MAGNETIC CORES - PART II - BASIC CIRCUITS
ฝัง
- เผยแพร่เมื่อ 20 ก.ย. 2024
- MAGNETIC CORES - PART II - BASIC CIRCUITS - Department of Defense 1962 - PIN 28375 - FEATURES AND FUNCTIONING OF SINGLE-DIODE, SPLIT-WINDING, AND INHIBIT TRANSFER LOOPS; APPLICATION OF THESE LOOPS, SINGLY OR IN COMBINATION, IN ADPS.
This was the first way of using cores. Very expensive as each core needed multiple turn windings and they were implementing computational logic via the windings. Later on when the computation was done with tubes or transistors all each core had to do was simply store a 1 or 0 and the rest of the machine did the work. The cores were arranged in grids and strung on wires that just went straight through, no turns; made it much cheaper. In the simplest form there would be two address wires and a sense wire through each core. Since it takes a certain amount of total current to write a 1 or 0 to a core just over 1/2 the current needed was passed through the horizontal and vertical address wires; only the core where they crossed saw enough total current to set the 1 or 0. To read the data on a core they would have to write to it. Same hysteresis loop, and same pulse when the bit was flipped as in the movie, that's where the sense wire comes in. If the core was already holding a zero and you write a zero to it, no pulse on the sense wire. If the core had a 1 and you write a zero the sense line pulsed. This was called "destructive read." If you cared, you had to write back to the core the value it was to begin with. Each grid was one bit of the computer word, so if you had a 16 bit word, you needed 16 grids or cores.
I sincerely appreciate you explaining that nuance. I was going back and forth between this theory and the practical cores strung up, confused. What computational logic were they doing with the windings before storing 1 or 0?
Amazing explanation
@@LydellAaronThey use similar ideas but everything is handled in a different way. The core of the technology they operate on is exactly the same. No pun intended.
I am a retired bank officer and feel that for the sake of developing deep understanding of the computer science these army videos are excellent source.
I'm watching this now. I love understanding this stuff from a deeper, electromagnetic level. I'm 43 years old and ALWAYS worked on computers but not much with electronics. Got my 1st computer at 11, learned BASIC (it would be like learning Python today - same 'basic' idea). So the whole binary thing always seemed silly. Now I'm 'getting' it. I'm gaining a HUGE appreciation for the miniturization.
+KENNETH UDUT Just to be sure, you realise that modern computers operate on much simpler principles these days? I.e., purely electrical, nothing magnetic going on (apart from in hard disk drives and switchmode power supplies, of course).
TheHue's SciTech
Oh indeed.
But knowing the evolution is very helpful to me.
When you consider the relationship of the behavior of electrons and "electron holes" in semiconductors and compare it to aligning magnetic fields... it really helps in comprehending "how it works" from several different viewpoints.
electrons and electron holes analogize to balls in cups moving around and aligning magnetic fields analogizes to changing water flow direction.
Yet even within semiconductor lingo, the "source/drain" metaphor remains. Whole thing is fascinating honestly.
Actually the basic principles. I.E. Binary, shifting bits around, etc is exactly the same. All that's really changed is the technology that stores and moves the bits around. The binary operations required for say, adding 1+1 would be the same using magnetic core as described here, or if you were using transistors. However the hardware layouts that implement those functions would be drastically different.
I hope i'm not the only one laughing when it approaches around the 20-minute mark.
Aside that, this is a very educational video.
This should be curriculum at schools. ! WOW - incredible !
tendré que esperar mas de 60 años mas para entender como es que funcionan en estos tiempos los puertos lógicos .Que explicación mas didáctica y profunda de este tema .gracias por el video
Presentations were practically an art form way back then! (Oof, I remember teachers in pre school using these velcro boards. I'm old.)
Very useful and information video, fellow Amerikan komrads. I am go to Miami beach with my American wife and have Thank's Giving and apple pie. God Bless Amerika. No commies here.
0:23 "we examined the basic properties"
Hmmm, the _core properties_, you might say?
My difficulty with all of this teaching is that because I don't understand the practical use of doing this I can't really understand very well what is happening and what the ultimate goal of all this is.
Well it's technology is 40-60 years old and is meant to be forgotten! :D Both this and the first video, though they could gear toward demonstrating any logic, both lean toward explaining a shift register which is still stock logic today in an IC 74595. One use of the day is explained, it could serve as a memory, and common today in light chasers and scrolling LED signs. You can make a light chaser with a 555 timer, flip-flop, and shift register.
Brek Martin Thanks, that helps a bit. A scrolling LED explains to me the part of wanting to shift the value from one register to the next. I was thinking that they were explaining how memory was stored alone. Which is why the shifting didn't make sense.
+Pianoman 70s style its simple its differrents circuits to make the differrent boolean operation , they show : not gate, or gate etc but they give differrents name for exemple" the split winding transfert loop" or "inhibit transfert loop"
listen at 28:02 he tell you the purpose.
