Author Topic: Home Brew Analog Computer System  (Read 105800 times)

0 Members and 1 Guest are viewing this topic.

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #250 on: April 29, 2014, 02:09:56 pm »
Thanks  :D

Here is the schematic for the R.O.M. board.
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline baljemmett

  • Supporter
  • ****
  • Posts: 666
  • Country: gb
Re: Home Brew Analog Computer System
« Reply #251 on: April 29, 2014, 02:48:22 pm »
Things have been a little hectic for me lately, but this evening I finally managed to get the (almost complete) Fourier synthesis character generator/display unit up and running.

That's some lovely work, GK - thanks for sharing!
 

Offline robrenz

  • Super Contributor
  • ***
  • Posts: 3035
  • Country: us
  • Real Machinist, Wannabe EE
Re: Home Brew Analog Computer System
« Reply #252 on: April 29, 2014, 03:10:17 pm »
Things have been a little hectic for me lately, but this evening I finally managed to get the (almost complete) Fourier synthesis character generator/display unit up and running.

That's some lovely work, GK - thanks for sharing!

+1 :-+

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #253 on: April 30, 2014, 01:01:27 pm »
Well, getting a little more serious now, this evening I removed the programming toggle switches, plugged a PIC16F874 into a bread board and wired it to the units programming port. A few quick test routines written for the '874 acting as the "remote" programmer and here is the result:



OK, I'm using a uC to program the display, but the character generator itself is 100% analogue and discrete digital logic (74HC' & 74HCT plus the SRAM).
Programming is completely asynchronous. You just put the data onto the 12-bit port and strobe the write pin. My simple discrete logic programming interface works flawlessly, transferring the asynchronously-received and latched data to the operating screen RAM during the blanking intervals without a glitch. Yahoo.

 
« Last Edit: May 01, 2014, 08:50:35 am by GK »
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline mamalala

  • Supporter
  • ****
  • Posts: 777
  • Country: de
Re: Home Brew Analog Computer System
« Reply #254 on: April 30, 2014, 03:06:15 pm »
Fantastic stuff! A big thumbs up!

Greetings,

Chris
 

Offline robrenz

  • Super Contributor
  • ***
  • Posts: 3035
  • Country: us
  • Real Machinist, Wannabe EE
Re: Home Brew Analog Computer System
« Reply #255 on: April 30, 2014, 03:35:43 pm »
Your anti screen burn orbit is very nice ;D

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #256 on: May 01, 2014, 08:49:40 am »
Your anti screen burn orbit is very nice ;D


I think I might still slow it down to make it less obvious. I've got a half dozen NOS CRT's as used in the Tek 422 oscilloscope, still in their original packaging, just for this kind of thing. Think I'll make a dedicated CRT display unit for this instead of utilizing a 'scope.
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline johnwa

  • Regular Contributor
  • *
  • Posts: 237
  • Country: au
    • loopgain.net - a few of my projects
Re: Home Brew Analog Computer System
« Reply #257 on: May 04, 2014, 12:13:23 pm »
Hi GK,

That is some really nice work with the Fourier character display! Producing that display using analogue circuitry looks impossible at first glance, but we can all see that it is quite practical. The smooth curves give the characters a certain aesthetic appeal, too.

Now, I was thinking about how this works, and also about the Rossler experiment last year, and I wondered: would it be possible to apply this technique to the synthesis of three dimensional objects? After a certain degree of head scratching, it appears that the maths works out, though I think building a physical implementation will be on the edge of practicality, due to the overall complexity required.

For a 5th order approximation, the character generator requires ten frequencies plus a DC term. With these, it is possible to parameterise a curve as a function of one variable (s), in two dimensions, with a total of 22 coefficients.

A 3D surface can be parameterised as a function of two variables (u and v). Therefore, instead of five frequencies, 25 will be required for the same order synthesis (5 x u x 5 x v). Allowing for the sin, cos, and DC components, a total of 121 terms would need to be synthesised. To cater for the three axes, up to 363 coefficients might be required (although many shapes would only need a relatively sparse coefficient matrix).

