The X-Ray image thread by Aurora - various electronics via X-Ray imaging

[Fraser]

Good news on the new Faxitron MX-20 USB.

The original PC was toast so I needed to load the Faxitron SR software onto another PC. Mike has already done this so I knew it was possible. I thought the following information might assist any future owners of a Faxitron MX-20 USB version.

I decided to do a bit of experimentation with my little Sony Netbook. I have the installation CD for SR versions 1.0, 1.5 and 1.6. I started with v1.5. The software requests a unique Key file as part of the installation. Without this file the software will not run. The Key file contains a complex hash that uses the MAC address of the licensed PC. You cannot, therefore use the Key file on another PC without tricking the SR software into thinking it is on the original licensed hardware.

To make the Key file valid on another PC you have to spoof the licensed PC's MAC address. This is easy as the required MAC address is detailed in the Key file itself. You can then use the Ethernet adapter advanced properties option to set the MAC address to that contained in the Key file.

This is what I did and, as Mike discovered, the SR software happily accepts the Key file and the installation completes with a fully licensed install. There is no 'calling home' by the software during installation.

If you do not have the original Key file you are not going to be able to complete the installation process. This is worth bearing in mind if you find a Faxitron MX-20 without its PC (a common occurrence with disposal equipment as the PC's contain patient data). The Key file may not be present on any accompanying disks as each customer is issued with their own unique key from Faxitron HQ.

There is also a file on the computer after the installation completes that contains the data from the Key file. Whether this can be renamed to be the Key file is not known. The Key file has an unusual name. I will attach a sanitised version of mine for others information on format etc.

It looks like I will be able to install the SR software on a replacement PC so I will not have to repair the original MSI Hetis 915 motherboard. Testing of the MX-20 with the Sony Netbook to follow.

In the attached Key file I have replaced the licensed company name with 'X' and the MAC address is set to 01.02.03 etc. You will see that there is a long Key contained in the file that I have replaced with '0's. This is a hash but I do not know what variables are used to create it. Possibly, licensed company name, expiry date sequence plus the MAC address of the host PC's Ethernet card. The key appears to be valid across all versions of the SR software so the Hash does not contain version specific variables.

Fraser

[mikeselectricstuff]

If you do find a machine with no PC or no software, it may be possible to drive the Hamamatsu sensor directly - I did find some Hamamatsu utils, don't recall if on the Faxitron install or at Hamamatsu, though didn't look close enough to see exactly what was possible with them.

[jdraughn]

A NFC tag might be interesting to xray, but maybe not, it would probably just show a coil of wire around a little rectangle and that's it.

[Fraser]

Well the Faxitron MX-20 USB lives 

I bought a replacement MSI Hetis 945 Dual Core 2 (2 x 2.4GHz) SFF PC to replace the dead Hetis 915 P4 3GHz that came with the USB Faxitron machine. I had to import it from Germany ! I wanted the MSI Hetis9xx as it is a decent little PC platform for use with the Faxitron and its software. The alternative that I considered was a Dell USFF PC as they are now very cheap on eBay.

I didn't transfer the hard disk from the 915 as the motherboard chip sets are different (Intel 915 Vs 945) and I wanted a clean XP install onto which to place the SR1.6 software. It avoided any issues with HDD transplantation.

All went well with installing both XP SP3 and SR1.6.The original imaging array Defect Map was transplanted into the new SR install easily. The MAC address on the new PC was changed to that contained within the Faxitron Licence Key file using the Ethernet advanced properties page. No problems there.

Once the new PC was refitted on the Faxitron Cart, and all cables connected, I fired the system up and allowed the Imaging array to warm up for 30 miniutes. The Imaging array was then calibrated and a test image taken.

