$40 Lens Hack Gives Your FLIR Higher Clarity

[Josh Oster-Morris’s] FLIR camera can see a bit more clearly now that he’s hacked it to have its own makeshift “macro” mode. You may remember [Josh] from his power distribution Motobrain project. He’s still improving the Motobrain, and he wanted to better understand the thermal characteristics of the high current draws (upwards of 100amps!)

After reading that the FLIR 4  could be hacked into a better version, [Josh] immediately purchased his own. The FLIR is, however, limited at close-range imaging, because the resolution of the FLIR’s microbolometer is relatively low.  He had fortunately decided to stay tuned in to [Mike’s] YouTube channel and saw his follow-up video a few days later on refocusing the FLIR camera with an external lens. [Josh] hit up Amazon for a Gallium Arsenide lens normally used for CO2 lasers, and found one for around $40. He then mounted this lens into a simple paper frame held together by tape and staples, and fitted it onto the FLIR.

After you’ve checked out [Josh’s] blog for more examples of how astoundingly clear the images become, check out [Mike’s] video detailing the hack below.

26 thoughts on “$40 Lens Hack Gives Your FLIR Higher Clarity

    1. If you are ‘the’ mike, it might interest you to know (if I didn’t mention this before) that the minimum warranty of electric equipment, which cover a rather broad range but basically is anything electric, is a mandatory 3 years in the EU . So if US stuff says it’s 2 they are actually mistaken since they have to give three in the EU.

      You can tell because a lot of stuff in the shop acts like they are doing you a special favor by putting in large print 3 years warranty on the box.

      I read about this in the news because it appears there is a lot of abuse too because even big companies try to get you to pay for an ‘extended warranty’ beyond 1 or 2 years not informing you you already get 3 years in the EU.

      1. The three years is an “expected lifetime” of a product during which you can contest the warranty, and it’s not a mandatory rule. Implementation depends on the country. The companies aren’t required to give such warranty, but consumers can still get covered to a part of the purchase price if the product fails and it’s a clear defect and they take up with the consumer protection board (or equivalent).

        In practice, it’s an uphill battle and you’ll most likely win only a fraction of the price as a refund, especially with electronics where the companies change modes every year and simply don’t sell the same product for longer than its warranty so they wouldn’t have to replace them.

        The extended warranties are shop warranties, because if you bring the device in after the shop warranty has ended they have no responsibility to take it in; you have no on-site warranty or equivalent and the usual warranty replacement rules (max 3 weeks return time etc.) won’t apply – you have to deal with the manufacturer/bureaucrats yourself.

    2. It doesn’t change the resolution of the sensor but does increase the resolution of the system at close range. It is a fundamental property of optical systems that the better the focus the higher the resolution.

  1. Is it the editors being that negligent, or did the submitter not pay attention? No is is not a GaAs lens (those are dark blue/brown). This is a ZnSe lens, which is a transparent yellow colour.
    Cool (or hot) hack anyway.

    1. There are a few links in this post, the video of mike does show the yellow lens, but the main link is not to mike’s stuff.and the lens does not appear yellow. He just emulated mike but obviously ordered his own lens based on what he could get.

      As the article says “He had fortunately decided to stay tuned in to [Mike’s] YouTube channel and saw..”

      This time it seem the commenter is the negligent one :)

  2. Well, this is a neat increase of spatial resolution but still, you have to bear in mind that one compley process you can not trick ist the calibration of the sensor and the lens. So, if it is about understanding what temperature does and where it goes, it’s fine – but as soon as you need to read exact temperature values all calibration what has been made to the original sensor and the original lens becomes obsolete because of the transmission curve of the GaAs Lens in the Longwave Band (7-14µm) of the Sensor and the original Germanium Lens coming with the camera. So it impacts the accuracy but will be ok if it is just for the gradients. Probably not all of the incoming radiation passes the GaAs Lens. Check Wiki for the transmission factor of GaAs in the a.m. Waveband. If it is 90% you can add approx. 10% to your absolute measurements.

