Apollo Guidance System Schematics
 

Some interesting historical stuff here! Check this out

APOLLO GUIDANCE COMPUTER (AGC) Schematics

:cool:

This is insanely cool.

Man - that brings back some memories of the 70's. Hardware - hardwired logic!!! I wuz there!!! Wellllllllll - I don't mean I worked on Apollo (but Uncle Buddy did). I worked on surface to air missile systems while in the USN and in those days most of the stuff was dedicated logic circuits - heck, the old stuff was analog circuits and relay switching. I was on the 1st ship to be converted over to a digital fire control system and it had a big ol mainframe computer running things. I think it was running at a blazing 1Mhz clock rate.

Very cool indeed!

Nowadays we can't even build a power amp or a pre-amp without an MCU!

I heard that the flight control computer on the Shuttle was 60's/early 70's vintage electronics as well - very simple, well tried and tested, zero risk stuff. Apparently all the PCB's are wire wrapped and IC's socketed and they did/still do this because a wirewrap connection is very reliable (multiple connections per wrap and 'gas tight' at the wire/post interface). I'm not a materials guy, so maybe someone with more expertise or direct knowledge could comment - is this true?

Its only when you look at stuff like this through the lens of modern electronics technology that you realize that a few great engineers acheived a hell of a lot with so little.

Come on Thomas, I know that you worked with Thermionic devices and mechanical computers. My brother did his FC school in the early 90's and he had to learn them, as the military still uses them (for obvious reasons).

Bonsai, great link. Always cool to see things from the past like this. Remember that the processor in your average calculator is as or more powerful than the computing that went into the Apollo program.:cool::cool:

Smart people don't need technology to bail them out, they make the technology!

Peace,

Dave

What - this old stuff???? FIRE CONTROL FUNDAMENTALS

Yup. KABOOM!:bomb::bullseye:

Peace,

Dave

Neat links! I remember the Nike bases and sonic booms.

The truth of the matter is that it takes soooooo long to design and seek approval for a major system that by the time it's finished it's about 10 years behind the current state of the art. You don't have the money or the time for a "do over" so you are stuck with your 10 year plus old technology - which will appear to be "simple" compared to the new systems. IF you go by the premise that processor computational power will double every couple of years Moore's law - Wikipedia, the free encyclopedia 10 years is 5 generations - or 1/16th the power available now.

Wirewrap was sold on the concept of improved reliability (in a number of areas) and a lot of military systems used it. Buuuuuut like all things made by man it did - and does - have it's share of problems. Oxidation of the wire wrap terminals is one concern - loosening of the wirewrap due to thermal expansion and contraction over time is another as well as vibration and strain placed up bundles of WW ties by harness straps.

I worked as a civilian at a military overhaul and repair depot and we fixed, overhauled, modified, all of that old beat-up stuff. 20 years of wear and tear stuck inside something like an aircraft or tank is NOT kind to anything - esp. if it is outside exposed to the elements (including shell hits). :eek:

WW can work really well in labs and low volume production runs - but watch out after about 10 years -

As to using socketed IC's - only ROM's - EPROM's sorta stuff. In fact IC sockets were viewed with a degree of concern because of the high vibration and g-loads encountered. What they liked to use - and the main reason they were developed - is surface mounted devices. Great savings on weight and reduction in size. Theeeeeen we started playing with custom gate arrays - and making our own circuits to customized functions and features.

I was involved in a project to introduce a digital fire and tracking radar control processor into the HAWK anti-aircraft missile system in early 70-ies. Just the A/D and D/A between the processor and launchers, radars etc was a huge 19 inch rack full of PC cards.
The processor memory initially was 4k core memory, 25 bit wide (24 data + 1 parity). Quickly that was doubled to 8k and a few years later doubled again to a whopping 16k of EPROM.
Coding was assembly of course, and you submitted your punched card deck for overnight compiling.
I remember trying to make very efficient and tight program loops because at one point we had 3 free memory slots left, and they STILL wanted additional features ;)

jan didden