Vacuum Rectifiers X-rays report / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Vacuum Rectifiers X-rays report

first previous 2 3 4 5 next last

Move Thread

LAN_403

radhoo

Wed Jan 19 2011, 09:32AM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

@Arcstarter, thanks, make sure those stabilizers you're planning to use don't have lead glass. I was surprised to see how efficient they are in blocking my 50keV x-rays.

@Proud Mary, There is iron inside, I saw this when doing the neodymium magnets test, not sure if it's the entire anode bell - but I don't think so, since the attraction was weak - probably I was sensing just the steel wires holding the internals together. Do you perhaps have another chart showing the emissions for tungsten? Got something here:

, this sure explains the big difference in choosing the anode.
For inverse polarization, I wonder how quickly the filament will get destroyed after these tests. And I'm guessing the electrons have one single favorite spot.

Time to post some new results, I got the field distribution shape, and it looks very good!

radhoo

Wed Jan 19 2011, 11:22AM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

06.A Field distribution test
Note: A 2X2 Tube in inverse polarization at 50KV. Setup looks like this:

Test1: The Z(vertical) axis:

Here is the last image again, with exposure/brightness increased in software:

Conclusion: As expected, on the vertical axis, the radiation is higher at the same level the anode is (the tungsten filament, under the bell). The bell doesn't make any difference, in fact the radiation is higher where the bell is (it passes straight through).

Test 2: First 360Â° (X/Y) distribution results
The same setup was used and four images of the fluorescent screen where taken with tube at A=0Â° , A=90Â°, A=180Â°, A=270Â°. Distance from the tube was D=6 cm (another test with D=10cm was performed, but with the same results).
It was very important not to move the camera at all , not to change any distances, fluorescent screen position, etc, so the results would be comparable. Here are the 4 shots, with exposure/brightness improvement:

And after Segmentation Threshold was applied (Level=140):

Here's is the tube, at 0Â°,90Â°,180Â°,270Â° in the same positions used to generate the images above:

Conclusions:The field distribution is uniform. The small changes in shape are only caused by the small metal plate, under the bell (easily visible by comparing the tube images with the blue fluorescent images). The source of radiation resulting from the 2X2 tube, should be visualized as a HOT, vertical rod, places inside the Bell. This rod, is , of course, the tungsten filament positioned vertically. The metallic bell produces almost no perturbation on the x-ray escaping it, however the small metal plate, under the bell is a strong blocker (it is not steel! what metal could that be?)
LE: the reason we get less radiation under the bell level is not entirely related to the small metallic plate, but also to the electric field distribution that causes the X-ray emissions: The electrons move from the bell to the inner tungsten filament, and produce x-rays, so we only get them between the bell and the internal filament. The metallic plate only blocks/attenuates a small ammount of rays that are oriented to the bottom of the tube. Most of them however are emitted perpendicular on the filament direction. X-RAY emission is therefore distributed like this:

The blue arrows are the electron paths, and the greens are the photons emitted. The internal filament represented in red is connected to HV+ and the external bell is connected to HV-.

Test 3: Another approach to 360Â° (X/Y) field distribution verification
A simpler test was performed, by surrounding the 2X2 with a tube made out of fluorescent material. Some extra attention was payed to keeping the 2X2 as close a possible to the center of the outer tube:

Segmentation on the fluorescent screen images:

A very nice cylindrical,uniform emission field! The projected band appears to be quite narrow, related to the size of the vertical tungsten filament inside the tube. The picture to the right was added to observe the position of the tube elements as compared to the fluorescence projected on the intensifier screen.

This nice emission field can be used to project good quality radiographs, using this very cheap vacuum tube - the 2X2 .
The only concern is that the emission is radial (cylindrical shape with the center in the tungsten filament) but with some geometry knowledge unaltered, flat projections, can be obtained.

Proud Mary

Wed Jan 19 2011, 11:24AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

X-ray production efficiency is roughly kV x Z x 10E-6, where Z is the Atomic Number - 74 in the case of W.