The coefficients can be implemented fairly simply using resistors, so the main obstacle to using this technique would appear to be the generation of the signals at all of the different frequencies and phases. To generate them directly would require 100 analogue multipliers. It might be feasible to mix all the u harmonics and all of the v harmonics, multiply the sums, and then filter off the individual (m*u * n*v) components, but I suspect that this would involve a similar level of complexity.

Ideally, the low frequency parameter would change in discrete steps, while the high frequency parameter swept continuously, to give a rectangular U-V grid. However, I think it would be simpler if each parameter changed continuously. This would result in a slightly skewed U-V grid, similar to a television raster scan. It would also be desirable to be able to switch the U and V signals, in order to give grid lines in both directions.

I don't have any plans to actually try to build this - it would be huge! However, if anyone is crazy enough to give it a go, I might be able to offer some suggestions as to how to go about it.

I have attached a couple of Matlab simulations of this method. Note that it is not limited to solids of revolution, though these are the only ones that I have managed to figure out the coefficients for so far  :)
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #258 on: May 06, 2014, 09:02:40 am »
That looks neat. Another approach might be to use spherical polar coordinates and generate objects by just synthesizing the three signals for the radial distance, azimuth and polar angles. It's easy enough to transform the spherical polar coordinates to xyz Cartesian values, and then to 2-D X-Y projective views.   

I am considering building a spherical polar coordinate to xyz Cartesian coordinate transformation unit to complement the function to my already-built 3-D projective unit. However the problem is finding an analog computing application for it. I'm not aware at the moment of any physical system analogs that provide three variables that are readily plotted and unambiguously readable in spherical polar coordinates. 

I have my polar coordinate oscilloscope adapter up and running and am planning to build another, more powerful MK II version. The current version can only generate plots over a polar angle of 2pi radians. The unit I am currently planning will be able to do 0.5 to 1024 pi radians in 0.5 pi radian steps.
   
« Last Edit: May 06, 2014, 12:33:34 pm by GK »
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline Clocky

  • Newbie
  • Posts: 3
Re: Home Brew Analog Computer System
« Reply #259 on: May 06, 2014, 02:59:37 pm »
That's really cool, I love this kind of stuff.
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #260 on: May 09, 2014, 02:16:06 pm »
Well then, here is a little more  ;D

The polar rose is a mathematical curve easily generated by my completed MK I polar coordinate oscilloscope adaptor.
Here is a demo of a pair of 2 pi radian roses. The first is generated with a swept frequency sine function and the second with a triangle wave.

« Last Edit: May 09, 2014, 02:21:38 pm by GK »
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline johnwa

  • Regular Contributor
  • *
  • Posts: 237
  • Country: au
    • loopgain.net - a few of my projects
Re: Home Brew Analog Computer System
« Reply #261 on: May 12, 2014, 12:25:01 pm »
That looks neat. Another approach might be to use spherical polar coordinates and generate objects by just synthesizing the three signals for the radial distance, azimuth and polar angles. It's easy enough to transform the spherical polar coordinates to xyz Cartesian values, and then to 2-D X-Y projective views.   


Yes, depending on the object, it might be possible to get better definition with fewer terms using polar coordinates. Although objects that deviate significantly from a spherical shape might still be better of in cartesian.

Anyway, this was just a random idea I had, I don't think I will take it any further at this stage - too much other stuff to do  :(.

Your rose figures are pretty neat too. I think there is a vaguely similar process for manufacturing faceted shapes on a lathe. I think it uses planetary gearing of the spindle, to give the part a complex motion. See, e.g. http://en.wikipedia.org/wiki/Hypotrochoid Obviously, the lobes of the pentagram get chopped off, and only the centre section remains. I imagine it would be fairly easy to plot these figures on your scope GK, though it might take a bit of work to get the equations in to the correct form.
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #262 on: May 14, 2014, 02:06:15 pm »
That Hypotrochoid looks interesting, but at a glance it looks like it needs 6 pi radians of rotation to be plotted in full. Unfortunately my MK I polar coordinate adapter is limited to 2 pi radians. While consumed in the technical side of the design I didn't realize how much of a limitation a fixed "sweep" of only 2 pi radians (360 degrees) for the angular coordinate would be; that came afterwards, reading up on geometric curves and finding applications for the completed unit. The Hypotrochoid will have to wait for the MK II version.

