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

[Ed.Kloonk]

Q: Have you at all managed to baulk any of the target devices as a result of being under the beam? Any noticeable ill-effects on your subjects?

[Ultrapurple]

@Ed.Kloonk - No, no borkage to date.

My X-Ray machines are relatively low power, low energy. I can't remember precisely the spec of the one I used for this image but it is something like 55kV, 0.5mA and exposures are limited to five or six seconds. I wouldn't put my hand inside (even if I decided to defeat the safety interlocks) but I've not seen any permanent damage to anything.

If I put a running video camera in the path of the 55kV beam then, as expected, there is a veritable snowstorm of noise.

A DSLR at high ISO in the lower power machine (microfocus tube, 35kV max, c 50µA) will also show a lot of noise, but I have no need to do that as the machine is equipped with a roughly 100mm square solid state imager.

I have yet to try putting anything close to the higher power tube aperture. I imagine that the (very much) higher intensity might give some semiconductors something to think about, but I don't know enough about the effects of X-Rays on silicon or germanium devices to make any meaningful guesses about what might happen. (The limited experience I have with bulk material seems to suggest silicon is largely transparent and germanium largely opaque, but there are relatively few germanium ICs available these days...)

Flash memory devices, which I'd intuitively expect to be more susceptible than, say, a power transistor, don't seem to show any problems after casual X-Raying.




Most of the memory card manufacturers warranty their cards as invulnerable to airport-type X-Ray machines, some of which use rather higher energies and powers than my little toys.

[Alex Eisenhut]

P6042 transformer. Did I already post this? 

[Fraser]

Ed.Klondike,

Care is required when exposing Electret Microphones to X-Ray. The exposure reduces the charge on the microphone diaphragm and the result is permanently reduced performance or complete failure. This damage is irreversible by the end user. The damage is relative to the total exposure however so quick tests in a MX-20 may not create noticeable change but is still best avoided if the electret microphone is valuable. For those wondering why this effect occurs, research the old Pen type radiation dosimeters that used a charged filament to detect radiation. The radiation discharges the pre-charge on the filament within the ion chamber.

Some care is also recommended when exposing sensitive memory cells to X-Ray radiation. Corruptions can occur with the memory cells. If in doubt, protect the memory chips with lead foil. Better still, do not X-Ray important memory chips that contain unique, non backed up, data ! X-Rays at sensible levels do not normally harm semiconductor junctions.

Fraser

[jmh]

CV56 magnetron (https://valvecollector.uk/cv56.htm)

CV64 magnetron (https://valvecollector.uk/cv64.htm)

Some packaged triode oscillator, forgot what!

These were rather fun to take though many years ago now. Someone in the local hospital did these in an x-ray suite when the place was empty. A nurse walking through was rather surprised at the devices laying on an x-ray bed!

[Ultrapurple]

I thought the LED display of the 'mystery calculator' warranted closer examination so I put it in my microfocus tube-equipped Faxitron MX20. Orienting the calculator at an oblique angle and giving it 19.9 seconds at 35kV resulted in this fine image. Click to embiggen.



I really like how clearly we can see the vias and the bonding wires from the LED chips to the PCB substrate. And if you look closely at the via to the right of the rightmost digit you can see the ghostly outline of the bubble magnifier. I was also interested to see that the substrate is clearly made to accept nine digits (there are even nine bubble magnifiers) but this calculator and its brethren only used eight. The ninth digit place was unpopulated - not surprisingly, as early LED digits were still quite expensive. (Looking at the substrate layout I guess it's possible the extra position was intended for a different type of LED, for example +).

(added in edit) The subject of this little fun competition used the digit to the left of its most significant digit to indicate "-" for negative values, thus reducing the display precision from 8 digits to 7 for all negative numbers. I believe a later version in the same family used a very similar LED display but with with a "-" in the 'empty' position, thus making the display accuracy the same for positive and negative values alike.

Full res original below.

<edit> I wound an amazing gallery of macro photos of the internals (and externals) of all sorts of semiconductors on Wikipedia by 'Mister RF' - well worth a look. In amongst it was this fine shot of an early 7-segment LED not unlike the ones in this calculator. Click image for full size (2800 x 2640).