For an anode voltage less than 69.5 kV, there can be no K-characteristic radiation from a tungsten target, as the incident electron energy must exceed the K-shell binding energy for emission to occur.

1295434567 543 FT106277 Tungsten Spectrum

This paper contains some very interesting observations on W target emission:

Soole BW The Attenuation of X-Radiation Generated at Constant Potentials Sufficient to Excite K-Radiation in a Tungsten Target Phys. Med. Biol., 1971, Vol. 16, No. 3, 427-437

But let's not jump to conclusions that the emission comes from W!

Certainly the heater filaments will be either W, or a W/Mo alloy, but there are other metals besides in the construction of the valve.

Edit: You were posting yours above, while I was writing this.

You have made a very good job of this Radhu. Excellent!

radhoo

Wed Jan 19 2011, 11:39AM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

07.Magnetic field influence
Note: A 2X2 Tube in inverse polarization at 50KV. A very strong Neodymium magnet was placed outside the tube, right next to the bell:

I took three pictures:
1) with the north pole of the magnet set towards the tube
2) without magnet
3) with the south pole of the magnet oriented towards the tube
The results:

Now the threshold (L=150):

In this setup, the magnet produces no visible influence on the emission field.

I am quite pleased of this method of highlighting the results by using image processing on the fluorescent screen shots. It shows a great potential of observing the x-ray phenomenon in more detail. Segmentation can be binary, as used up until now in my reports, but it can use several threshold levels and multiple colors to indicate various radiation levels. The following example shows two segmentation levels of 180 , corresponding to very intense radiation, and 140, corresponding to a decrease in radiation level:

An entire scale of intensities can be built this way with a corresponding color map legend.

Proud Mary

Wed Jan 19 2011, 11:52AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

I feel you should try the magnet experiment again, with the magnets lower down so the magnetic field bisects the open end of the anode bell and the electrode connection leads, where electron field emission is likely to be occurring.

There will be almost no electrons available for interaction with the magnetic field as you show it in the set-up in the picture. The anode bell will shield any electrons flying around inside from the magnetic field.

radhoo

Wed Jan 19 2011, 12:00PM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

I also need to see the angle at which the narrow emission band expands with distance. This one:

To determine that I will use two fluorescent screen cylinders at a few cm apart.

LE: I will also write a little software to do the segmentation automatically and create a map of colors. If I knew the Fluorescent screen fluorescence coefficient (must be something related to the energy of incident x-ray photon and the amount of visible light emitted), I could link it to the light intensity and compute the x-ray energy of the color map areas.The fluorescence surface would indicate the counts. That would a neat x-ray dosimeter. Eg:

Doing the same thing, but using several camera-captured frames, at a well known time interval, allows the time component to jump in the equation, so the Sv/h can be computed for a given color-map area.

Proud Mary

Wed Jan 19 2011, 12:13PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Radhu, if you use the shadow of a metal washer, you can easily plot the beam divergence, by measuring the amount the washer's circular shadow expands with distance from the valve.

You can also assess the quality of the beam by means of a metal gauze or grill with square holes. Distortion of the squares imaged directly relates to field non-linearity.

radhoo

Wed Jan 19 2011, 12:15PM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

Indeed, geometry offers endless possibilities to solving a problem.

Proud Mary

Wed Jan 19 2011, 12:22PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

I added this to my post above whilst you were posting yours:

You can also assess the quality of the beam by means of a metal gauze or grill with square holes. Distortion of the imaged squares directly relates to field non-linearity.

radhoo

Thu Jan 20 2011, 11:13PM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

Unfortunately I couldn't find such a metal piece with square holes. I'll see if I find something else. Some more interesting finds are to follow, meanwhile, here are some radiographs I took:

SDCard:

Mogo Bluetooth mouse ( it has a metallic side):

A piece of coral - blocks the rays completely!

The Kvarts DRSB-01 Dosimeter

As can be seen in the pictures, the optimal position of the tube is at 45 degrees, with the anode towards the fluorescent screen. This way a maximum uniform illumination surface is obtained (see the nice big square green glow with uniform intensity - less shadows). More on this in the next post.

first previous 2 3 4 5 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob