UBUHANZI
UBWIZA N'UBUZIMA
UBUKORIKORI
UMUCO N'AMATEKA
IMYIDAGADURO
IBIDUKIKIJE
IBIRIBWA N'IBINYOBWA
UBWUBATSI BUSUBIRA INYUMA
UBUMENYI
SIPORO
IKORANABUHANGA
IBYAMBARWA
The Transistor Logic Gate
Volt

Yaremwe na

Volt

27. Kanama 2026SE
33
0
0
0
0

The Transistor Logic Gate

Every limit in this batch has been physical rather than logical. The relay was correct and took milliseconds because it had to move metal. The valve switched a thousand times faster and burned a heater continuously, so a machine with eighteen thousand of them consumed a house’s worth of power and failed somewhere every few hours. The transistor, already in this catalogue, does the same switching with no moving part, no heater, and a fraction of the power — and it is small enough that thousands fit where one valve stood. Nothing in the logic changes at all: the gates are the same gates Shannon described in 1937 and the arithmetic is the same arithmetic. What changes is that they become cheap enough to use in quantities that make the previous chapters look like sketches.
Hagati
5 hours 30 minutes

Amabwiriza

1

Build the same three gates, a thousand times faster

Rebuild the relay blueprint's gates with transistors and compare them directly.

  1. NOT: one NPN transistor, base through a resistor to the input, collector through a resistor to supply. Output at the collector.
  2. NOR: two transistors sharing one collector resistor — either input high pulls the output low.
  3. NAND: two transistors in SERIES — both inputs must be high to pull the output low.
  4. Verify each truth table, then measure propagation delay with the oscilloscope.

Series is AND and parallel is OR, exactly as with relay contacts — the topology did not change, only what is doing the switching. The delay will be tens of nanoseconds against the relay's milliseconds.

Note that the natural transistor gates are the INVERTING ones — NAND and NOR — because a transistor pulls its output DOWN when driven. A non-inverting AND costs an extra inverter. That is why real logic families are built from NAND and NOR, and why the next step matters so much.

Ibikoresho by'iyi ntambwe:

Transisitori zinyuranyeTransisitori zinyuranye1 ibikoresho
Ikusanyirizo ry'Ibirwanya AmashanyaraziIkusanyirizo ry'Ibirwanya Amashanyarazi1 ibikoresho
Ibikoresho bya LED zerekanaIbikoresho bya LED zerekana1 ikirundo
Urubaho rufite utwobo / urubaho rw'igeragezaUrubaho rufite utwobo / urubaho rw'igerageza1 igice
Insinga yo guhuzaInsinga yo guhuza1 reel

Ibikoresho bikenewe:

Ahantu ho GushongeshaAhantu ho Gushongesha
OsilosikopuOsilosikopu
Igikora Imirongo y'AmashanyaraziIgikora Imirongo y'Amashanyarazi
Igipimo cya Digitale cy'Urwego rw'UbushakashatsiIgipimo cya Digitale cy'Urwego rw'Ubushakashatsi
Isoko ry'Amashanyarazi ry'Ameza RihindurwaIsoko ry'Amashanyarazi ry'Ameza Rihindurwa
2

NAND is enough for everything

Ahakorerwa Blockly

Loading Blockly workspace...

Ibikoresho bikenewe:

Mudasobwa yo ku MezaMudasobwa yo ku Meza
3

Fan-out, noise margin and why digital is reliable

Gupakira ikaye ya Jupyter…

Ibikoresho bikenewe:

Mudasobwa yo ku MezaMudasobwa yo ku Meza
4

Measure the power, and find the real wall

The valve machine's limit was heat. Measure whether that limit really went away.

  1. Measure the current your transistor gate draws with the output high, and again with it low.
  2. Now drive it with a square wave and measure average current against frequency.
  3. Plot it and extrapolate.

Static current is small; average current rises roughly in proportion to switching frequency, because every transition charges and discharges the capacitance of the next stage, and that energy is dissipated as heat every single time.

So the heat problem never disappeared — it was postponed, and it returns the moment you have enough gates switching fast enough. That is precisely the wall processors hit around 2005, when clock speeds stopped rising and manufacturers began adding cores instead: not because faster was impossible, but because the heat was.

CMOS was the escape from the static half of the problem — it draws almost no current when it is not switching, which is why it displaced everything else and why a modern chip's power is almost entirely switching energy. ENIAC's engineers left their valves powered continuously because switching them on and off killed them; the same instinct, for a different reason, is why your phone aggressively idles its cores.

Ibikoresho by'iyi ntambwe:

Transisitori zinyuranyeTransisitori zinyuranye1 ibikoresho
Ikusanyirizo ry'Ibirwanya AmashanyaraziIkusanyirizo ry'Ibirwanya Amashanyarazi1 ibikoresho
Agafuka k'ibibiko by'amashanyaraziAgafuka k'ibibiko by'amashanyarazi1 ibikoresho

Ibikoresho bikenewe:

OsilosikopuOsilosikopu
Igikora Imirongo y'AmashanyaraziIgikora Imirongo y'Amashanyarazi
Igipimo cya Digitale cy'Urwego rw'UbushakashatsiIgipimo cya Digitale cy'Urwego rw'Ubushakashatsi
Igipimo cy'amashanyarazi gifata AC/DC 600 AIgipimo cy'amashanyarazi gifata AC/DC 600 A
Isoko ry'Amashanyarazi ry'Ameza RihindurwaIsoko ry'Amashanyarazi ry'Ameza Rihindurwa
5

Assemble the batch into a computer

Everything needed has now been built. Put it together and notice that nothing is missing.

  1. Gates — from this blueprint, or from relays if you prefer to hear it working.
  2. An adder — from the binary adder blueprint.
  3. Registers — flip-flops from the Eccles-Jordan blueprint.
  4. Memory — core, or simply more flip-flops at small scale.
  5. A program counter, an instruction decoder and a conditional jump — from the stored-program blueprint.
  6. An assembler — from the previous blueprint, to make it usable by a human.

That is a complete general-purpose computer, and every part of it was invented before 1950. Everything since has been the same architecture made smaller, faster and cheaper — which is not a small thing, since it changed the world, but it is not a different thing.

The batch's real lesson is the one it has repeated at every step: the logic was finished in 1937 and the engineering was never finished at all. Shannon's algebra describes a relay machine, a valve machine, a transistor machine and the processor rendering this page, without alteration. What changed a million-fold, twice, was the switch — and each time the switch improved, the same ideas simply became affordable in larger quantities. If you want to know what computing does next, do not look at the algebra. Look at what is replacing the transistor.

Ibikoresho bikenewe:

Mudasobwa yo ku MezaMudasobwa yo ku Meza

Ibikoresho

6

Ibikoresho bikenewe

7

CC0 Umurenge rusange

Iyi blueprint yasohowe munsi ya CC0. Ushobora gukoporora, guhindura, gukwirakwiza no gukoresha nta kwemererwa.

Shyigikira Umuremyi ugura ibicuruzwa binyuze muri Blueprint ye Komisiyo y'Umuremyi byashyizweho n'Abacuruzi, cyangwa kora verisiyo nshya y'iyi Blueprint ukayinjiza nk'isano muri Blueprint yawe kugira ngo musangire inyungu.

Ibiganiro

(0)

Injira kugira ngo ujye mu biganiro

Gutegura ibitekerezo...