[Ultrapurple]

I did my utmost with the Faxitron but this is the best I could achieve in terms of seeing through the metal-lidded chip. Unfortunately, its 35kV isn't enough to see clearly through the (presumably copper or steel) lid. But at least we get a hint of the bond wires and die size.



I suppose I could de-cap the IC, but somehow I can't really bring myself to vandalise a perfectly good vintage calculator that has survived for nigh-on 50 years.

(added in edit) I did try using my 55kV machine but that was also unable to penetrate the IC cap. I think the only non-destructive way I have of seeing anything inside would be to take an image at a very steep slant angle, using the small space between the die and cap. But at a guess the space is only going to be 1mm or so, meaning I'd really be looking at more or less glazing incidence.

Full res image below.

[Ultrapurple]

And another clue, perhaps. This particular copy of the calculator was made in Great Britain. It's a relatively late version, with the keyboard joined to the main PCB by soldered wires. Earlier ones - particularly  those made in the USA - used a plug and socket arrangement, though the connections were in the same places. Other manufacturing sites included Mexico and (I think) Hong Kong.

This version has a SRBP board; many earlier ones were fibreglass.

[Ultrapurple]

I think the excitement has died down so I'll reveal the answer: the calculator is a Rockwell 18R.

In 1973 Rockwell purchased the British company Sumlock Anita Electronics (who designed and, in 1961, launched the world's first all-electronic desktop calculator) and, among other things, used Sumlock Anita's factory in Portsmouth in the south of England to manufacture a number of their calculators for the British and (presumably) European market. The Rockwell 18R seen here was made in Great Britain.

There is a really good simulator of the Rockwell 18R on anita-simulators.org.uk (and many other calculators too).

[Ultrapurple]

<snip> Pen type radiation dosimeters that used a charged filament to detect radiation. The radiation discharges the pre-charge on the filament within the ion chamber.

Hmmm - I have one of those sitting on the desk beside me (I also have the necessary charger). Perhaps I'll put it in the Faxitron for a while and see if it clocks up the exposure.

I did try it in my higher-energy machine a while ago and after a total of about a couple of minutes' irradiation across multiple exposures it did begin to register something. Unfortunately I didn't keep appropriate records so I don't know how many exposures (=total time) it received.

[Ultrapurple]

I have been doing some experiments with trying to extract 'colour' from X-Rays by examining the same subject at a range of different energies (10kV to 35kV in 5kV steps) and then combining the results in various ways.

My basic theory is that if something is transparent at (say) 35kV but opaque at (say) 20kV, subtracting one image from the other will reveal those differences, which can then be highlighted with colours.

So far all I've done is produce a series of more-or-less pretty pictures, but the fact that I am able to get some kind of colour is interesting (at least to me).

Here are some results. The subject was a one-chip USB hub with wired connections, encased in a plastic box.

[Ultrapurple]

Here are some of the source images (alas, not at full resolution due to size constraints, but at least they're a reasonable size and uncompressed PNG). The numbers in the top left are the kV figures.

[Ultrapurple]

I am happy to make the original DICOM files available to anyone who wants them, but as they're a little over 10MB each it won't be possible to post them here (they don't ZIP by any significant amount). Send me a PM and we'll work something out.

[Ultrapurple]

And one more for luck. The 'graph paper' effect is caused by me being too lazy to go through the X-Ray sensor calibration routine before making the images.

[harerod]

I like these! I wish I knew someone like you in central Germany.
Back in the old days it was helpful to be friends with a dentist or any other kind of practitioner who did x-rays. Several of my manufacturers have x-ray, but I doubt that their operators have your level of experience in those matters.

[Ultrapurple]

Several of my manufacturers have x-ray, but I doubt that their operators have your level of experience in those matters.

I expect they are A LOT better than me, particularly in their own specialist professional realms. I like to tinker - something that is rather frowned-upon in most jobs, especially where parts of people are being exposed to ionizing radiation.

