Needing more than a spark test?

[graham-xrf]

I know myself well enough to recognize when I have invested enough thought, and poked in enough speculative holes, that I must be really interested. Even though it will have to come behind more important things, like the building changes at my place, and have to get along in parallel with my machine stuff, I will try to progress this. I want to check this idea out.

So now I have put in some real money for stuff that could just end up as a cool experiment. There will be something come out of it, even if we discover the scheme has flaws. I expect I will likely have some interesting fun on the way. The best outcome is that between us, we figure out an affordable gadget that many HM members would want to have, and can reasonably put together.

Any tested circuits and proper knowledge can be posted here. Nobody else need embark on any part of it until we know something works - unless they really want to, which I welcome. I am happy to have anyone help me figure out bits of this, because for sure, there is stuff you will know better. If you make something that it works a bit better, or even just looks a bit cooler without compromising function, I am all for it.

Sensors - let's check them out
I agree that when it comes to uncompromising performance, the PMT tubes seem hard to beat. Scoring cheap PMT tubes from eBay could easily result in some great bargains, but could also lead to disappointment. I would not advise a "pre-owned" PMT. By their nature, the photo-voltaic material, especially from older tubes, can be life-limited from a range of causes. "New other" might be OK, but still a risk. I notice many of them come from Ukraine, or Poland.

New PMTs and also a huge variety of solid state X-Ray and other sensors are out there. There are other manufacturers, including some USA firms Start with Hamamatsu.
--> Hamamatsu HERE
Then get closer and check out PMTs
--> Hamamatsu PMTs HERE
--> Avalanche Photodiodes (like solid state PMTs)
--> Spectrometers, FTIR modules, Raman
--> X-Ray sensors
I am sure HM folk can trawl the website and data sheets as well as I can. Move on from Hamamatsu, and at least have a look at other manufacturers. It's OK to use any PMT one can get hold of. The circuits are all similar, and can be designed to be adaptable. Do not fear the high voltage. It is a tiny feeble low current thing. Last time I did one, I made it jump little 6kV sparks to the end of my clutch pencil.

The sensor area - go for bigger
Try for something large - maybe 19mm (3/4") or more. The bigger the front face, the more of the radiation hitting the steel gets used.

I don't know how much it costs, but one can get the whole thing, high voltage and all, in a module.
(See attachment)

Look for the best scintillator
I understand some are hygroscopic. I read that CsI is "slightly hygroscopic". OK, so it takes up water. What does that mean? Does it still work? Will it end up in a pile of dissolved mess (or was that "deliquescent")?
I see there are scintillator materials made of plastics. What?

So far, the first part I am sure of is using the smoke detector innards. That is the easiest, cheapest, most convenient, long lived excitation source I have come across - unless someone knows different.

- - - - - - - - -
Note: I have to hide out from this COVID-19 thing for longer than most. It may limit my options a bit.

 

[homebrewed]

I know myself well enough to recognize when I have invested enough thought, and poked in enough speculative holes, that I must be really interested. Even though it will have to come behind more important things, like the building changes at my place, and have to get along in parallel with my machine stuff, I will try to progress this. I want to check this idea out.

So now I have put in some real money for stuff that could just end up as a cool experiment. There will be something come out of it, even if we discover the scheme has flaws. I expect I will likely have some interesting fun on the way. The best outcome is that between us, we figure out an affordable gadget that many HM members would want to have, and can reasonably put together.

Any tested circuits and proper knowledge can be posted here. Nobody else need embark on any part of it until we know something works - unless they really want to, which I welcome. I am happy to have anyone help me figure out bits of this, because for sure, there is stuff you will know better. If you make something that it works a bit better, or even just looks a bit cooler without compromising function, I am all for it.

Sensors - let's check them out
I agree that when it comes to uncompromising performance, the PMT tubes seem hard to beat. Scoring cheap PMT tubes from eBay could easily result in some great bargains, but could also lead to disappointment. I would not advise a "pre-owned" PMT. By their nature, the photo-voltaic material, especially from older tubes, can be life-limited from a range of causes. "New other" might be OK, but still a risk. I notice many of them come from Ukraine, or Poland.

