SPi RazIR mini thermal camera -Teardown and comment

[Fraser]

Dear Readers,

I have just purchased a SPi RazIR mini thermal camera. This is actually a Wuhan Guide Infrared Co. Ltd MobIR M3 model. 160 x 120 pixels with 60fps output. These thermal cameras are best known under their SPi branding as Wuhan Guide Infrared Co. Ltd did not have much market presence until relatively recently. Chinese thermal cameras can be a 'mixed bag' when it comes ergonomics, quality and design. They range from good to appalling !The same goes fro their firmware and PC software.

The RazIR and MobIR models have been around a few years now and sold for around $10K to $12K new ! I never liked them due to their cheap appearance and 'flip-phone' format.  I bought this unit only because it is in need of repair and I want to see what hides inside the case. I expect to find a decent quality French ULIS 160 x 120 microbolometer and a Germanium lens driving an electronics package that resembles the ULIS reference design.

The RazIR is unusual in not only its shape but also its technical design. It contains two LCD displays. The display that is visible with the cover closed is basic STN technology but the main display under the cover is a normal TFT type. Both displays can present the thermal scene to the user.

I am not expecting great things from the firmware or analysis software but I am willing to be surprised. The brochures claim a capable image capture engine with alarms etc and an equally capable image analysis software package for the PC. Connectivity to a PC appears to be USB 1.1 via a custom connector on the camera.

Once I have the camera (it is en route to me now) I shall be dismantling it for repair. This thread will contain pictures of the teardown so that others can see what $10K would buy you. Do not expect to be that impressed ! This is a relatively early generation of Chinese Thermal Camera. It is quirky in many ways but we shall see what sort of pictures it can produce.

Watch this space !

Fraser

[Fraser]

The User manual and brochure:

[Fraser]

Some exterior pictures:

[joe-c]

i hat a M2 a while ago and build a pc interface for that:
http://joe-c.de/pages/posts/waermebildkamera_120x120_80.php

The Image quality is quite good. but i don't like the storage only in internal flash and the limited way to get the files to pc.
No Problems to use 32bit WinXp, but finding a 64x driver was hard long time ago.

I also had a M4 and repaired the Flex:
http://joe-c.de/pages/sonstiges/flex_reparieren.php#M4

both cameras was sold, but i have some info's from inside if you want (need some days, starting vacation in garden now  ).

[Fraser]

Thanks Joe-C !

Great to hear that you have already been inside these cameras

I have taken a look at your linked pages .... excellent work. In the case of the M4 and its broken FPC ..... I have faced similar challenges and such a failure is not an easy repair.

I would love to see more internal pictures as and when you have time to find them. I have already identified the microbolometer used in theses cameras thanks to your pictures of the M4. As expected they use the well known French ULIS 160 x 120 '02 05 1' microbolometer that I have come across in several cameras. This particular microbolometer is temperature stabilised but its package sadly has a weakness at the glass seals for the leads. If the seals crack, the capsule vacuum is lost and the microbolometer is scrap. A good friend in the industry warned me to handle these Microbolometers with great care to avoid exerting undue tension on the leads. It is a decent microbolometer though and is more than capable of producing a decent thermal image.

I have a ULIS 02 05 1 that I extracted from a camera. I shall upload a picture of it. Decent engineering that made use of an industry standard package.

A thread I created about the ULIS 02 05 1 may be found here....

https://www.eevblog.com/forum/thermal-imaging/the-ulis-02-05-01-microbolometer-an-insight-into-teh-technology-by-fraser/

Fraser

[Fraser]

Pictures of the ULIS 02 05 1 microbolometer.

Note the bulky mechanical packaging compared to modern MEMS Microbolometers such as those used in the FLIR LEPTON and SEEK Thermal cameras.

Fraser

[sam1275]

Thank you very much for let us know that SPI is in fact a Chinese company, some of their products looks so Chinese and their website is also a bit suspected(over advertise), but they disguised as a well US company, that's really annoying. At least I know this before buying anything from it, thanks!

[Fraser]

Sam1275,

Sorry, I should have made it clear....... SPi is a USA company that sells various brands of thermal imaging camera. They import the Wuhan Guide M3 in an SPi branded livery. No different to many other companies who source products from China.

SPi are an 'interesting' company with somewhat amateur videos on you tube.

They obey ITAR and do not export or even answer questions from outside the USA. Not a company that I have managed to have any dealings with.

Fraser

[sam1275]

Sam1275,

Sorry, I should have made it clear....... SPi is a USA company that sells various brands of thermal imaging camera. They import the Wuhan Guide M3 in an SPi branded livery. No different to many other companies who source products from China.

SPi are an 'interesting' company with somewhat amateur videos on you tube.

They obey ITAR and do not export or even answer questions from outside the USA. Not a company that I have managed to have any dealings with.

