• Please review our updated Terms and Rules here

Analog Computer .. Questions I love to hear the answers to

I

inotarobot

Veteran Member
Joined
Feb 7, 2014
Messages
1,090
Location
Melbourne, Australia
OK all you Digital computer buffs here are some Survey questions I am posing on analog computers

AKAT-1.jpg AKAT-1 pic 2.jpg This AKAT-1, by the way, is my favorite analog computer, as the look is soooooo cool. Wish I could own one.

Q1. If you could locate an analog computer in your budget range
a. would you buy it ?
b. If one you really liked came up, BUT outside your budget, how much would you consider paying for one in US$ or Euros ?

Q2a. If you were given or left a working analog computer, would you

A. keep it or
B. sell it, or
C. give it away ?

Q2b. If you were given or left a NON working analog computer, would you

A. keep it and work to get it fully operational or
B. sell it in a non working state, or
C. give it away, again in a non working state ?

Q3. Ok so lets for a now assume, you suddenly had a working analog computer come into your possession,

what task would you think you could do with it ?

Please DONT Google for an answer about the tasks one can do with such a machine, just answer from your existing knowledge base, as it interesting to me

Q4. If you had the chance to build a brand new state of the art Analog Computer, would you

A. jump at the chance
or
B. go nar, I would prefer to wait to get a vintage machine ?

Q5 Have you ever or do you currently owned an analog computer

A. Yes

B. No

Ok Q6. so early analog computer were made using various technologies.

A. all Mechanical ie Gears and Differentials or Rotatable Cardboard Disks
FLOW-SCHEMATIC-COMPUTER-MK-1-2.jpg

B. all Vacuum Tube Valves
gapr2.jpg
r500 photos 01.jpg

C. Transistors era
Welsh Scientific Cat 7568.jpg

D. early integrated circuits.
next 2 pics are from this wonderful web site
http://www.earlycomputers.com/cgi-bin/item-report-main.cgi?20051031a
amf-ec-i003.jpg amf-ec-i013.jpg

So out of A,B,C and D styles

the question 6 is from the 4 constructional types just mentioned, do you have a preference of type you would like to own ?


Q7 Do you own any paper books on Analog computers ? and if so how many ?

and the last question for now is

Q8 Do you have any parts of or from an Analog computer ?

if so A. are you looking to find more of that machine ? maybe to get it working.
or B are you maybe thinking of selling or disposing of these parts

Looking forward to hearing answers

thanks in advance for reading and hopefully answering

As a parting note this is a pic of my EAI TR-20 Analog computer. Its such a great machine and I am so proud to have it
EAI 10a.jpg
 
I have a Donner 3500 and a Heathkit EC-1. They both work pretty well, some calibration needed. No practical use for these, but they're occasionally fun to take out and hook to an oscilloscope to see how changing capacitors in the amplifier circuit works. Might be fun to use with a plotter if I ever decided to try something like that. I think that's what I'd do with one of these, plot stuff.
 
billdeg said:
I have a Donner 3500 and a Heathkit EC-1. They both work pretty well, some calibration needed.
Click to expand...

I have an EC-1 on my to do pile. Nuts & Volts mag recently had an article on restoring one, which I've been meaning to get ahold of.
 
I don't have any analog computers around, but I once got a visit to the fire control room of a WW2 USN Battleship. The computer that aimed those massive guns and turrets was the most awesome analog computer I've ever seen.
 
In my earlier days, I played a bit with the then-new Heath EC-1. It was interesting, but not nearly as much as the digital computers of the day. I remember doing the bouncing ball simulation.

Sometime around that time, I visited Argonne National Labs who were very proud of their hybrid digital-analog computer. I saw it, but don't otherwise remember a thing about it.
 
Analog computers are possibly still useful in simulating
various things.
Most things are described in the form of differential equations.
You'd need enough math knowledge to convert them to integral
equations that you'd then simulate.
I did once have one of mine run a plotter and do rose engine
stuff.
Dwight
 
Whilst analog computers can still solve differential equations there is one major problem with using them in the modern world. That is the accuracy of the solution cannot be measured. On the other hand you can generate approximations to the solutions to a much greater range of differential equations to any required accuracy.