New PMTs and also a huge variety of solid state X-Ray and other sensors are out there. There are other manufacturers, including some USA firms Start with Hamamatsu.
--> Hamamatsu HERE
Then get closer and check out PMTs
--> Hamamatsu PMTs HERE
--> Avalanche Photodiodes (like solid state PMTs)
--> Spectrometers, FTIR modules, Raman
--> X-Ray sensors
I am sure HM folk can trawl the website and data sheets as well as I can. Move on from Hamamatsu, and at least have a look at other manufacturers. It's OK to use any PMT one can get hold of. The circuits are all similar, and can be designed to be adaptable. Do not fear the high voltage. It is a tiny feeble low current thing. Last time I did one, I made it jump little 6kV sparks to the end of my clutch pencil.

The sensor area - go for bigger
Try for something large - maybe 19mm (3/4") or more. The bigger the front face, the more of the radiation hitting the steel gets used.

I don't know how much it costs, but one can get the whole thing, high voltage and all, in a module.
(See attachment)

Look for the best scintillator
I understand some are hygroscopic. I read that CsI is "slightly hygroscopic". OK, so it takes up water. What does that mean? Does it still work? Will it end up in a pile of dissolved mess (or was that "deliquescent")?
I see there are scintillator materials made of plastics. What?

So far, the first part I am sure of is using the smoke detector innards. That is the easiest, cheapest, most convenient, long lived excitation source I have come across - unless someone knows different.

- - - - - - - - -
Note: I have to hide out from this COVID-19 thing for longer than most. It may limit my options a bit.

Seems like you found the Wikipedia article about CsI. Yes it's disappointing to find it's somewhat hygrosopic. It might be possible to seal it in something like a vacuum-sealed "food saver" bag to get around that. Plastic film are pretty transparent to X-rays so vacuum-sealing the scintillator probably won't incur much of a performance hit. As a side-note, the wizard who ran our WDX system made his own replacement detector windows out of stretched/thinned polypropylene plastic bags. If carefully stretched, the plastic became both thin enough and strong enough to produce very good low-energy X-ray analysis results AND resist the pressure differential between the low-vacuum SEM chamber and detector gas, which was around atmospheric. Must've looked like a balloon, but he got the goods. Like I said, a wizard...too bad he's mentally gone now....

 

[graham-xrf]

Sometimes, with cheap bits, you can make something great.
I consider Dan Berard's achievement outstanding,. He built a homebrew STM (Scanning Tunnelling Microscope) from a piezo noise buzzer - which is one ot the "other" parts found in discarded smoke detector alarms. He managed to image atoms using John Alexander's micro version of a milling machine table.

Cut across the top electrode to divide it into quadrants, then get the tiny X-Y scan motion by flexing the piezo noise-maker using op-amp ramp waveforms. How he manages to make atomically sharp STM tips by cutting stressed wire at a slant, or sharpened by electro-etching it dipped in hydroxide.
--> Home Built STM

Like these graphite atoms, or the 600nm wide bit of gold (that's about 24 millionths inch)



This is something like the tiniest equivalent of of moving a mill table under a dial indicator.

 

[graham-xrf]

Calculation Anomaly?
I need an arithmetic check... and yes - feeling stupid and embarrassed.
I was calculating the wavelengths of what the alloy metals emissions might be, and it messed up.

Example: Sodium
Sodium has the characteristic yellow emission, the two well-known lines at 589nm and 589.6nm


OK - we won't be trying to identify sodium, but this test is just to make my calculation work.
We get the energies from the document in posting #9.
They agree with the values in my old chemistry book (Kaye & Laby).


They are in KeV, so multiply by 1000 to get 1040.98eV and 1071.1eV

We need also:
h = 6.626070E-34 (Planck's Constant)
c = 299792458 (Speed of Light m/sec)
eV = 1.602176634E-19 ( Electronvolt in Joules from electron charge moved through 1 volt)

Wavelength λ = (h * c)/E metres (The Planck-Einstein relation - our main equation)

So I tried h * c = 1.9864458121E-25
Also, get the energy E1 = 1040.98 * eV = 1.66783383246e-16 Joules
and E2 = 1072.10 * eV = 1.71609139268e-16 Joules
Now go for the wavelengths..
h * c = 1.98644581218e-25

λ1 = (h * c)/E1 = 1.19103343E-09 m
and
λ2 = (h * c)/E2 = 1.15754083E-09 m

This is not what I expected! Multiply 1E9 is nanometres. I wanted to see 589nm.