  3. Thanks for the hint on the lense focus method. I have had a Flir i5 for a while for some pcb work for my small business and suffered poor focus. I did a search for lenses and came across the “Shenzhen Reliable Electromechanical Technology Co Ltd” in China. They said that Gallium Arsenide was obsolete, but they carried a range of Zinc-Selenium lenses (znse) which were better (see: http://www.szret.com/en/znse-co2-laser-focus-lens-40.html) The 19mm lenses with 50mm focal length were about $40, but because of the stepped-taper in the front of the I5, it was going to be hard to fit. (like the inside of a DEVO hat if you know what I mean) So I ordereded the bigger 25.4mm lense with the 50.8mm focal length for closer to $100, This was dearer, but it fitted nicely, almost flush with the front of the I5.. I made up a round holding plate from some pcb scrap to keep it in place. Results were amazing. At 50mm I could fill the screen with a really sharp image of a small coin, whereas before it was just a vague round blur. Totally worth it! The camera is now a much more usable tool for workshop use. In retrospect, it is a small crime that the flir distributor that sold me the camera a couple of years ago had no interest in highlighting the limitations of the standard I5 camera for electronics work.

    Thanks for the inspiration to make the upgrade. Ian, Drouin West, Australia.

        1. By reducing the focal length for sharper images with the extra lens, the unit has to be closer to the objects to maintain the focus. I think 100mm would be blurry. Maybe 50mm for focus would be more realistic. I would upload some sample jpg captures if I could figure out how to send them to this site.

          1. Sorry typo. I was trying to ask 100m is there a lens that would allow the unit to work for bigger distances or is the unit ok at long distance. For hunting dear at night.This would be a cost effective way but the back light may be a problem.

          2. Ah, Ok Brent. I understand. Actually, they work well at a distance. Not quite like the IR images you see from police helicopters following robbers climbing over fences, but the I5 is ok at a distance. I am on about 8 acres of farmland here. I have 3 cows and 5 alpacas. At night it is pitch black in the paddocks and I routinely use the IR camera to find out where they are. At 100 metres plus, they tend to be moving blobs because of camera pixelisation, but it is in stark relief to the surrounding terrain which is much colder. Because of their short hair, the cows are extremely visible to spot on IR. The alpacas at this time have a lot of wool insulation, You can still see them, but their faces and knees which are not covered, show up very brightly under IR. Cats are also easy to find at night.
            If you have an email address, or another method of posing images to this site, I can post some animal images to you as examples.

            Another use I found was for checking water tank levels. Night or day I can point the instrument at a water tank (concrete or plastic) and see at a glance how full the tank is because of the thermal gradient. I have speculated that a water carrier selling water to remote properties could do well going up the road with an IR camera to spot tanks that are low, then leaving a business card at those locations. A good business opportunity for someone.

          3. Thank you for the info Ian Thats exactly the info I need . is that with the seek unit not the flir one http://www.flir.com.au/flirone/ . would the flir unit be better for this application . i have been looking at a flir C2 or even the flir FX (very interesting but can it streem thermal imaging http://www.flir.com/flirfx/ )?? Thanks for the help
            We canot get the cameras in australia so may neend to get them in the US and have them posted over, anty thoughts . brenthudson2@gmail.com

          4. I hadn’t seen that FLIRONE before. Looks like a pretty good thing. I don’t know if they are available in Australia. Maybe you could contact the manufacturer and look for a distributor closer to home. The Australian dollar is not great at the moment against the USD.. I am about to send some sample animal images from my I5 to your email address. Good luck.

  4. “As soon as you need to read exact temperature values”, either you’re using the wrong instrument (and should use a contact thermometer) or you need to calibrate for the thing you’re looking at and the way you’re looking at it regardless of using an added lens or not (for which calibration you’ll need to use a contact thermometer). (And even contact thermometers won’t measure the temperature exactly, due to being in thermal communication with the environment independently and differently from the thing whose temperature you want to know, though I don’t think that usually matters much.)

    This applies to any infrared thermometer or thermal camera, unless it uses multi-wavelength methods to take different surfaces’ emissivities into account, and even that only works somewhat. Emissivity (and thermal emission spectrum) varies with substance, texture, thickness, coatings, viewing angle, etc., and a non-contact temperature-reading device (of whatever type) usually has no way to know those factors or their impact on the target’s emissivity, other than a single number that the user probably didn’t remember to guess at. It only shows the amount of infrared light (and only that within its passband) coming from each point on the target object, possibly scaled by a constant factor or formula (derived from the user-entered emissivity setting) so it appears on a “temperature” scale. And not all of that light coming from the target is actually emitted by it, because most things have nonzero thermal reflectance—try pointing a thermal camera at any flat and bare or thinly painted metal surface and you’ll see it makes a pretty good mirror.

Leave a Reply to Ian JacksonCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.