I have had a few interesting conversations with hospital radiographers on the occasions I've been in for X-Ray imaging, but of course one has to be a little bit careful. Seen from the viewpoint of someone used to dealing with big, potentially very dangerous, largely unshielded, high-power, medical-diagnostic imaging X-Ray equipment, the idea of someone having an X-Ray machine in their kitchen sounds really scary. And of course it would be extremely foolish for anyone to have an unshielded machine of that type in an uncontrolled, domestic area. Operator training is an essential part of making sire that the equipment can be operated safely. My X-Ray devices, fortunately, are fully shielded cabinet-type machines that are designed for use in a general office or lab environment with no special additional precautions. But I expect the original operators had to have some training - though at a guess this would mainly be on general principles and saying it's a really bad idea to try to defeat the safety interlocks.

[harerod]

...making sire that... <- that is a Freudian, if I ever saw one. *lol*

I was under the impression that working on that kind of equipment is/was your profession?I am pretty sure that access to those devices, even components, is regulated?

I have been designing medical devices (among other things) for decades. The fun stops if you are connected to the business end of one of your machines and the graphs don't look so good. In that particular case it was neuro and cardio diagnostics stuff.

[Ultrapurple]

I was under the impression that working on that kind of equipment is/was your profession?

No. Although I work in technology, my work does not (and never has) directly involved ionizing radiation.

Regulations vary by jurisdiction. I believe for example there is extremely tight control on X-Ray equipment in Australia, whereas here in the UK things are a little more relaxed. That doesn't mean one can go out and irradiate one's neighbours (or oneself) with impunity, just that there are fewer legal impediments to obtaining the equipment, if you can find it. (It may be that there are specific rules for more powerful systems, I don't know).

Some while ago I do recall a former professional radiographer who, after retiring, started making fine art works using X-Ray equipment. I think he got a little too well-known and the regulatory authorities had a little word with him - certainly, the website became a lot more circumspect about the equipment and techniques, and many sample images disappeared. I don't think the authority had any particular problems with this individual's actions - I don't recall any suggestion of him doing anything illegal - but I do think the idea may have been to discourage less well-informed copycats.

I suspect most of us have seen people (usually young men - women are too sensible) taking the magnetrons out of microwave ovens, mounting them on a stick and seeing what happens when they shine the 'magic moonbeams' on things. I have also seen similar things done with X-Ray tubes (though usually in darkened rooms). I wouldn't want to go near an operating magnetron and that goes more than doubly so for X-Ray sources. So mine stay firmly in their shielded containers with all the safety interlocks intact, just as the designers intended. I know more than enough to defeat the interlocks - but I also have a pretty good notion that this would be a Really Bad Idea.

[KalleMp]

Your multi spectral/energy images are beautiful.  If you manage to automate the workflow you would indeed have a powerful analytical tool.

I have been doing some experiments with trying to extract 'colour' from X-Rays by examining the same subject at a range of different energies (10kV to 35kV in 5kV steps) and then combining the results in various ways.

My basic theory is that if something is transparent at (say) 35kV but opaque at (say) 20kV, subtracting one image from the other will reveal those differences, which can then be highlighted with colours.

In very similar ways the airport/security scanners do the same to detect 'organics' in your bags so the border controls can target you for contraband or the security can quiz you about lumps that look like they might be energetic materials.

I expect that many of the latest generation machines may incorporate some forms of backscatter detection as well to further discriminate the material composition as much as possible.

One problem with 100% screening is that there will always be false readings, positives waste time, negatives increase risk due to ignorance and the whole process breeds complacence in the hands of the security apparatus.

I recall pulling some elements of the sensor arrays from an airport scanner (two actually) at a scrap yard 30 years ago and they were small DIL hybrids with a row of  8 or 10 square 0.1" pitch photocells covered with a scintillator so pretty low resolution but one of the units had two detectors stacked with different energy sensitivities I suppose to achieve basic energy discrimination.  So to use the same method you could have a system that slides a shutter with very thin foils of various materials either after the tube or before the sensor to modify the energy spectrum.  Given the low energies the filters would have to be pretty thin but might give interesting results.  Lots of work to dial it in though.

[rickells]

Hi,

 Can you send me any HD image for the Faxitron ?
   Or know who could provide any disk images for these units  ...
Or if they are archived anywhere available for Download ?

thank you