If you want to try this for yourselves GNU Octave is a free "MatLab like" program which will do this...

https://www.gnu.org/software/octave/

Note that unlike an analog computer which actually integrates, digital tools only generate an approximation of the integral. You can control how accurate it is, and make the error as small as you want, but there will always be some uncertainty in the answer. The smaller you make it the more calculations you will need to be performed, and so the time to produce a solution will extend. One popular tool which solves differential equations numerically is the SPICE circuit modelling tool. It has settings which allow the acceptable error to be controlled.

If any one wants to play with Analog Computers but does not have one, there are some components for LTSpice which allow one to be simulated..

http://www.edn.com/design/analog/4376400/A-virtual-analog-computer-for-your-desktop

of course this is really using digital approximations under the covers so it isn't real.

Analog Computers continued in use for long after digital computers could solve these problems. This was for a couple of reasons, firstly they were much cheaper than digital computers so could be widely deployed. Secondly, and IMHO most importantly, is that when the differential equations are with respect to time, which many are then an analogue computer solves these at a constant rate proportional to that time.

So for example in an Aircraft Simulator where many differential equations may need to be solved, analog computers can easily be used, because the inherently work in real time. Doing the same task with digital computers is harder because you have to sync the solution to real time, which can be harder than actually solving the equations....
 
g4ugm said:
Whilst analog computers can still solve differential equations there is one major problem with using them in the modern world. That is the accuracy of the solution cannot be measured. On the other hand you can generate approximations to the solutions to a much greater range of differential equations to any required accuracy.
Click to expand...

Hi Dave, thanks for that full and informative reply. At this moment One point I like to make, I think this paragraph is missing the words "with Digital computers" to make it completely non ambiguous.

This I feel what you are saying is
"Whilst analog computers can still solve differential equations there is one major problem with using them in the modern world. That is the accuracy of the solution cannot be measured. On the other hand with DIGITAL computers you can generate approximations to the solutions to a much greater range of differential equations to any required accuracy."

This I feel is what you are saying. Hope this is what you meant ?

regards
David
 
Last edited:
Q1..2: Probably I would buy it, keep it and if something is damaged, try to get as many blocks into operation as possible.

Q3: What is it for... solving equations of course, it may be a bit hard to get OS on it ;). This AKAT-1 you shown, built by Mr Karpiński, is loosely based on early ARR - Differential equations analyzer made in 1953.

Q4: Depends on state of my trash bin in my basement. If there are lots of spare parts, I would try to build something.

Q5: A slide rule only.

Q6: Because of space requirements, the only analog computer that will fit in my cave would be transistor-based. Or hybrid circuits/ICs. Maybe some small one, see this Soviet small educational unit with switchboard, generators, filters, integrators and a small scope.

Q7: I have lots of books about analog computers, some date from early 1960s. Most in Polish. Well, from this knowledge I could build one.
BTW but the strangest book in my collection is still one about building binary logic circuits using... hydraulics.

Q8: Probably not.
 
A lot of this stuff was from the pre-IC era. One of the earliest linear ICs I experimented with was the Fairchild uA709--introduced in 1965--a operational amplifier. Its successor, the uA741, debuted in 1968 and is still in production today. This was a remarkable advance, as vacuum-tube and discrete transistor op-amps were expensive.

So, given that this essential piece of analog computing is today never easier to source, how many have constructed their own analog computer using monolithic op-amps?
 
Is this electronic based analogue computing, or is mechanical based included as well?

There are lots of 50s era mechanical computers around, thing the Mk4/4A ground position indicators from the Canberra or Vulcan. There are also a Mk6 but only seen one. The bombsight computers from WW2 are available. I have a dead reckoning table that recorded the track of a ship that needs rebuilding.

Has anyone been to HMS Belfast and seen if that still has the analogue fire control computers? Really would like to go and see, but that is London.

These are of course specialised computers but wheter thay can be sensibly adapted for other uses I don't really know.
 
woodchips said:
Is this electronic based analogue computing, or is mechanical based included as well?

