Nostr notes by Ken Shirriff

Credits: IOP boards provided by Richard, 4 Pi board from ...

2026-06-28T16:30:48Z

In reply to nevent1q…7ezz
_________________________

Credits: IOP boards provided by Richard, 4 Pi board from Tube🌞 Time (npub1xvg…62t2).
Photo of the IOP and CPU courtesy of RR Auction. IOP documentation from Mike Stewart. Exploded diagram from https://ibiblio.org/apollo/Shuttle/IBM-74-A31-016%20-%20Space%20Shuttle%20Advanced%20System,%204%20Pi%20-%20Prototype%20Input,%20Output%20Processor%20(IOP).pdf#page=145

This diagram shows how the circuit boards (called pages) were ...

2026-06-28T16:29:31Z

In reply to nevent1q…50ge
_________________________

This diagram shows how the circuit boards (called pages) were plugged into the I/O Processor's aluminum-alloy case. For storage, the system used magnetic core memory, larger pages at the back. This system didn't use a microprocessor; it was built from 11 logic pages crammed with simple chips.

IBM had a series of aerospace computers called System/4 Pi. These ...

2026-06-28T16:29:01Z

In reply to nevent1q…swy6
_________________________

IBM had a series of aerospace computers called System/4 Pi. These computers all used standard-sized boards (top). Except that the I/O Processor's boards (bottom) were one inch wider. I guess they needed the extra space for more circuitry.

The Space Shuttle had five general-purpose computers that ...

2026-06-28T16:25:31Z

The Space Shuttle had five general-purpose computers that controlled, monitored, and navigated the Shuttle. Each computer consisted of two boxes: the CPU (right) and the I/O Processor (IOP, left). The IOP connected the computer to 24 high-speed networks. Let's look at two boards from the IOP... 1/N

Intel introduced the 8087 floating-point chip in 1980 to make ...

2026-06-13T16:21:58Z

Intel introduced the 8087 floating-point chip in 1980 to make math faster. The chip does tangents, square roots, and so forth, but fundamentally everything is computed through additions and subtractions, performed by an adder at the heart of the chip. Let's look at the adder's circuitry... 1/N

Yes, the thyratron is very similar to an SCR (also known as a ...

2026-06-07T19:02:50Z

In reply to nevent1q…t69w
_________________________

Yes, the thyratron is very similar to an SCR (also known as a thyristor: thyratron transistor).

For more on the IBM 604 and this thyratron tube module, see my ...

2026-06-07T17:29:42Z

In reply to nevent1q…hpej
_________________________

For more on the IBM 604 and this thyratron tube module, see my latest article: https://www.righto.com/2026/06/ibm-604-thyraton-tube-module.html

Pluggable tube modules were an important innovation. They made it ...

2026-06-07T17:29:23Z

In reply to nevent1q…awpq
_________________________

Pluggable tube modules were an important innovation. They made it easy to repair tube-based systems, they packaged components densely, and they paved the way for IBM's tube computers in the 1950s. IBM featured the modules in this 1952 ad from Time magazine: "Fingers you can count on". 4/N

The tube is not exactly a vacuum tube since it has a bit of xenon ...

2026-06-07T17:28:12Z

In reply to nevent1q…nkz5
_________________________

The tube is not exactly a vacuum tube since it has a bit of xenon gas inside. It is called a thyratron: when it turns on, the xenon ionizes, allowing a large current to flow through the tube. In the IBM 604, thyratron tubes provided high-current pulses to the electromagnets to punch holes in cards.

The tube module acts as a high-current switch: I hooked it up to ...

2026-06-07T17:26:34Z

In reply to nevent1q…92f2
_________________________

The tube module acts as a high-current switch: I hooked it up to control a light bulb. The first button turns the module on with a small signal. The second button cuts the power to turn the module off. 2/N

The IBM 604 Electronic Calculating Punch was introduced in 1948. ...

2026-06-07T17:24:35Z

The IBM 604 Electronic Calculating Punch was introduced in 1948. Not quite a computer, it was a programmable calculator that became very popular due to its low price of $550/month, and over 5600 were produced. I have a vacuum tube module from this system, so I powered it up... 1/N

I did this work with the Opcode Collective, especially Smartest ...

