First Analog Simulation (publ. 2025-02-04)

I completed enough components on my analog computer to be able to wire up the Damped Oscillation simulation from Analog Thing. My components are based on the Grappendorf analog computer schematics, since I cannot afford to purchase an Analog Thing computer at this time. Here is a photo of the computer as it currently stands, with wires hooked up for the simulation:

my analog computer wired up for the damped oscillator simulation

The breadboard currently has

- 2x integrators

- 2x inverters

- 1x adder

- 3x coefficient potentiometers

A few notes on the adventure of building this:

- I had a lot of trouble with the mechanical switch toward the top not making good contact with the breadboard, but I was able to get that worked out. I then fixed the switch in place with a glue gun.

- Thankfully, I was able to get one of those breadboard jumper wire kits, with thinner wires of various lengths. These are easier to work with than the 18 gauge wire I had.

- I realized, when I went to tie the components together for the simulation, that I have one more inverter than is necessary.

- I'm using the blue screw terminal blocks for now to tie together the inputs and outputs. Am hoping to be using a patch panel of some sort in the future.

For running the simulation, at first I tried an initial condition of only 1 volt. But there is bad noise from my power supply of around 500 mV p-p, so I could see this was going to make for some ugly screenshots:

early output with 1 volt initial condition

So I brought up the initial condition voltage to 10 volts, where the noise is still there but difficult to see on the scope.

Here is a baseline simulation, with the spring coefficient (stiffness), damper coefficient (friction or resistance), and mass coefficient all set to about midway range.

base simulation screenshot

Something I was wondering about from the beginning, was whether or not I should be setting some specific initial condition for the second integrator. I decided to just attach it to GND (0V). Hopefully that doesn't have some weird effect on the results. Anyway, here are outputs for various adjustments to the coefficients. In each case, I would make sure the other two coefficients were set to about mid-way on their range.

The oscilloscope is hooked up to the output of the inverter, which I believe represents the velocity of the system.

low spring coefficient (stiffness)

high spring coefficient (stiffness)

low mass

high mass

low damper coefficient (friction or resistance)

high damper coefficient (friction or resistance)

Maybe I can find a spring at some point and see if these results seem consistent with reality. But I don't have one handy at the moment.

I hope at some point to set up a patch panel along with large, mounted potentiometers. I think the multiple IC/RUN switches could be replaced by one switch and some solid state relays. I would like to make a multiplier component also, but the AD633 chips are running at about $20/each right now and I am on a tight budget this pay period.

I'll mention again that I started channel #analogcomputing@irc.libera.chat, hoping that I would be able to interact with others who are interested in or playing around with analog computing.

Copyright

CC BY-SA 4.0 Deed