I am pleased to report that all went well with the first test of the 'new' USB Faxitron. I attach the first image that was captured using the automatic exposure system. This tries several different X-Ray exposures to establish which produces the best image of the DUT. The end result looked too dark to me so I used the image 'window' and centre controls to pull out a little more detail. The image is captured as 14 bit depth ADU's and, like many thermal cameras, the user may select the range of ADU's to be displayed on the screen. Very nice and the window function has improved since my earlier DR software (used on my older Faxitron - Bioptics MX-20).

Here are some pics of the new PC internals and the first images captured by the USB Faxitron. The DUT is a Logitech mouse. I have not had to time to experiment with the images or the defect map as its too darned cold in the garage to play at the moment  the images seem nice though so at least this machine is a good one 

Fraser

[Fraser]

Mouse

[Fraser]

More Mouse

[mtdoc]

Cool stuff! 

We have a high quality digital X-ray machine in my office. I wish I new how to operate it so I could go in after hours and play.

[Fraser]

If its correctly managed, it will be disabled with a key lock to prevent unauthorised use. Access is not usually enough 

Mine have their activation keys removed when not in use.

Fraser

[Fraser]

My final challenge on the USB Faxitron is to find the administrator password that unlocks advanced options such as creating a new defect map and changing various exposure defaults etc. No joy finding a likely text string in any of the SR program files yet. I will try the UK Faxitron agent, but not holding out much hope of them helping. The systems administrator is given the password but none of the masses of documentation supplied with my unit has it written down. Pity.

Fraser

[Fraser]

For anyone interested in the USB Faxitron MX-20, I attach some pictures of the rack and the user interface. It is currently roughly assembled for testing and is sat in my garage. It will be installed properly next to my other MX-20 in due course.

The user interface is very different to that of my other MX-20 that is running the earlier DR software. I am not used to it yet but first impressions are that the SR software is less user friendly than the DR version. The SR version is likely more capable when it comes to image manipulation however.

The fact that the SR software fully remote controls the MX-20 settings and carries out auto exposure to achieve the best combination of kV and exposure time for a DUT is a significant improvement though.

The external appearance of the USB Faxitron is very similar to that of the Bioptics based Faxitron but the camera connector is obviously different ion that it is a standard USB connector Vs a dedicated SCSI IV type connector. In use it is clear that the firmware in the MX-20 is different in the USB version. The kV steps are the same but the time is shown in seconds and tenths of seconds as opposed to just seconds on the older unit. This enables the SR software to accurately set the exposure for optimum image.

Sorry the pictures are not more interesting but the unit is actually pretty simple in terms of appearance.

Fraser

[tautech]

For anyone interested in ......

5000+ views 
Please keep entertaining us Fraser. 

Dead hedgehogs, squirrels, birds.....anything.

[Fraser]

Slightly off topic but I have just purchased a Gendex GXS-700 USB digital X-Ray imaging plate. It is not working but may be repairable as the cables are normally the issue with these units. If it has corrupted serial number (another known issue) it may be scrap but I decided to take the risk as it was cheap.

These X-Ray imaging plates are used by Dentists and I have bought the larger No2 size. It is a fully integrated solution that needs no special interface box etc. Basically a CMOS USB camera with some on board processing. The resolution is excellent. These things normally cost $7000 each !  The software to drive the sensor is free from Gendex so it should be possible to use the sensor without any super expensive dental image management software. A TWAIN software interface is also available.

The construction of the unit is very nice. A highly integrated piece of electronics is located inside the thin sensor head. The technology used is not revolutionary. A high resolution scintillator plate produces light when exposed to an X-Ray beam. The light is passed through a fibre optic light channel array and illuminates the pixels of the CMOS imaging chip. The read-out electronics then converts the pixel data to a format that may be sent over USB for processing by the host PC. The sensor electronics also detects when X-Ray is present and triggers a capture event automatically. Auto exposure functionality stops the image capture automatically.