2026-05-30T17:48:19Z

In reply to nevent1q…37h5
_________________________

I did this work with the Opcode Collective, especially Smartest Blob and Gloriouscow.
The full ROM image: https://8087.martypc.net/#dzi=tiles%2Ftiles.dzi
A description of the 8087 microcode: https://github.com/a-mcego/granite/blob/main/tools/8087mc/bin/8087.md

We are reverse-engineering the meaning of the 8087's ...

2026-05-30T17:47:35Z

In reply to nevent1q…8s43
_________________________

We are reverse-engineering the meaning of the 8087's microcode. For more details, see my latest article:
https://www.righto.com/2026/05/microcode-inside-intel-8087-floating.html

For more on the Spacelab computer, see my latest article: ...

2026-05-23T16:54:05Z

In reply to nevent1q…lr30
_________________________

For more on the Spacelab computer, see my latest article: https://www.righto.com/2026/05/reverse-engineering-spacelab-computer.html

The Space Shuttle could hold a flying laboratory called Spacelab ...

2026-05-23T16:49:16Z

The Space Shuttle could hold a flying laboratory called Spacelab in its cargo bay. Three French-made computers ran Spacelab. I opened up a Spacelab computer and found that instead of a microprocessor, it is built from a multitude of simple chips. Let's take a closer look at the computer...

Fair enough :-)

2026-03-29T22:22:12Z

In reply to nevent1q…z3r5
_________________________

Fair enough :-)

I forgot to explain the 4 Pi name. IBM's System/360 ...

2026-03-29T17:30:27Z

In reply to nevent1q…6pgx
_________________________

I forgot to explain the 4 Pi name. IBM's System/360 mainframes got their name because they supported a full 360º circle of applications, from business to scientific. The 4 Pi aerospace computers, however, were designed for the 3-D world, a "sphere" of applications. While a circle has 360º, a sphere has 4π steradians of solid angle.

There were many more models of System/4 Pi computers, including ...

2026-03-29T17:11:56Z

In reply to nevent1q…mt32
_________________________

There were many more models of System/4 Pi computers, including refrigerator-sized computers used on submarines; see my article for details: https://www.righto.com/2026/03/ibm-4-pi-computer-history.html

In 1991, the Space Shuttle upgraded its computer to the AP-101S, ...

2026-03-29T17:10:28Z

In reply to nevent1q…feky
_________________________

In 1991, the Space Shuttle upgraded its computer to the AP-101S, replacing two 60-pound boxes with one box. Since there were five computers on the Shuttle, this change freed up 300 pounds for additional cargo. The CPU was implemented in three boards full of chips, one below. These boards were much larger than the earlier 4 Pi boards.

The Space Shuttle used a more advanced System/4 Pi computer ...

2026-03-29T17:05:00Z

In reply to nevent1q…ghkn
_________________________

The Space Shuttle used a more advanced System/4 Pi computer called the AP-101B. A separate computer, the I/O Processor, provided 24 data buses to all parts of the Shuttle. These boxes still used individual TTL chips and magnetic core memory.

IBM's 4 Pi computers were built from simple TTL integrated ...

2026-03-29T16:59:44Z

In reply to nevent1q…pmz6
_________________________

IBM's 4 Pi computers were built from simple TTL integrated circuits and used magnetic core memory for storage. "Flat-pack" integrated circuits on multi-layer circuit boards provided dense logic. The board below was round to fit inside a tactical missile, but most of the computers used standardized rectangular boards called "pages".

In 1967, IBM introduced the System/4 Pi line of aerospace ...

2026-03-29T16:53:11Z

In 1967, IBM introduced the System/4 Pi line of aerospace computers, packing mainframe performance into a compact box. 4 Pi computers powered everything from military aircraft to the Space Shuttle to sonar systems on submarines. Thread...

The 8087 doesn't help with 16-bit integer math.

2026-02-14T22:09:42Z

In reply to nevent1q…acn2
_________________________

The 8087 doesn't help with 16-bit integer math.

Early calculator chips are pretty weird. They operate in decimal ...

2026-02-14T22:08:57Z

In reply to nevent1q…3ma4
_________________________

Early calculator chips are pretty weird. They operate in decimal and are serial, so they operate on one bit at a time with a one-bit ALU. So all the data is constantly in motion.

