass filter on analogue out), or be readable by the system software/FPGA to prevent use at higher rates.
Regulation-wise, as long as they don't tell the user how to change it they're not producing something that falls under the regs. If someone modifies or hacks it, the problem is with that person, not the original manufacturer, as it is the modifyer who is effectively manufacturing a device whose performance brings it within the scope of the regulations.

[mikeselectricstuff]

As far as I know, ITAR is a US regulation that covers technology developed in the USA and exported from the USA.
So the simplest way to avoid falling under these regulations is to independently develop equivalent functionality outside the borders of the US.

They can still affect you though - There are mechanisms which can make it illegal for any US company to deal with you, and many other countries have similar regulations.
And the US has managed to persuade some countries to roll over like a bunch of pussies to extradite people (e.g. Christopher Tappin).
And that's before the NSA and other spooks do whatever the hell they like.
   

[mikeselectricstuff]

I wonder if whoever thought up the 9fps rule considered that you could buy three or four 9fps cameras and strap them together....?
 

[mikeselectricstuff]

Just spotted one TIC that offers resolution upgrades :
http://uk.rs-online.com/web/p/thermal-imaging-camera-accessories/7776710/?origin=PSF_437238|acc

[EEVblog]

From what is said it appears the 9Hz refresh is a software configuration matter and if set can be exported under lesser restrictions than full frame rate units without the 9Hz limitation imposed.
All very odd as any software limitation can often be reversed.

It says "Factory set" - OK it almost certainly is a software setting, but it could also be something like a different FPGA config file, which is for all intents and purposes different hardware. If they wanted to get really serious they could put a fuse on the sensor die that was blown to either limit readout rate (e.g. lowpass filter on analogue out), or be readable by the system software/FPGA to prevent use at higher rates.

The document says the sensor itself is a hybrid containing the sensor array and an ASIC chip. It is possible they could OTP (laser or whatever) some config bits in the ASIC.

[EEVblog]

And the US has managed to persuade some countries to roll over like a bunch of pussies to extradite people (e.g. Christopher Tappin).

Or lean on your government via back channels through another government to spend millions of dollars, go against international law, and make complete public arses out of themselves by holding certain Australians hostage in embassies.
If they are willing to do that, imagine how fast they'll comply with a trivial request against a simple video blogger 

[TopLoser]

Cheers Aurora... I'm now the proud owner of a Talisman I just bought from a nice man in Doncaster. Picking it up ASAP!

[MFX]

Seems everything about thermal imaging is expensive :-
http://uk.rs-online.com/web/p/thermal-imaging-camera-accessories/0446594/

£91 for a car charger! Now that's really taking the mick.

[mikeselectricstuff]

Just spotted one TIC that offers resolution upgrades :
http://uk.rs-online.com/web/p/thermal-imaging-camera-accessories/7776710/?origin=PSF_437238|acc

Are you sure that upgrade isn't just like the Testo one (http://www.testosites.de/export/sites/default/thermalimaging/resources/downloads/Background_article_SuperResolution_EN.pdf)? They offer a program where the camera takes two images a few milliseconds apart and fuses them in software to increase resolution. You've got to pay for the program though, and it doesn't look at all applicable to other manufacturers' cameras. I would love to do this with an ix, but the i-series has a terrible capture rate so it's doubtful whether that kind of program is even possible. ttp://www.testosites.de/export/sites/default/thermalimaging/resources/downloads/Background_article_SuperResolution_EN.pdf

Probably  - I just assumed it was crippleware on a real 320x240 sensor

[tom66]

Hmm. Sounds like they're stalling.

Over the last few weeks we've moved into new office space and now have all of our equipment set up again. The team was still able to get a lot of work done during the chaos, even when that meant taking equipment home to work. I'll follow up with more news next week.

But, I'll give them credit: they are at least staying in touch.

[EEVblog]

I find it funny how people are still asking for a delivery date estimate, as if that's magically going to make it happen.
Until they show that working prototype, it's all just pie-in-the-sky.
At least this one has been funny to watch and follow along, although sadly at the expense of those who contributed.

[MacAttak]

Denial isn't a river in Egypt.

[Fraser]

This is the long and painful death of a poorly planned project. They should have the balls to put it out of its misery. If any money still remains they should partially refund the investors as a small token of 'doing the right thing'

[MFX]

It does seem that the natives are starting to get restless rather than blindly defending it as they were. As for the office move, 3 weeks? And this meant they had no internet access at all to let people know what was going on?

[CanadianAvenger]

Maybe they moved to hide from the angry mobs coming to bang on their doors requesting their cameras.

[Corporate666]

It boggles the mind that people are still posting things like "Guys, can we please get an update?  I would like to know if you have decided what color the housing will be" or some such.

If they had ANYTHING, they would be shouting from the highest rooftops.  The project is following the classic manner in which such things go down the tubes.  They give out practically zero info and do so irregularly, and it's only naivete and denial that keeps people going.  Nobody wants to think they lost their money, and some are greedy enough not to want to miss out on something they think they're getting for $200 that usually costs 10 times that much.

That's how all cons work.

[MFX]

Although for a "con" it's weird they've kept it going this long and are still giving refunds when they could have done a runner with the money ages ago. Also someone running a con would likely be doing all they could to keep the marks happy (regular fake updates etc.) so if they are running a con they don't seem to be very good at it. And they do seem to have a PCB with what looks like a genuine Thermal imaging lens on it even if they can't show any images from it. I agree that the whole project seems impractical for the cost (I bailed out ages ago) but mysteries still remain.

Martin.

[Corporate666]

I don't think it started out as a con.  I think it started out with naivete and maybe they thought they had an inside line on some way to do thermal imaging cheap (like using a traditional imaging sensor).

I think a project can transgress from "honest if naive attempt" to "con" when the project creator realizes the idea they presented is untenable.  In the case of Mu, the secret sauce of their whole project was the imaging sensor.  They said after the close of funding that the imaging sensor manufacturer had dropped the ball in some manner, so they were switching.

Which is sort of like someone selling you on their miraculous discovery of gold in their yard... then after you invest, telling you the gold turned out to be someone's lost ring, but no worries, because they also struck oil in the neighbors yard and he's going to let you have some for cheap.