It help store bits like a register would.
Apparently this presentation is about combinatorial logic with memory function. But I don't get that. By 1961 technology was sufficiently advanced that this logic could be implemented using transistors. Only for memory (I mean real storage) cores were used. But these circuits implement flip-flops and latches used to build counters and accumulators and such. In those days the military was even ahead of the civil developments. Or would those circuits have been more resilient to EMP and is that the reason no semiconductors were used?
The explanation as such is superb. Instructions like this have not being made anymore for a long time. And why not? I know education techniques have advanced during the past 50 years, but errr.... to what exactly? Advanced for the worse?
3 years without an answer. Yikes.
These where developed to overcome the issues of the transitor as they had them up tp the 1960's.
And that is to say vacuum tubes, and discreet components, that where all highly lossy, slow in comparison, and very dependant on current and heat.
Of course modern transistors are still dependant on current and heat sensitive but not as bad as they where back then, but magnetic logic itself is not volatile, that is it keeps its data when the power is removed, and that brings down operation costs, very important when a simple computer fills a room.
In addition the advanced windings could offer an half adder with a single core, although I do not know the schematic, but that alone brought the size of these machines down to a single wall, and that was using 1 inch cores, before the transistor and it's ability to be integrated overtook the speed and matched the efficiency of the core logic.
But this is just the tip of the iceberg when it comes to the logic methods we abaonded, and are now looking to revive due to issues with further shrinking. We can produce tighter tolerances and some smaller features allowing for some gains, and still see the advertising bragging about about smallest features but transistors and capacitors have reached thier limits.
This is a very good and informative. Educational video. Now I kann build a frequency devider without ICs, with this technology
back in the day, would it have been physically possible to use capacitors in lieu of the cores?
for example:
while capacitor [C1] held a charge, it would be a binary One and when the charge of C1 was passed on to the next capacitor [ C2 ] the value of C1 would be a binary zero and C2 would now be a binary one.
Also using Electrolytic caps, the polarity could have been controlled.
with the Pos. to pos. / Neg to neg connection being a binary one and the inverse of that [ Pos to neg and Neg to pos ] would be a binary zero.
+DUSTERDUDE238 : In theory, no. You cannot "pass on" a charge. When you hook up two capacitors together, the charge flows only halfway from one to the other, until each has 1/2 the voltage. You also loose half of your power due to resistance losses. Now, if you add an inductor, you may be able to transfer a lot more power from one cap to the other (sort of like a transmission line), but you will eventually loose all of your charge after enough stages. You would need amplifiers to recover the charge, and if you actually have amplifiers, you have a switch that you can use to directly implement digital logic (you could easily make all basic digital gates from a op-amps and resistors).
However, you might find the topic of "dynamic logic" interesting, because it uses a capacitor on the output of every gate. Wikipedia has a good intro on the subject.
Kevin Harrelson
Ok Kewl Thanks :0
The caps would slowly discharge
If you built a circuit that functioned on current instead of voltage yes you could, but due to power dissipation being a function of current the power consumption of such a device would be high.
Using ferric loops like this exploits some properties of matter.
However you can picture the modern transistor as being capacitance based computing since the electron depletion zone within junction has a very small capacitance.
Pfff , boning to watch but you can get lot's of knowledgeof it.
Thanks for uplioading anyway.
I dig how Invader Zim is in these videos... just placed within the mixture
. XD
Hi :) Was there ever a part 3 for this series?
Why didn't the army run the department of education. So practical and common sense teaching.
In Father and my Husband Jesus' Name, Amen ✝️✨
Interesting
Interesting video and hilarious because he is same to agent K of men in black
how did they made such nice schematics in 60s?
I wonder why the first and second parts were released a year apart unless it was just on the verge of a new year.
Recorded at the same time even: somewhere in part 1, he walks away from the board with diagram that is explained in this part.
I wonder if that’s when the videos became declassified
Alguien sabe si esta traducido al español o castellano o LATINO o SUDAMERICANO..?? gracias
19:38 there it is
And to think... now we've managed to get most of our processors down to just two or four cores. ; )
To think that today most people can't even do LED blinkers without a microcontroller...
+szlatyka Imagine that using uC may be the cheapest way to build LED blinker today :O
+brys555 : A 555 timer will at least require the chip, plus a couple of resistors and caps. Since you can get a micro for under a buck that does not require those passives, why not? Some micros are available in an 8-pin package and about the same price as a 555.
Kevin Harrelson
Some PIC10F.. and ATtiny10 are available in 6-pin SOT23 package.
Flip-flop circuit.
People talking about a 555 or a flip-flop are overkilling it. Just make a bistable multivibrator with a couple of transistors and a couple of caps, plus a few resistors to charge/discharge the thing.
Does anyone else see something wrong at 19:50? =P
DRUGMONSTER It's wireless :D