An Amateur X-Ray System
Neon tester and Geiger counter.
Demo-type cathode-ray tube and vacuum flask.
Film, film safe, cassette, chems., trays & thermom.
High-voltage system using auto ignition parts.
Simple system with one coil and one ignition box.
Twin ignition coil hookup.
Stun gun parts in box.
The first X-ray.
The fourth X-ray.
The ninth X-ray, using only one coil.
X-rays have fascinated me ever since I first put my foot into a Pedoscope
in a shoe store in the '40s. There were the bones of your toes wiggling
around inside your shoe. A friend would put his hand in the opening so you
could see his hand bones and then you'd do that for him to see your hand.
This project is a high-voltage system that is made from common auto and
electronic parts suitable for making X-rays for hobby applications.
It was fortuitous that Roentgen discovered X-rays at all. By chance, he
had the absolute minimum equipment necessary to create and detect them, and
all his equipment was state-of-the-art at that time. The power supply was a
new and very costly induction coil producing 35,000 volts at 8 pulses per
second. His X-ray source was a Crooke's gas-discharge tube with a very-low-
for-the-time vacuum of 1/10,000 atmospheres. His viewing screen was a barium
platino-cyanide powder which he was using to see the mysterious radiation
coming from gas-discharge tubes and radioactive salts. Many college physics
departments were using similar equipment at the same time. In fact, Crooke
missed discovering X-rays first only because he had wrongly assumed that his
photo plates, which were stored near his equipment while it was working, were
fogged accidentally at the factory before he received them!
The first good human body-part X-ray, of Roentgen's wife's hand, required
a 15-minute exposure. Roentgen received the very first Nobel prize for his
discovery. He also never patented the X-ray system, knowing it was much
too important to waste time going thru the lengthy patenting process. Within
three years, every major medical institution in the world was using X-rays.
They were even used at the front during the War of the Sudan. Early
experimenters used twin X-ray exposures to produce the only true stereo-
scopic (3-D pair) images of fractures ever made, predating 2-D-only CAT scans
by seven decades!
Early amateurs discovered that a Wimshurst machine produces the proper
voltage with adequate current at an affordable price. Periodicals aimed
at the "handy electrician" showed how to determine the polarity of the
sparks from these machines and how to identify a Crooke's tube which would
produce X-rays well. Photographic equipment and film was common at the time.
Newsletters gave how-to information, X-ray exposure times, best films to use
and advertised new, specialized X-ray equipment. Soon companies that made
Crooke's and other gas-discharge tubes started making dedicated X-ray tubes.
We began by looking around for a way to make about 75,000 volts with
1 to 6 mA of current using available parts. Using marble-size metal balls
to get a true measure of voltage by the length of a spark, we found that a
stun gun makes 40,000 V at .3 mA, despite advertising claims. But these
cannot be left on for any length of time. A neon tester probe gives 45,000 V
at an unknown current, but probably in the range of .1 mA. A neon sign
transformer produces 15,000 V at 30 mA, but the arc can be drawn to over an
inch while travelling up a Jacob's Ladder. This translates to 50,000 V and
the current would drop. There is a solid-state induction coil sold by school
science suppliers that claims to produce 65,000 V at .5 mA. The Pedoscope
produced 50,000 V at 3 to 8 mA and released from .5 to 6.0 Roentgens/second.
Dental X-ray machines produce 75,000 V at up to 15 mA during their brief
exposures on the small films used.
Our Geiger counter, used to show radioactivity from mineral samples and
cosmic (gamma) rays, measures up to .01 Roentgens/hr., not Roentgens/sec.
So to test your X-ray equipment output with one, you need to be some
distance away and extrapolate back to the source output. Or use roofer's
lead sheeting as an attenuator and compare your system's output to a
dentist's or vet's X-ray unit output.
An induction coil is a transformer with the primary switched by a set of
contact points, like on an older car's ignition, to put a sharply rising
voltage into the primary. The sharper the rise rate, the higher the induced
voltage out. A stun gun, US patent #4,253,132, is an Oudin coil. It uses about
400 volts from a smaller transformer switched by a spark discharge from a
capacitor that is then sent into a voltage multiplier and capactor. A neon
tester is a combination of a small induction coil and a high-voltage auto-
transformer. (The standard probe on it has a spark gap inside the white
plastic girdle and should be defeated by using a solid wire probe instead.)
We decided to use common car ignition coils for the high-voltage source.
Two coils can be hooked together in anti-parallel to produce 3", hot, loud
sparks. A scary sight to all those who are not used to crazy experiments!
The film was placed between fluorescent "screens" in a "cassette" to
reduce the X-ray exposure time needed down 1/20th to 1/150th of that required
when using film alone. Our cassette is a Kiran blue-sensitive "Hi-Plus".
Fuji RX-B 5" x 7" X-ray film was used. (8" x 10" X-ray film is more common.)
Processing was with common Kodak black-and-white photo chemistry.
The car coils can be switched by a relay driven by a simple 555 timer
circuit and a MOSFET. The relay points need a capacitor across them to reduce
arcing so that the switching will be as abrupt as possible. (The points will
still get rough and need polishing frequently.) Use a .7 to .9 mF, 200 V
cap across the relay contacts. A ballast resistor of .8 ohms is used in
series with MSD 8223 coils. We used a garage battery charger for the
power with a total of 20,000 mF, 35 V caps for ripple filters. But you could
use a car battery instead of a garage charger and all those capacitors.
Twin aftermarket ignition systems can be employed to drive the coils.
We used twin MSD 5900s in parallel. When using the 5900s, no ballast
resistor is needed for the coils. This gave us 90,000 V at an undetermined
current, but the loud, snapping sparks indicated that it was adequate.