Fraser

Hi Fraser, sorry for my misunderstood and I appreciate your explain. So how can we spot a product that is "source from China"?

[joe-c]

Hello Again,

I have here a package with some info's and images:
https://drive.google.com/file/d/0B-YjTGxntt5KRGZzczAzX3B3alk/view?usp=sharing

I recorded the DataStream with 2 different Analyzers, so you need 2 Applications to read them:
for *.scana files: https://www.ikalogic.com/scanastudio_v2/
(not the 2.4 Beta, use the 2.3xx regular release)
for *.logicdata files: https://www.saleae.com/downloads

there should be also a record for the i2c Traffic. if I remember right, there 2 digital Temperature sensors, one near the Detector core and one near the top of the Display, inside the cover, next to the speaker.

I hope it's usefully
best wishes
Joe-c

[Bill W]

Hi Fraser, sorry for my misunderstood and I appreciate your explain. So how can we spot a product that is "source from China"?

It is just a question of knowing who are 'real' manufacturers and (sometimes) who are distributors.
In this case we have Wuhan, a known Chinese manufacturer who sell products in their own right, and a US source selling identical items fairly cheap barring some labels.
Therefore this one is a Chinese product being imported.

In contrast if you have say a known UK/EU/US manufacturer product and find a very similar Chinese product on Alibaba it is either someone winging it, fishing for business or a clone job.

Just taking it apart is no real clue as several FLIR products look very 'Chinese' inside apart from the core.  Take the H series, untidy burnt wiring held together with glue and bad soldering.  Chinese box on a US core.

regards
Bill

[Fraser]

Joe-C,

Many thanks for the information and pictures. Very much appreciated 

Fraser

[nidlaX]

Apparently SPi is a trusted name in the science and research community. I wouldn't have known given their branding and marketing.

Did Wuhan Guide Infrared ever make a 320 x 240 version?

[Bill W]

Not sure of the 'flip phone' design ever had a 320x240 in it, but Wuhan did several cameras and modules based around ULIS 320x240's - even if some of their publicity images are of ISG product 

Bill

[Fraser]

The M3 had arrived safely. It came with two battery packs that were declared 'dead'. Having tested them and got 0V that was accurate.

The battery pack contains a single cell and the case is glued at the edges. Some careful taps with a small hammers broke the glue bond very easily. These packs are eminently rebuildable if you can find a suitable cell.

The metal cased Li-Ion cell used in the battery pack is a LIS983448B. It is rated 3.7V 1550mAh. Not the easiest cell to find. I have ordered a 1000mAh LiPo cell that will do for now. Charging will be via an appropriate LiPo external charger. The cameras internal Li-Ion charger may be OK for the LiPo cell but that will need further investigation before using it.

The original cells had bulged and provided virtually no p.d at their tabs. There is a separate protection board present as would be expected.

More when I get time to look further at this weird looking camera.

Fraser

[Fraser]

Pictures of the disassembled battery........

[Uho]

Hey. I did for the Mobir M3 eyecup and changed the magnification of the lens. This thermal imager does not cost as much money. More like a toy. To use it is not convenient. I was disappointed and did not make good photos.

[Fraser]

Uho

I agree. It feels and looks very cheap and would not normally be my first choice of thermal camera. It does have a decent ULIS microbolometer in it though. Poor images are likely the result of inadequate image processing.

This unit was very inexpensive so it was an opportunity to take a look at the design. You did a very nice modification for the viewfinder. You have excellent skills Uho

Fraser

[Uho]

Yes. It is a very bad image processing. No anti-aliasing. At the same resolution, the 2000AS module works better. I made a thermal imager on it. The image is much better.
https://youtu.be/Z5ocJQ6wmo8

[Fraser]

Uho,

How do you make such nice cases for your projects ? Are they metal or plastic ? The case on that thermal scope looks good enough to be a commercial product !

Fraser

[Uho]

The case is made of plastic. I printed it on a 3d printer. Grinding with grinding machines. Painted in several layers of paint for plastic. The cover was sewed by the wife. The 2000AS module itself needs a control board and a voltage converter board. I developed them myself. Here I compare c flir ps24. Look at the comparison of the diameter of the lens.
https://youtu.be/KDzDS8_u-EM

[Fraser]

And to think some people say you cannot make nice looking cases with a 3D printer    Your thermal scope case proves them wrong.

I print in ABS so that I can sand the surfaces smooth and even glue parts together, like a model kit  I have yet to make a project casing though.

Very nice work 

Fraser

[Fraser]

I have just rebuilt the battery pack for the RazIR/M3.

I elected to use a commonly available Canon Li-Ion battery that fits neatly in the original battery case. The Canon battery is lower capacity at 1050mAh, but that is no big deal for me. The batteries are very reasonably priced at 3 for £10 and contain their own protection board. For my purposes they will have more than enough capacity.

I tested the M3 camera and was pleasantly surprised at the images that it produced. The user interface could be better but just takes some getting used to.