So am I doing something really silly? How does this work? I have to get it right if I want to figure what comes from (say) vanadium, or nickel.
[Edit - OK, we got there. End of post#77]

 

[rwm]

I may be misunderstanding this but I thought spectral lines were related to the outer shell electrons (like chemistry) while characteristic x-rays are related to the K shell electrons. They are therefore completely different energies and properties. Sodium emission lines are visible light wavelength while characteristic x-rays are, well x-rays. The characteristic x-rays of Na would not be the 589nm spectral lines.
Robert

 

[RJSakowski]

Robert has it right; different shells. This may help to clear things up.

Electromagnetic spectrum - Wikipedia

en.wikipedia.org

 

[graham-xrf]

Me too (misunderstanding).
I know that if you toss some salt into the blue propane flame, you get that colour, and that may not be the kind of wavelengths we are talking about when we hit metal (even sodium) with some smoke-detector specials.

I just took the document of shell energies for all those elements (from posting #9) and tried out some, looking for any match-up to any spectra at all. I can only suppose that the calculation is correct, and that the real "colours" coming out are 1.19nm and 1.15nm

Take c = 299792458 (Speed of Light m/sec) and divide by one of those wavelengths - and cringe.
They would be X-Rays - 251petaHertz. X-Rays is considered 0.1nm to 10nm

One transition is an L --> K transition (Kα )
The other is a M --> K transition (Kβ), that is, from M shell right past L, all the way to K
The configuration is 1s22s22p63s1.

The thing is, with only 11 electrons, there are a limited number of ways you can beat the thing up (with smoke detector shots). There are only two close-space strong lines in the spectrum. There are only two energies given. They are close together. That might be confusing me.

Try this:
The bright yellow lines that are seen, even in light from the Sun, are not what we have to deal with. They are visible light, not X-Rays!
If you do enough to knock out a K electron, then one from M will drop in there and take it's place, OR, maybe one from L region will drop in and take it's place. Each transition will deliver some X-Rays.

The yellow light
I think the change in energy involved might correspond to just the energies in the K region, the "outer shell" if you like. The transition between the ground state and one of the spin states 3p3/2 and 3p1/2.

@RJSakowski has replied while I was typing - so I press reply now.

[Edit - I think I got it. 589nm is good for transition from -5.14eV to -3.04eV, corresponding to 3s, and 589.6nm is good for transit to 3p spin state. Who knew I would ever get this deep into quantum mechanics?]

 

[graham-xrf]

Moving swiftly on - we need to be detecting X-Rays. That, I suppose, is what the scintillator material is all about. A frequency down-conversion to tiny flashes that we detect with the photo-electric goodies.

Re: Hygroscopic
The joint to the detector can be sealed up. The outside of the CsI (T) or whatever can be blocked with grease, water-glass, paint, whatever. The front surface has to be exposed, so store it between outings with a cover with dessicant, like the silica gel bags that come with your new micrometer. Maybe there is a "dry it out" procedure. What if one takes the butane torch to it? Bake it in the oven along with the cookies?

Seriously - which X-ray scintillator material do we adopt?

 

[RJSakowski]

Moving swiftly on - we need to be detecting X-Rays. That, I suppose, is what the scintillator material is all about. A frequency down-conversion to tiny flashes that we detect with the photo-electric goodies.

Re: Hygroscopic
The joint to the detector can be sealed up. The outside of the CsI (T) or whatever can be blocked with grease, water-glass, paint, whatever. The front surface has to be exposed, so store it between outings with a cover with dessicant, like the silica gel bags that come with your new micrometer. Maybe there is a "dry it out" procedure. What if one takes the butane torch to it? Bake it in the oven along with the cookies?

Seriously - which X-ray scintillator material do we adopt?

The link I posted in post #64 has a good rundown on various scintillator crystals, pick on Radiation Detectors-Scintillator, scroll down to your selection and pick a scintillator from the menu at the right.

 

[rwm]

I think CsI may be the most widely available. It is used for slightly higher energies in medical imaging. Typically the 140KeV photon of Tc 99m however it will detect lower energies well.
Robert