Instruction decoding in the 8087 has many more complications: ...

2026-02-14T17:45:31Z

In reply to nevent1q…7cyp
_________________________

Instruction decoding in the 8087 has many more complications: hardcoded instructions handled by the "Bus Interface Unit", instructions that load a constant from the constant ROM, and handling the instruction queue. For more details, see my latest blog post:
https://www.righto.com/2026/02/8087-instruction-decoding.html

The microcode inside the 8087 chip uses conditional jumps for ...

2026-02-14T17:43:59Z

In reply to nevent1q…ay9w
_________________________

The microcode inside the 8087 chip uses conditional jumps for many purposes. Some conditions examine the instruction, so the microcode can act differently for different instructions. Here's the PLA that stores target addresses for microcode jumps. Target 0 is special: it does a three-way branch depending if the instruction is multiplication, division, or addition/subtraction.

Like many processors, the 8087 uses a "Programmable Logic ...

2026-02-14T17:38:43Z

In reply to nevent1q…3suw
_________________________

Like many processors, the 8087 uses a "Programmable Logic Array" (PLA) to decode the bit patterns in instructions. A PLA consists of grids of transistors, with some of the transistors missing. The pattern of transistors controls the PLA's function. The grid is a dense and flexible way of implementing this logic.

In 1980, Intel released the 8087 floating-point chip, making math ...

2026-02-14T17:32:53Z

In 1980, Intel released the 8087 floating-point chip, making math much faster. I'm reverse-engineering this chip, 46 years later. Most of its instructions are implemented in microcode, but some are implemented in hardware. Let's look at the circuitry that decodes instructions and decides what to do...

Thanks, I meant the 1402 card reader/punch. I've edited it ...

2026-02-05T05:10:34Z

In reply to nevent1q…fy0u
_________________________

Thanks, I meant the 1402 card reader/punch. I've edited it now.

The card reader/punch (IBM 1403) can punch 250 cards per minute, ...

2026-02-04T22:29:15Z

In reply to nevent1q…cewl
_________________________

The card reader/punch (IBM 1403) can punch 250 cards per minute, so the heart cards zipped out at high speed. However, this machine can't print text on cards, so I needed to go back to the keypunch...

To duplicate the heart cards, I toggled a short program into the ...

2026-02-04T22:26:04Z

In reply to nevent1q…uqvw
_________________________

To duplicate the heart cards, I toggled a short program into the 1401 computer to read a card and print out copies. The 1401 has a special feature called "column binary" that allows it to read and punch cards even if the hole pattern isn't valid.

To make the heart, I had to figure out the right sequence of ...

2026-02-04T22:23:00Z

In reply to nevent1q…y7q9
_________________________

To make the heart, I had to figure out the right sequence of characters to form the hole pattern. Inconveniently, some holes don't form valid characters, so I had to "multi-punch" in some columns. This worked for one card, but the keypunch can't duplicate multi-punched cards.

The Computer History Museum has a presentation tonight on the ...

2026-02-04T22:16:41Z

The Computer History Museum has a presentation tonight on the history of computer dating: "Algorithms of Love". I made a special punch card for the event. It wasn't easy to make this card so I'll explain how I did it...

If you want more on the 8086 processor's ALU, see my latest ...

2026-01-23T17:51:13Z

In reply to nevent1q…2teu
_________________________

If you want more on the 8086 processor's ALU, see my latest blog post:

https://www.righto.com/2026/01/notes-on-intel-8086-processors.html

The control signals in the 8086 need to be amplified to drive the ...

2026-01-23T17:49:35Z

In reply to nevent1q…3hxk
_________________________

The control signals in the 8086 need to be amplified to drive the circuitry for 16 bits. The driver circuit is very unusual: it is a "bootstrap" circuit that uses stray capacitance to boost the output voltage higher than normal, similar to a charge pump. The transistor with +5 on the gate acts as a diode, preventing the extra voltage from escaping. On the die, you can see that the drive transistors are much larger than the regular transistors.

Each bit of the 8086's ALU is implemented with an unusual ...

2026-01-23T17:44:15Z

In reply to nevent1q…08u2
_________________________

Each bit of the 8086's ALU is implemented with an unusual circuit. It uses two lookup tables (trapezoids), loaded from six control signals. This reprograms the ALU's logic to implement the desired operation. This is similar to how an FPGA uses lookup tables to implement different logic functions.