However that will be a while off. ATM I'm currently struggling in my free hours here and there to complete the construction of my character generator and write it up on my webpage. I need to clear to bench to get onto others things that need attention.

All that's left is installing the power supply, some fabrication of panels and screwing the case together. I also designed an additional, 5th PCB, giving LED status indication to the 12-bit parallel programming port (green for the 7 address lines, yellow for the 5 data lines and a single monostable-stretched red LED for the write strobe line) . This 2-relay-rack analog/hybrid computer is going to have plenty blinky LEDs to make it look really technical ;D



« Last Edit: May 14, 2014, 02:12:12 pm by GK »
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #263 on: May 24, 2014, 01:54:48 am »
Here is a particular example of dither (these days known as triangular probability density function) that predates digital audio and video processing by a few years. I am currently applying the technique to my diode function generator design and it works well, turning the sharp diode discontinuities parabolic.

This page is from my earliest analog computing text out of several that summarizes the technique: Analog computing techniques, Johnson, McGraw Hill, 1956. None of my old texts actually use the word "dither", however. Not sure when that term was first applied. Dithering techniques go back even further than this, applied to mechanical analog computers by means of vibrator motors to compensate for quantization errors incurred by gear sticking and backlash.

The reference cited (8) is: Stone, J. J., Jr.: Smooth-curve Function Generation Using Diodes, IRE Airborne Electronics Digest, 1954.

Does anyone out there know of any earlier examples of the application of electronic dithering?

« Last Edit: May 24, 2014, 03:17:47 am by GK »
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #264 on: August 16, 2014, 07:22:40 am »
All that's left is installing the power supply, some fabrication of panels and screwing the case together. I also designed an additional, 5th PCB, giving LED status indication to the 12-bit parallel programming port (green for the 7 address lines, yellow for the 5 data lines and a single monostable-stretched red LED for the write strobe line) . This 2-relay-rack analog/hybrid computer is going to have plenty blinky LEDs to make it look really technical ;D


I finally got this thing finished and written up:

http://www.glensstuff.com/fouriersynthchargen/fouriersynthchargen.htm
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #265 on: January 19, 2015, 01:09:00 pm »
After many other diversions I'm back to allocating time to this ongoing project again. I've finally competed all 30 of the programmable integrator modules and have them attached to their brackets and ready to be chassis mounted. Should get them chassis-mounted this coming long weekend, and at least partially wired. As pictured earlier in this thread, I've had the panels drilled and the 366 banana jacks installed for over a year already. Also have the sine/cosine circuit boards completed and an order for another ~ 1000 4mm banana jacks has recently arrived, so I have some further panel drilling to get on with...... The (twenty) Summing and (ten) Logarithmic amplifier PCBs are currently underway.....

The integrator board assembly was a little tedious as I had to pad the "timing" capacitors to the correct (measured) values. There are two selectable timing capacitance values. 10nF is permanently wired into the integrator and an additional 990nF (for a total of 1uF) is switched in by a reed relay. In each example I got the 990nF to within 1nF and the 10nF within 100pF. For low dielectric absorption I used polypropylene for the bulk of the capacitance (at least 820nF and 8n2 respectively) and padded with COG ceramic for the 100's of pF and polyester for the 10's of nF. Low value (nF's) polypropylene capacitors only seem to be available in very large packages with high voltage ratings (like 2kV and 1" lead pitch) and are expensive. The polyester was the most cost effective option. Back in the day the big analog computer manufacturers had custom polystyrene caps made to uF values with fraction of a percent tolerances - somewhat beyond the DIY budget.   

The low-leakage summing junction clamp/protection diodes at the JFET0switched inputs are heat shrink covered to keep out the light as discussed here: https://www.eevblog.com/forum/projects/glass-diode-photoelectric-effect/
 
« Last Edit: January 19, 2015, 01:17:40 pm by GK »
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 14546
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Home Brew Analog Computer System
« Reply #266 on: January 19, 2015, 10:55:36 pm »
Wow!
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Dave Turner

  • Frequent Contributor
  • **
  • Posts: 439
  • Country: gb
Re: Home Brew Analog Computer System
« Reply #267 on: January 21, 2015, 12:06:47 am »
Have you estimated the total power requirement of the finished product? Apologies if this has already been answered.
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #268 on: January 21, 2015, 12:29:05 am »
The computer is being built in modular units, each with it's own self-contained, regulated power supply. Each unit, housed in either a 2U or a 3U case will contain, in general, ten identical computing units - a computing unit being either an integrator, a summing amplifier, a multiplier, a comparator, a logarithmic amplifier, etc.