A homemade X-ray tube can be made using a shortened aurora tube if a
high enough vaccuum can be pulled. A tube like this and a cathode ray tube
are "cold-cathode" tubes, having no glowing filament to kick out a shower
of electrons, and do not produce X-rays as well as "hot-cathode" type tubes.
It should be stated that when ignition coils are not connected to spark
plugs, the voltage rises unchecked to a maximum and can damage the coil.
After a while one MSD Blaster 3 coil arced thru the center tower to the
brass ground terminal, carbonizing the plastic and ruining the coil. So
the good coil was immersed in mineral (baby) oil with just the tip 1" above
the oil. This stopped any wasteful corona and carbonizing. Remove both nuts,
solder wires to the top of the brass terminals and then lower the coil into a
glass that is about 2 & 1/2" wide inside and about 6 & 1/2" deep. Acetone
dissolves spilled oil. The MSD Blaster 3 (8223) coils were much better than
other MSD coils and better than two types of Accel coils.
The spark between the coil tips is brighter at the positive tip. This
tip should go to the X-ray tube anode. But the system is not pure DC, and the
X-ray tube itself acts like a vacuum-tube diode and rectifies the AC.
An old RAD-1 tube used with twin auto coils in open air and 6 volts going
to the filament took a 10 minute exposure to give us our first X-ray. The
image shows a tungsten strip at the top, a molybdenum ore-assay boat next,
then a tungsten wafer, then a lead strip and finally a wood block with a
steel nail, a steel brad and a brass brad in it. There are two holes drilled
in the wood block.
Other X-rays we did show that X-ray exposure time and development time
are critical to imaging what it is you wish to examine inside the items
X-rayed. Our fourth X-ray shows a skyrocket, tape measure, lighter and can
opener. But a shorter exposure time showed the skyrocket motor's internal
structure better. (See the Homemade Fireworks page.)
The neon tester connected to the cathode-ray tube (a cold-cathode tube)
with one side grounded to a water pipe required 15 minutes to produce a
very dim second and third X-ray. A skyrocket, a small Crescent wrench and
a cigarette lighter were placed on the film cassette. But only the outline
of the adjustable wrench is obvious.
The neon tester used with the old RAD-1 tube with filament current going to
it and with the anode grounded to a water pipe produced a light but quite
useable X-ray in 6 minutes.
Using only one 8223 coil immersed in mineral oil, the twin 5900s and with
the RAD-1 filament on produced a good X-ray in 3 min. The ignition coil tip
goes to the RAD-1 cathode and the anode was grounded to the battery charger
negative clamp. The X-ray shows a computer mouse, a pen and a photoelectric
night light. The pulse timing was 40/sec. An exposure of only 20 seconds
yielded a good X-ray of the rocket motor interior.
Using one 8223 coil in oil and the NAPA AR-108 relay with .9 mF across
the points produced a good X-ray after 10 minutes. That was with the RAD-1
filament on and using fresh D-cells. The .8 ohm ballast resistor must be
used when not using the ignition boxes. The relay points took a beating
after just one X-ray and needed to be filed and sanded smooth again.
When using one 5900, the MOSFET signal can go directly to the RPM sensor
inputs, violet to the MOSFET drain and green to the +9 volt bus. Of course,
with this hookup, the frequency can be set higher than 40 pulses/sec.
The best system at reasonable cost consists of a dental X-ray tube
(Toshiba D-088 or D-101, for instance), one Blaster 3 coil, one 5900
ignition box and no relay on the timer/pulser board. The white wire,
which is only for older cars which have contact points, is not used.
NOTES: X-ray film scratches easily when in the solutions, so use care.
Touch them only around the edges when wet. (OK to touch when dry)
All X-rays were taken by plugging in a long extension cord in
another room to power up the battery charger and start the exposure.
The 9-volt pulser battery and the 6-volt D-cell filament batteries
were connected manually before leaving the exposure room.
MSD Blaster 3 coil #8223
MSD Blaster Ignition #5900
Relay: NAPA AR-108 (if used instead of the 5900)
MOSFET: Radio Shack 276-2072 (IRF 510)
Caps for filtering: Radio Shack 272-1022 (4)
See eBay for used X-ray tubes, also called "inserts". (email@example.com)
Geiger counter: Frey 15598209 (optional)
Neon tester: Frey 15583518 (optional)
Induction coil: Frey 15583515 (optional) replaces timer/pulser, coil & 5900
Kiran Hi-Plus cassette, blue-sensitive (or equivalent) 5 x 7 or 8 x 10
Fuji RX-B film (or equivalent) 5 x 7 or 8 x 10
Kodak Hobby-Pac, 3 trays and safelight (optional: thermometer, paper safe)
PHOTO NOTES: Mix one pint of solutions at a time. In safelight illumination,
open film box and take out one sheet of film at a time from
black plastic bag. (end sheets are cardboard) Seal rest of
film back in box. Put film in cassette and close securely.
Light can now be turned on. After each X-ray exposure, turn
safelight on and light out. Remove exposed film and place in
the paper safe. (Or an old film bag.) Repeat for the next
X-ray. When ready to process film, pour solutions in trays.
Turn light out and safelight on. Use one hand to slip film
into and out of developer one at a time for the required time.
Use the other hand for stop bath and fixer. Keep a paper towel
handy to wipe the solutions off your hands frequently.
NEVER splash stop bath or fixer hand in developer! After sliding
films into solution so they are submerged, agitate the tray
gently by rocking. Times are 5 min. in developer, 10 sec. in
stop bath, 2 to 5 min. in fixer and 30 minutes washing in
running water. Lights can be turned on when films are in
fixer. Stand films on edge on paper towel angled with one
side touching a back rest to dry the films.