Technical details of the sensor

Sensor size : 31mm x 42mm

Pixels: 19.5um / 1842p x 1324p / 2.4 Megapixels

Resolution : 25.6 lp/mm

Scintillator : Csl

Camera technology : High sensitivity Enhanced CMOS imaging array

Interface to Host : USB 2.0

Power supply : 5V @ 350mA

The anatomy of the sensor may be seen here along with a description of the parts inside the sensor head. Dexis make the GXS-700 for Gendex. The Dexis version is called the Dexis Platinum sensor.

http://www.dexis.com/sensor-anatomy


As can be seen from the attached picture from an ebay auction..... the sensor is usable for imaging PCB's and chips.
If I get the sensor running I will add images to this thread. In case anyone wonders, I already have a working 65kVp Gendex Dental X-Ray generator that may be used with this sensor plate. I also have the non digital Faxitron MX-20 cabinet (spares for my digital camera units) and the Tel-X-ometer X-Ray experimentation platform. These should all provide enough energy to work with the Csl scintillator used in the sensor.

When used with the Gendex 65kVp dental X-Ray generator I will gain a greater penetration capability, but the down side is that it is open site X-Ray so much greater safety precautions are needed, such as an exclusion zone and stand-off to the user (me)

Fraser

[Wolfram]

Excellent thread, lots of nice photos and information.

I have some pictures to add as well. Unlike the other images in this thread, these were taken on photographic film. It's not as convenient as using digital sensors, but it gives good results at a low cost. The X-ray source is a modern dental x-ray head with a 0.7 mm focal spot, driven by a homemade driver with anode voltage regulation. All the following x-rays were taken with an anode voltage of 70 kV, a focal spot to film distance of around half a meter, and an exposure of around 60 mAs.

I had to scale and compress the images a bit to fit within the attachment size limits of the forum. The first image is an ATXMega Xplain card, the second one is a Noritake 2x20 VFD display. I'll post some more images when I get a chance.

[Fraser]

@Wolfram,

Nice pictures 

Had it not been for the availability of the Faxitron units at an affordable price I doubt I would have been able to venture into X-Ray imaging at home.  I used to use imaging plates that contained a scintillator screen that illuminated a monochrome Polaroid instant film. It was much like the old Polaroid instant cameras where the developer is in a sachet at the side of the film and is spread across it as it is ejected from the cartridge. I was considering using one of these units at home but the film became obsolete and had a short shelf life anyway. Resolution wasn't great either.

Dabbling with real X-Ray film or even normal consumer photographic film is not something I was keen to dive into as it can get expensive and messy  I am so pleased to have a digital solution as it makes my life a lot simpler. Your images are great though. Film with a fine grain structure is often superior to digital imaging anyway.

One aspect of the Faxitron Digital imaging system that is worthy of note is its amazing pixel data depth. The MX-20 cameras have an ADU spec of 16000 as they produce 14 bit pixel value data. That is to say, they can capture 16000 different shades (or tones) of grey. The computer monitor and human eye cannot cope with such excellent data depth, but using the software it is possible to view a window of variable width and centring within the captured ADU range.A nominal human eye can only differentiate between 30  720 shades of grey. I normally set it to 8000 ADU's per image as that is plenty of range for most tasks. This is an advantage over film as it is possible to pull out subtle hidden detail that may otherwise be missed.

For those unaware of the term ADU:

ADU -- Analog-to-digital Unit. ADUs are employed as a measurement of pixel value or brightness. Pixel voltages (numbers of electrons) stored during CCD integrations are converted to ADU integers representing the measured voltage compared to maximum (full pixel) voltages in terms of the full Base 2 dynamic range of the CCD system (12 bit = 2^12th, 16 bit = 2^16th, etc.).

There is plenty on CCD ADU's and pixel data bit depths on the internet.

