Mar 062013

Analog computing in the dark!

I got interested in analog computers over the summer and ended up buying a Comdyna GP-6 off of eBay. They still make the GP-10 version and the Comdyna finds application in teaching labs and in control systems engineering. You can read about the history of the Comdyna GP-6 here.

The unit works and as a first computation I patched it up to simulate the Lorenz system.

The Lorenz system of ODEs is given by:

\frac{dx}{dt} = s (y-x)

\frac{dy}{dt} = r x-y-x z

\frac{dz}{dt} = x y - b z

The classic Lorenz butterfly results when s=10, r=28, and b=8/3.
Here’s the resulting program:

The “program” on the Comdyna GP-6.

And here are a few time-exposures of the program output.

The x-z plane of the phase space.

Another view of the x-z plane.

The x-y plane.


 Posted by at 6:01 PM
Dec 192012

Since curve tracers can go to large voltages, you can look at the I-V characteristics of things like neon bulbs. After playing with a bulb for a few minutes, I wanted to see if the photoelectric effect was observable on the curve tracer. Sure enough, you can see the effects on quantum mechanics in a cheap Neon bulb!

Here’s the video:

 Posted by at 8:59 PM
Oct 022012

I picked up this Tektronix 575 Curve tracer at an estate sale this weekend for $50. It seems to basically work, although if you look closely, you’ll see that the out and back I-V sweeps on the 2N2222 IV curve don’t lay on top of each other.

This is the Tek 575 as as found it in a barn. Seventy pounds of awesome!

Restoring this unit will be my introduction to vacuum tube electronics! Like I need a new project!


This looks almost OK, but notice how the up and down voltage sweeps aren’t the same.

Here’s the collector I-V of a transistor as the base current is stepped. The fact that the curves are almost horizontal tells you the collector current isn’t much of a function of collector voltage; it depends mostly on the base current, which is stepped up and down to generate the family of curves.

Here’s the collector I-V curve of a 2N2222 transistor as the base current is stepped. Each I-V curve for a given base current should lay on top of each other.


And finally, I shorted the collector to ground, turned off the base stepping, and set the x-axis to display base volatge (should be constant). I don’t get a vertical line.


Collector shorted to ground via a 1k load resistor. Should be a vertical line. About 2V of collector sweep.

I’ll document my progress getting this unit back into spec. The fact that it basically works makes me hope that it’ll be a relatively straightforward job.

 Posted by at 5:58 PM