There are lots of 50s era mechanical computers around, thing the Mk4/4A ground position indicators from the Canberra or Vulcan. There are also a Mk6 but only seen one. The bombsight computers from WW2 are available. I have a dead reckoning table that recorded the track of a ship that needs rebuilding.

Has anyone been to HMS Belfast and seen if that still has the analogue fire control computers? Really would like to go and see, but that is London.

These are of course specialised computers but wheter thay can be sensibly adapted for other uses I don't really know.
Click to expand...

Hi there,

yes this post is about all mechanism types of analog computer. In fact in my initial (a) part Q6 I had a photo of a mechanical gun computer,
So yes this post is for discussion on all types,

I have not been to see HMS Belfast, maybe in September on my way back from Europe as I head home to Aus, I will try an go look.

Would you be able to post some pics of that dead reckoning table you have here please ?
 
MCbx said:
Q1..2: Probably I would buy it, keep it and if something is damaged, try to get as many blocks into operation as possible.

Q3: What is it for... solving equations of course, it may be a bit hard to get OS on it ;). This AKAT-1 you shown, built by Mr Karpiński, is loosely based on early ARR - Differential equations analyzer made in 1953.

Q4: Depends on state of my trash bin in my basement. If there are lots of spare parts, I would try to build something.

Q5: A slide rule only.

Q6: Because of space requirements, the only analog computer that will fit in my cave would be transistor-based. Or hybrid circuits/ICs. Maybe some small one, see this Soviet small educational unit with switchboard, generators, filters, integrators and a small scope.

Q7: I have lots of books about analog computers, some date from early 1960s. Most in Polish. Well, from this knowledge I could build one.
BTW but the strangest book in my collection is still one about building binary logic circuits using... hydraulics.

Q8: Probably not.
Click to expand...

Thank you MCbx for your detailed answer. I was interesting to see your responses to each question, especially a couple of them/

I think a lot of computer foke forget that a slide rule is actually a form of analog computer.

Your reply to Q6 was very very interesting, I clicked on that Soviet link.. Its an AMAZING machine. Thank you for that link.. I RECOMMEND all reading this post do have a look at it.

http://www.leningrad.su/museum/show_big.php?n=2360

I also like the comment on the odd book of building binary logic circuits using... hydraulics.

I saw where a MIT Grad made part of a computer logic gate using water
http://www.blikstein.com/paulo/projects/project_water.html

Have not read the whole article.

maybe you should look at trying to find remains of one in your 'neck of the world'
 
When you get to analog mechanical systems, the question inevitably arises "What's a computer?"

If you say "that which computes an output from more than one input", then we've been surrounded by computers for at least 100 years or more. Consider the ordinary water meter. It takes a flow over time and integrates it to show consumption. Is it a computer?

When I was still going to school in the dark ages, I spent my summers working as an instrumentation tech in a steel mill. While there were certainly electronic systems, there were also pneumatics and hydraulics as well. You could start with a differential pressure across a venturi and perform calculations to show flow and consumption all using pneumatics. There were adders, subtractors, multipliers and even square-root extractors--all pneumatic/mechanical. No wires--just lots of 1/4" copper tubing carrying signals (IIRC) from 3-15 psi.

Nobody, to the best of my knowledge, called them "computers".

So were they?
 
In fact every circuit which has characteristics corresponding to some other physical phenomena may be used as some analog simulational computer.

->inotarobot
About hydraulic/pneumatic computers, I was asking. Simple logic elements are still in use in industrial applications, when one pressure must inhibit another or work depending on state of other pneumatic parts, yet it seems that no computer has been built this way. If looking more to the east, there is an information that in Soviet Union in 1930s there was operating hydraulic integrator. There is some information about it in Wikipedia.
So generally I asked a man who was working in Polish N.O.T. (commitee which tried to regulate development and cooperation) and he told me that in 1960s there was a failed attempt to develop a digital machine based on pneumatics. Typical gates have been made using machined/extruded plastic and were "the size of toy block". I don't know in which university such thing has been made nor who tried it.
 
You must log in or register to reply here.
Back
Top