What's New

November 26, 2006

I updated the site with some photos of the new production prototype. I made many
improvements to the design, including:

- More iron around coils to improve flux linkage,
- Using bonding magnet wire to make the coils much stiffer,
- Replaced slitted brass firing tube with Garolite one, removing all eddy currents due to tube,
- New sensor board using reflection instead of interruption detection,
- Reduced the number of batteries to eight, moved them all inside the grip,
- Handling battery charging with PIC micro instead of separate IC,
- Improved efficiency of capacitor bank charging circuit,
- Transparent acrylic covers.

Test results for kinetic energy and efficiency will be posted soon, followed by video of me
shooting various items.

I have formed a company, "Spaztronics Corporation". All design, manufacture, marketing
and sales will take place through this entity.

Finally, thank you again to the hundreds of people out there interested enough in this project
to email me. I apologize for being unable to reply to many of these individually.

January 21, 2007

I made a fairly inexpensive yet dramatic change to the pistol design last weekend.

Originally, the design had a mechanical pre-accelerator that stored the energy from the
shooter's trigger finger in a spring to inject the projectile into the firing tube. To get the
initial velocity I wanted (6-8 m/s) either the trigger travel had to be unnaturally long, or else
the trigger gives too much resistance to the trigger finger.

I am replacing it all with a solenoid with a spring-return plunger, triggered by an electrical
switch that is contacted by the trigger. This not only changes it into a purely electromagnetic
gun, it increases the final muzzle velocity as well.

January 27, 2007

A couple of new developments:

1. I decided to ditch the Garolite firing tube and go back to slitted brass tubes. The reduction
in eddy currents (to basically zero, which was great) turns out to be FAR outweighed by:
- Decreased rigidity. Sighting through the barrel in the assembled pistol revealed a rather
curvy path for the projectile.
- Increased thickness of the tube wall. This made a much larger air gap between the projectile
and coil core (0.79 now versus ~0.5 before).
- Increased wall friction.

2. I am writing a program to find the optimum coil parameters for the pistol. The math for
the coil/projectile dynamics were made possible by Bill Slade, who is a mathematical genius.

I am combining his code with genetic algorithms in order to search the coil gun solution-space
(the set of all possible coil guns). Anyone who ever tried to designed a coil gun knows how
many dimensions there are to search, because almost every parameter depends on one or
more other parameters. Trail and error on the test bench can only probe tiny points of the
"solution universe".

The basic idea is that all coil gun parameters (wire gauge, pulse time, number of layers,
detector lead distance, etc) are encoded as genes in an individual "coil organism". This
"coilganism" gives birth to descendants every generation each with slightly modified genes.
The "fitness" of each coilganism is calculated (those yielding high final kinetic energy primarily,
high efficiency could also be selected for), highly fit individuals have more offspring, and their
offspring will also tend to share these good genes. Each of these reproduce and so on. The
technique uses the concepts of inheritance, mutation, selection, recombination, etc to do what
its biological counterparts do; find peaks of fitness (roughly). Random mutations keep
families from getting stuck on local peaks and never finding higher peaks.

I will release the program and my results here when it's finished.

April 24, 2007

Obviously my target product launch has come and passed. I apologize for any
disappointment this has caused. The IGBTs I had been using all along became "obsolete" by
the manufacturer and the "plug in replacements" definitely don't behave exactly the same. I
have had to add some gate clamping components, etc which has added some time.

As a consolation, please check out this beta version of the Coil Gun Simulator! It has some
bugs, but I will be updating it fairly regularly.