How to produce X-rays: DY802

How to produce X-rays: DY802 tubes

Subsection DY802

This part focuses on rectifier tubes (e.g. DY802), which I tried first. Rectifier tubes are cheap and relatively easily available, but have disadvantages: They are not designed for high power dissipation, and most don't take very high voltages. Rectifier tubes are operated in cold cathode mode with fixed operating voltage.

The setup I used:

The above picture shows the general setup I normally use. From left to right: 8kVss transfomrmer, 9-stage cascade, big metal ball on isolating pillar to prevent spraying. Then, from background to foreground, the tube together with a 500MOhm resistor (both in the white pipe). Not shown (or not visible) here are the instruments to measure current (a must) and voltage across the tube.

And these are the tubes I use. The upper one (type DY802) is much more effective and mostly used. See also the diagram of the DY802 type.

More details and explanations:

As already mentioned in the introductory part, the filament is not heated for our purposes (actually, in my first tube, I blew it up in an earlier experiment :-). The tube is operated in a mode that's called "cold cathode discharge", and which is normally unwanted, as the tube loses part of its rectifying characteristic then. The onset voltage of this cold cathode discharge is much higher (40kV) when the cap is negative (i.e. reverse polarity, as it's normally positive) than when the filament is negative. To make the tube work reliably, the following points habe to be considered:

An obvious problem arises from the fact that the tube was not made for 40kV. I had arc-over at 30kV. This doesn't at once damage the tube, but it lowers the voltage :-) I soldered longer wire on both the cathode and the anode and glued PVC tubes to both ends of the tube. Pour a lot of epoxy glue into the PVC tubes, this also reduces mechanical stress on the tube contacts (long stiff copper wire = long lever).
I found that from a certain onset voltage onwards, current increases very rapidly with the voltage, and is only eventually limited by your source. Have a look at the so-called characteristic of the tube. It resembles that of a Zener-Diode very much. Once the onset voltage has been reached, it remains nearly constant over a wide range of current. This means that you won't get much further than this onset voltage, even at considerably high current. The tube has a more or less fixed operation voltage.
This implies a problem: just as a Zener diode, the tube can not be operated stable on a constant voltage source. It has to be driven with a constant current source, or over a big resistor from a usual voltage source. The voltage of this source and the resistance then determine the current, as the voltage drop over the tube remains constant. Example: My tube has an onset voltage of 48kV. I operate it over a 500MOhm resistor from a 75kV source. This results in (75kV-48kV)/500KOhm = 54uA. As this is a bit too high, I usually add another 200MOhm, resulting in 39uA.
Also, the tube won't handle arbitrary power. Some watts are enough. More will heat up the tube considerably and eventually destroy it. E.g. I overheated my first tube with the result that the remains of the filament bent and made an internal short. From my experience, 40uA at 50kV (=2 watts) is just enough!

Technical details: 1st update 10 March 1998

40uA is probably not just enough but more than enough. I seem to have destroyed one of my best tubes using it at his power level. It either outgased or leaked. Fact is that I cannot reach any useful operating voltage (below 20kV) and that instead there is a visible gas discharge inside. I am now using another DY 802 tube at 45kV/10uA. At this level, it gets only slightly warm.
I also redesigned hv-proof incasing. I use wider plastic tubing that fits well over the glass tube, and I stick them together more towards the middle. Several layers of adhesive tape fill the gap. The reason is that I had problems with arcover when I had the lead shield on the tube. Discharges went through the expoxy glue at one end of the tube into the lead shield, and back through the epoxy at the other end. The plastic tube is not so easily broken through.
I measured the operation voltages of several types of tubes. As I couldn't read the type number on some of them, I give a sketch (cross section) of the electrode geometry instead. Several voltage values mean different tubes of the same type. Operation voltage was defined as the voltage at 10uA current.

Type ??? ??? DY 802 GY501

Voltage/kV <25
30 <25 28
30
32
45
30-60
(larger
than
DY802)

Shielding

To operate the tube safely, it has to be shielded, or you have to go away quite a bit. The measurement with a commercial dosemeter showed, that a safe distance in the legal sense would be 3m or more!

For an effective shield, use lead 1mm thick (as used by roofers). This is enough to reduce the radiation below normal background. Of course, you've got to leave a hole to let some radiation through :-) Just don't forget that

a hole of 6mm diameter a few mm above the surface of the tube leaves radiation into a cone of 30 degrees opening angle - broad enough to accidently hit you!
there is also radiation coming out along or near the axis of the tube. A simple cylinder will let this through!

The following two pictures show the shields I made (lovely sunset light on the left one :-).

The three cylinders leave a gap to let air circulate (and cool the tube), but due to the slightly conical form prevent radiation coming out through this gap. The radiation coming exactly along the axis is absorbed by the wax-filled PVC-tubes. The narrow pipe is used to make a really narrow beam.

[Back to "Technical"] [Back to Index] [Other links]

Jochen Kronjaeger
Kronjaeg@stud-mailer.uni-marburg.de


Type	???	???	DY 802	GY501
Voltage/kV	<25 30	<25	28 30 32 45	30-60 (larger than DY802)