Generating x-rays using radio valves, driven in flash mode.
By Leslie Wright
Several people including myself have built x-ray machines, using common radio valves/tubes, driven in cold cathode mode, from a high voltage source, such as a tesla coil, or cockroft walton mutiplier.
Whilst these setups can generate copious amounts of x-rays, the tubes are prone to early failure, mostly from the excessive heat generated.
My first x-ray machine was simply an old radio tube, with a tin-foil band wrapped around the glass envelope, and grounded. The pins were connected to a small 60kV tesla coil.
Of course driving a tube in this way, whilst workable, caused failure of the glass envelope due to heating. Even using high power triodes such as the 6EN4, and using its intenal electrodes (rather than capacitively coupling through the glass envelope) will eventually exceed the tubes thermal limiits, as radio valves are poor emitters of x-rays, and they really have to be pushed to generate a useful amount of radiation.
The solution for this, is really simple, and can provide excellent results (better even than running a tube CW!), and will even work with certain power triodes that have leaded glass!
I was doing a little research on Flash X-ray tubes,and did some patent searches on them. These small tubes, (some are as short as 8 inches and 1 inch diameter) are capable of producing MULTI-MEGAWATT! pulses of x-rays, with energies ranging from 100keV to well into the MeV range!
Such tubes are used for imaging things like very thick cable joints (such as on pylons), and are used by the military to image Nuclear warheads! They are also used to capture short lived events, such as Sabot`s penetrating armour (since the x-ray flash is very short, a few nanoseconds).
Above: a commercial flash x-ray tube.
Below: There are just two electrodes in a flash x-ray tube. a small pointed anode, and a large cathode, with many small fine points on it. When a sufficiently high voltage, and high current pulse is applied to the anode, electrons (e-) are drawn off the fine points of the cathode by field emission, and accelerated toward the pointed anode. The resulting beam (r) is perfectly circular, and very intense, and emerges through a thin metal or glass end window.
To generate the high voltages, at enormously high currents, a Marx-Generator is employed. as shown in the diagram below, each of the capacitors, is charged up in parallel, and discharged in series, once all the spark gaps (SG) fire.
In the diagram shown below, there are only four stages, but many more stages ar possible. My Marx generator has 10 stages at 30kV per stage, giving me 300kV at upto 5 pulses per second! The Erected capacitance of the Marx generator is 470pF, and I find this is about right, in terms of current, if you wish to try to replicate the setup.
Note the final spark gap (SG*) This gap is very important to the succesful operation of flash x-ray tubes. During operation in air, a marx generator, will always produce some corona between each of the spark gaps. This provides a conductive channnel (albeit high resistance) across each of the gaps, which very loosely connects each capacitor in series, even before the gaps fire!, almost like a CW voltage multiplier. The result, is that a voltage of around 60kV is present on the output sphere of the Marx generator. If we were to connect our cold cathode tube, DIRECTLY to the generator, then the 60kV or so, would soon heat the internal components of the tube in a CW fashion, which is what we are trying to avoid.
The purpose therefore, of the final gap, is to ensure we have fast, very steep switching from 0V to 300kV. In my design this consists of two 100mm diameter aluminium spheres placed 35mm apart.
Using common radio valves.
Not everyone is able to get their hands on a flash x-ray tube. They rarely show up on the suplus market ( I am lucky, as I managed to buy one a while ago), so I decided to see if radio valves could be used as flash x-ray tubes.
I have found the following power triodes to work well in pulsed mode from a marx generator, and display very little heating:
PD500
6EN4
A267
The 6EN4 often comes in two types, the older thin walled tubes, and the more modern thick leaded glass version. The leaded version is easy to spot as it is much heavier than its older counterpart, and normally has an x-ray warning stamped on the side.
These tubes are useless for use with an ordinary DC supply, since the leaded glass will attenuate a significant portion of x-rays upto about 60keV, meaning you would have to push the tube to melting poit, to get any appreciable output!
In a Marx driven regime such as this, however, there is almost no difference in output, between the leaded and unleaded versions of the tube, since the energy of the beam is greater than 100keV!
6EN4
6EN4 warning label.
In the first experiment I simply placed a PD500 power triode across the output of my marx generator, and fired it a few times, whilst taking measurements. The tube generated copious amounts of radiation.
The tubes must be well insulated in oil, as the shockwave produced by any flashovers will shatter the glass tube. Marx generators are VERY powerful! Once sunk in oil, the tubes can be operated almost indefinately.
Damaged PD500
Below are a pair of tubes in oil. They are simply glued into plasic cups. The base pins are all connected together, and grounded via the copper wires at the bottom of the picture. The anode top-cap`s protrude slightly above the oil, to allow connection to the Marx Generator.
The cups WILL flash over occasionally, and although alarming, does no harm to the tubes.
To test this assembly, I placed a pneumatic valve assembly, and a piece of circuit board on an Agfa Curix Blue cassette with standard varible contrast paper loaded.
Pneumatic valve. The large aluminium section in the middle is 25mm thick.
The above photo was taken with an A267 radio tube. The exposure time was 90 seconds, at one pulse per second, the distance from film to tube is 300mm. I work exclusively with paper, as large sheets of film, are far too expensive, using film would probably reduce the exposure time down to a few shots!
The aluminium block is rendered almost transparent on this radiograph.
The radiation exits the tube, in all directions, and is most intense, around its middle.
The approximate radiation pattern of the tube, where length = intensity.
As can be seen in the above diagram, the radiation pattern is quite diffuse, which is why the radiograph is so blurred.
This situation could be improved by shielding the tube with thick lead, and leaving an aperture in it, so as to roughly collimate the beam. This would give a much smaller `apparent` focal spot size.
Return to x-ray Pages
Copyright © 2007, Leslie Wright, All Rights Reserved.