The "integrator chassis", for example, of which there are three, for 30 computing nits in total, is a 3U case. Power consumption per integrator chassis is about 40W.
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline Dave Turner

  • Frequent Contributor
  • **
  • Posts: 439
  • Country: gb
Re: Home Brew Analog Computer System
« Reply #269 on: January 21, 2015, 05:02:03 pm »
I, for one, eagerly await the completion of this project or at least it being sufficiently complete to demonstrate with a video or two.
 

Offline Alex Eisenhut

  • Super Contributor
  • ***
  • Posts: 2146
  • Country: ca
  • If you can buy it for 4$ on eBay, why design it?*
Re: Home Brew Analog Computer System
« Reply #270 on: January 21, 2015, 05:14:15 pm »
If you could build a Scanimate, that would be awesome.
*Except AC/DC adapters on eBay. Avoid them all!
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #271 on: January 28, 2015, 12:04:31 pm »
An integrator chassis slowly coming together. The banana jack functions are as follows:

White: initial condition (I.C.) input
Yellow: x1 input
Blue: x10 input
Red: signal output
Green: earth

The orange LED below each bank of integrator outputs is for individual overload/over-range indication. Each overload indicator output also feeds an internal alarm bus. 

The brown and orange jack pair are for external logic control signals. The integrators will be ordinarily under the control of the computers timing/control unit, but a selector switch in the front panel will alternately route the control signals from these two banana jacks. There are 3 modes of operation: 1) Initial condition - the integrator is preset to the (inverted) DC potential applied to the I.C. input. 2) Compute/Run - the voltages applied to the summed x1 and x10 inputs are integrated. 3) Hold - operation is paused/halted and the integrators hold their value.

There will be three other selector switches on the front panel. One will be "Pot(entiometer) Set On/Off", which is used when setting up a "program" to assist the setting of the computers coefficient potentiometers. Another switch will be the time constant select, activating the timing capacitance reed relays of all integrators in parallel to select either the 10nF integration capacitance or the 1uF integration capacitance. The last switch will be a 10 position rotary switch used to select which one of the integrator signal outputs to route to the chassis' metering bus. By means of a selector which matrix, every single "computing module" signal output will be selectable for voltage measurement on the computers built-in 5-digit voltmeter (there will be a separate chassis associated with the timing/control chassis called the "metering chassis").





« Last Edit: January 28, 2015, 12:12:52 pm by GK »
Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline GK

  • Super Contributor
  • ***
  • Posts: 2607
  • Country: au
Re: Home Brew Analog Computer System
« Reply #272 on: February 01, 2015, 10:25:34 am »
Getting closer........

Bzzzzt. No longer care, over this forum shit.........ZZzzzzzzzzzzzzzzz
 

Offline netdudeuk

  • Frequent Contributor
  • **
  • Posts: 408
  • Country: gb
Re: Home Brew Analog Computer System
« Reply #273 on: February 01, 2015, 02:43:45 pm »
Looks amazing.  It wouldn't even matter if it didn't work  :)
 

Offline peter.mcnair

  • Contributor
  • Posts: 14
  • Country: gb
    • analog-ontology
Re: Home Brew Analog Computer System
« Reply #274 on: February 01, 2015, 03:45:00 pm »
Excellent!

From one analog computer builder to another:

Beyond Rössler...

dx/dt = - ax - cos(by),

dy/dt = -ay - cos(bz)

dz/dt = -az - cos(bx) .

I just finished my computer yesterday - well almost - still needs a back panel - and maybe a fan (or two...).

Given that I can now do trigonometric functions, I tried patching up the above equations...it's awesome - the results looks like soldering iron fumes!

(details in http://analog-ontology.blogspot.co.uk/2015/02/beyond-rossler.html)


Above equations from:

Experimental Study of the Sampled Labyrinth Chaos. Tomáš Götthans, Ji?í Petržela, Radioengineering, Vol. 20, No. 4, December 2011.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf