Email John
Email Garett
Last update 02/26/2014

John's Very Large X-ray Tube

Custom Search
since 02/11/2014
UPDATE 02/18/2014 - Tube in action.

UPDATE 02/24/2014 - 2.73 R/second!

UPDATE 02/26/2014 - The HEI ignition coil driver page is up.

This is a very large X-ray tube with a stationary anode, electrode spacing for 200-300kV and a water-cooled anode.  It has a very narrow spot size of about a mm, at least considering the size of the tube.  I believe it to be an imaging tube, though it could be a therapy tube.  If you have info on this device, please email me <----




This shows the overall size of the tube as compared to a calculator.  The tube is well-used as evidenced by the dark brown radiation-damaged glass envelope.


This is the opposite side showing where the glass has been ground thin above the X-ray exit port.


A view of the filament and cathode. Also shows the part number.


The manufacturer's logo.


A close-up of the cathode structure and filament.  The area around the filament is designed to electrostatically focus the beam as a narrow line on the tungsten anode.

This view shows the X-ray exit port on the copper shield and a sliver of the tungsten anode.  The copper shield greatly reduces the amount of radiation emitted in directions not desired.

A better view of the mouth of the anode.  Note how the mouth is rounded.  Even in a vacuum, high voltage surfaces must be smooth to avoid electron emission and arcing.

The anode end.  The white stuff are mineral deposits, evidence that the tube was cooled with non-distilled water.  You will again notice the rounded shape of the connector, designed to cut down on corona stress.

The anode cooling water entrance, again showing signs of mineral deposits.


 I have designed and built an X-ray tube driver based on GM HEI ignition coils. (a page on that project is forthcoming).  Out of oil, the coil is good for about 100kV.  Operated in an oil bath, it is good for several hundred kV, depending on the drive voltage.

This photo shows the treatment the coil's high voltage tower got to make it withstand 100kV.  The neon GTO wire is attached to the tower connector and the area on and around the conductors is doused in Shoe Goop (polyurethane resin).  Then several layers of heat shrink tubing.

The Tek high voltage probe in the background measures the voltage.  The probe is rated to 45kV but in typical Tektronix over-engineered fashion, is showing no distress at 100kV.

At 100kV, little leaders started streaming from the base of the high voltage tower over to the white capacitor.  So, next step: oil.


This photo shows the Very Large Tube set up for testing.  A current limited bench supply (off-screen) is driving the filament.  The current from the HEI coil enters the anode to the right.  The RO-2 ion chamber is set up to measure the output.  It is set on the 50R/hr scale.  The aluminum block is just a scotch to keep the tube from rolling.


Here is the tube in action.  The RO-2 is reading 40R/hr.  I varied the filament current and noticed no change in output so the tube is not emission-limited.  It just wants more juice!

PS: For those of you concerned about my exposure, this run lasted 5 seconds.  My self-reading pocket chamber accumulated no dose at chest level.  The copper shroud around the target does a good job of shielding unwanted radiation.

I will normally operate this tube with a bag of lead shot below and another above it.  Several inches of shielding.  That would not have made a very good photo, though :-)


My new Victoreen Model 570 R meter arrived yesterday.  This is the most accurate radiation measurement device available and was used as a transfer standard until Victoreen went out of business. My old one got stolen so this is a welcome replacement.

The design concept is dirt simple.  The removable probe on the right consists of an air equivalent ion chamber and a capacitor.  The unit consists of a high voltage source and a quartz fiber electrometer.

Operation is simple.  The probe is inserted into the base unit and is charged to about 1000 volts until the quartz fiber indicates 0.  The probe is removed, capped and exposed to radiation for a measured period of time.  Radiation ionizes the air in the chamber which lets a little current flow, discharging the capacitor.

After exposure the probe is re-inserted into the instrument and the quartz fiber indicates how many Roentgens the chamber was exposed to.  Divide that by the time and the result is dose rate.

This arrived just in time because I've finished the design of the driver board.  It is designed to drive 4 ignition coils but I only have 2 on hand at the moment. 

With only two coils working, I operated the driver at line voltage for the first time.  Over 600 watts input at about 90% efficiency.  About 100kV at 6 ma at the tube.

The result was 41 R over a 15 second exposure, an astounding 2.73 R/second!  That's 9,840 R/hr with the chamber an inch away from the exit window. 4 coils should double that amount.

The beam was aimed safely into a concrete block wall backed by dirt.  Dose at my operating position was 1 mR/hr.  I'm not gettin' much nukey :-)

This will probably be the last update on this page.  Shortly I'll have a page up on the driver which is open source.  Bare boards and assembled and tested boards will be available from Fluxeon.

I get the inevitable question, "Why?"  a) because I can :-) and b) because I'm interested in the effects of radiation on materials.  Strange things happen when common materials are exposed to high integrated doses of radiation.  Glass and common table salt turn black.  Salt releases the stored energy as green flashes when heated.

A high power X-ray source is a BUNCH more convenient than a large radioactive source.  It can be turned off and the radiation beam is directional.

Another question is, "Where can I get a tube like that?"  Beats me.  I was gifted that tube by a friend.  Best I can tell it is a 50s vintage radiotherapy tube.  Nowadays accelerators are used for external radiotherapy.  Now that you know what to look for, maybe you'll get lucky.