I now need to either find or change the connector type that is used for the USB and video I/O port on the camera. I attach a picture of it. it appears to be an 8 pin connector similar to that found on older mobile phones and PCMCIA cards.

More about this camera when I find time and feel motivated enough to dive inside it  I need to construct a new hinge for the display section and that will take some planning. My ABS 3D printer may be needed to make a new part.

Fraser

[Fraser]

In case anyone is wondering, the rebuilt batteries will be charged outside the camera on a custom charger dock as I do not trust the internal charge circuit (the original cells were bulged). I may investigate the camera charge circuit later to see what, if anything, is wrong with its design.

Fraser

[Fraser]

The I/O connector appears to be a very slightly modified 4 pin mini USB type. Two keyways are missing either side of the central keyway at the top. I will need to check the pinout but it looks like some careful plastic removal on a standard 4 pin mini connector USB cable will work.

This particular 4 pin USB mini USB connector format is from HIROSE. There is a different 4 pin mini USB connector by Mitsumi.

[Fraser]

It appears to be a tweaked version of the Hirose MQ172 series. Datasheet attached.

Fraser

[Fraser]

The 4 pin mini USB lead has arrived.

It certainly looks the same as what is required and the main dimensions match.
The normal 4 pin HIROSE type mini USB plug has three alignment 'keys', one wide one at the mid point and one at each edge. The 'keys' at the edges need to be removed in order to use the plug with the RazIR/M series cameras. The removal of the two thin 'keys' appears quite simple and I shall first try a scalpel. If that does not work, a very fine file should make short work of them

I then need to check for correct fit and pinout compatibility with the standard USB A plug at the other end of the lead.

I previously stated that this USB port is used for video output as well as USB comms. That was incorrect. I have the video cable and it plugs into another port on the side.

More on this when I get a chance to play with the USB lead

Fraser

[Fraser]

A useful picture I just found, showing different digital camera communications connectors.
Keep it safe somewhere, it could have saved me quite a lot of time searching !

Fraser

[Fraser]

Well I just found a really interesting cable OEM site. Very useful !

It shows the type of 4 pin HIROSE USB connector that the RazIR/M series uses. It even includes the wiring details. I still need to check the camera pinout though.

http://www.dcscsi.com/product/363.html

Another source of cable connector identification pictures

http://www.dcscsi.com/product/cat-6.html

Fraser

[Fraser]

Cable connector modified with ease........ it fits ! I did have to slightly modify the connectors outer plastic case as it was too large to allow the plug to fully seat in the slightly recessed socket. I actually removed more material than was required.

The wiring of this generic cable was correct for the RazIR/M3.

Once connected to the PC, I was able to access the cameras stored images and make radiometric measurements. Different palettes may be selected and the .IRI radiometric files may be saved as .BMP or .JPG for easy viewing.

I provide some pictures of the cable connector modification and some pictures of Oliver downloaded from my camera. I tried different palettes as well. The saved images are at original resolution and are not upscaled so they are quire small.

Fraser

[Fraser]

Pictures of Oliver........

[jagans]

Hi there. I bought one of these for 6500 dollars from SPI quite a few years ago. They no longer support these cameras, and when you could get them the batteries, a lousy 3.7 volt Lipon brick, was around 110 dollars. INSANE! I guess it was for an IR camera so they figured, why not rip them off, they were already stupid enough to pay 6500 for a POS. These cameras cannot take the rigors of outdoor PM use. They are like toys or like a Chinese Sailboat, JUNK. Companies tat do not support what they sell are assholes. HP did it with plotter drivers. These companies should be ashamed of themselves. LG is another one. Lousy Engineering, Lousy Products, Lousy support. $6500 dollars is still a lot of money to me. Im not the government that steals everyones income and calls it taxes.

[jean-paul]

Hello,

I bought that camera 2 weeks ago. It 's good for the price! FOV is perfect, and manual focus is good when you take pictures inside!

But it seems that there's a bug in the software, when you export a picture from their proprietary format (IRI) to bmp/jpg without the palette you cannot remove the min/max temperature.

So I tried to reverse engineered their file format and I managed to display pictures, but it’s in “raw” format. Do you have an idea on how to convert it to a radiometric image? It seems that it depends on the camera internal temperature ?

Thank you,

[jean-paul]

Hello,

I managed to reverse engineered a part of their IRI file but I haven't managed to find out the formula to get temperature from raw data.

The good thing is that the Guide software tell you the raw value and the corresponding temperature, so I did a table with a fixed emissivity, humidity, distance and a variable ambient temperature.

Then I did a regression using this formula : https://github.com/cran/Thermimage/blob/master/R/raw2temp.R

The problem is that there are 5 unknown! So I did an optimization with 2 differents ambient temperature. With a fixed ambient temperature it works great but if ambient temperature changes then it's not working anymore