Fraser

[Fraser]

Faxitron and Gendex useful information

For anyone in need of such, I now have the advanced menu password for the Faxitron SR software. This provides access to the Defect Map generator for the Hamamatsu camera array. The utility does everything needed to identify dead pixels etc and create the required map files. Very useful

I also have the firmware serial number reprogramming utility for the Gendex GXS 700 digital dental X-Ray plate. They are known to sometimes forget their serial number, rendering them inoperable.

Fraser

[bktemp]

A nominal human eye can only differentiate between 30 shades of grey !

I doubt that.
Maybe the human eye can identify 30 shades if only a single shade is visible at a time, but if you put the steps next to each other a several 100 steps are necessary to make an even gradient without individual steps beeing distinguishable.
I have no idea how the greyscales in the x-ray images are beeing processed, but normal pictures with 256 greyscales use a non linear scaling: Most displays use a gamma value of 2.2. This means the colour value 1 does not have 1/255 the maximum brightness but (1/255)^2.2=1/196964. So a 8bit greyscale image has a dynamic range of more than 5 orders of magnitude (in an ideal case, most TFT LCD monitors have a dynamic range of around 1000:1). That is why most cameras use a 10, 12, 14 or even 16bit ADC to capture the full dynamic range.

[Fraser]

This was my source for the potentially incorrect 30 shades of grey differentiation comment.........

http://www.popsci.com/humans-can-only-distinguish-between-about-30-shades-gray

Colour is a very different situation as our eyes perform well in that domain as stated in the above reference.

Fraser

[Fraser]

Some Googling on human grey scale abilities lead me to this interesting paper that discusses this very topic of grey scale perception in medical monitor situations........

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3043920/

They detail the complexities of the situation but appear to suggest that a human can detect around 720 shades of grey over a wide luminance range under optimal conditions.

I stand corrected on my previous statement of 30 shades capability of the human eye. The Popular Science posting appears to be inaccurate, unless I misread the article.

Fraser

[Fraser]

My USB Faxitron had a single defective array line that would appear when the higher ADU values were included in an image. I suspected this was an issue with the SR software Defect Map as the file was dated 2007 ! Pretty old as things can change in the array with years of use.

I enabled the advanced menu options this evening and carried out a full warm-up (2 hours), Auto Calibration and then ran the Defect Map generator. The defect mapping utility first carried out a dark field capture, followed by a flat field capture. Defects in the two produced images were identified by the utility and then combined into a single defect map. Out of tolerance pixels or array lines that contain more than 80% defective pixels are captured and dealt with in the image correction algorithms. It is possible for the user to change the acceptable pixel tolerance and percentage of defective pixels in a line that rule the line defective. I used the default settings.

I then took a series of X-Ray images to test the Defect map. I am pleased to see that the one rogue vertical line that used to appear in some wide ADU images is no longer present 

If you own a USB Faxitron that is using an old Defect Map, it may well be worth making a fresh one to capture changes in the imaging array. The old one is saved as a .bak file.

Fraser

[Fraser]

Pictures of the Dark field and Flat field graphs that the Defect Map utility produces. Just for those with an interest in such 

[Fraser]

I am still getting used to the SR software but thought I would upload some of my test images for interest. The SR software seems to provide an image that needs very little tweaking in most cases. It is easy to change the ADU window width and centring to pull out hidden detail as well. It behaves very much like a thermal camera in this respect. Narrow ADU window ranges allow the centre setting to be moved up and down to show only certain material densities on the screen.

I hope the readership enjoy these initial images that I produced tonight. I also include some 'Mystery' devices for you to identify. I know what they are but do you ?

Enjoy 

[Fraser]

More STAMP close ups

[Fraser]

NEC LCD Monitor PCB with 80% JPEG quality setting to fit into this forums file size limit.

[Fraser]

Mystery item No1 

A hint........ I can accurately say that it is old 

The band at the top of the second image is actually PVC insulation tape that is across the front of the item covering some terminals. The material around the central dense item is felt padding.

[Fraser]

And a second Mystery item.

I suspect that this is an easy one for the old boys