What's the XI mux? To make the 8086's microcode more ...

2026-01-23T17:41:29Z

In reply to nevent1q…sqp5
_________________________

What's the XI mux? To make the 8086's microcode more compact, it often doesn't specify the ALU operation, using a special code "XI" instead. The hardware looks at the machine instruction, determines if it is an ADD, SUB, XOR, or so forth, and substitutes the correct ALU operation. Thus, one microcode routine can implement 8 different machine instructions.

The 8086's microcode specifies what ALU operation to perform ...

2026-01-23T17:35:57Z

In reply to nevent1q…uvzc
_________________________

The 8086's microcode specifies what ALU operation to perform along with which temporary register to use. The control circuitry holds this information in latches. A PLA (Programmable Logic Array) decodes the instruction into dozens of control signals. The Lookup Table (LUT) PLA generates six signals that configure the ALU to perform the desired operation.

The arithmetic/logic unit (ALU) in the Intel 8086 processor ...

2026-01-23T17:31:57Z

The arithmetic/logic unit (ALU) in the Intel 8086 processor (1978) is more complicated than you might expect, performing 28 different operations from addition and logical AND to shifts and BCD adjustment. A special control circuit reconfigures the ALU for each operation. Let's look closer...

If you're interested in obscure details of the microcode in ...

2025-12-30T19:51:11Z

If you're interested in obscure details of the microcode in the Intel 8087 floating-point chip, I have a new blog post...
https://www.righto.com/2025/12/8087-microcode-conditions.html

In 1980, Intel announced the 8087 Math Coprocessor, a chip that ...

2025-12-09T18:33:38Z

In 1980, Intel announced the 8087 Math Coprocessor, a chip that made floating-point 100 times faster. I opened up the chip, took photos of the silicon structures, and analyzed its circuitry. It's a very complex chip for its time. Let's take a look inside...

I've been using HTML forever, but I just learned that <div ...

2025-09-15T17:22:01Z

I've been using HTML forever, but I just learned that <div /> isn't the same as <div></div>.
In the first case, the div isn't closed and everything that follows is inside it.

Can anyone identify these mystery terminals? The first is from ...

2025-09-10T16:28:19Z

Can anyone identify these mystery terminals? The first is from "Los placeres ocultos" (1977). The second is from "Policia" (1987). (Someone wants help identifying computer hardware in Spanish movies.)

Can anyone recognize this front panel from a mainframe computer ...

2025-09-09T20:33:19Z

Can anyone recognize this front panel from a mainframe computer from a 1978 Spanish movie? The black color resembles IBM System/370 panels, but the knobs don't look right. Other information suggests maybe an IBM-compatible FACOM/Fujitsu computer.

Marilou Schultz's rug is on display at SITE Santa Fe's ...

2025-09-06T15:30:13Z

In reply to nevent1q…tsdr
_________________________

Marilou Schultz's rug is on display at SITE Santa Fe's current exhibition: "Once Within a Time"; let me know if you see it!

Photo credit: rug photo from First American Art Magazine.

To learn more about the 555 timer rug, see my latest post: ...

2025-09-06T15:29:48Z

In reply to nevent1q…qwe5
_________________________

To learn more about the 555 timer rug, see my latest post: https://www.righto.com/2025/09/marilou-schultz-navajo-555-weaving.html

Marilou Schultz first made a chip rug in 1994, when Intel ...

2025-09-06T15:29:16Z

In reply to nevent1q…scxw
_________________________

Marilou Schultz first made a chip rug in 1994, when Intel commissioned a rug based on the Pentium as a gift to AISES (American Indian Science & Engineering Society). The Pentium weaving used natural dyes, while the 555 weaving uses aniline dyes and some metallic threads for more intense colors.

Marilou Schultz based the rug on a photo by Antoine Bercovici ...

2025-09-06T15:28:14Z

In reply to nevent1q…y8dx
_________________________

Marilou Schultz based the rug on a photo by Antoine Bercovici (Siliconinsider). He used a special dark field microscope that produces a black background, highlighting the metal wiring on top of the silicon. The rug (left) mostly matches the photo (right), but there are some artistic changes.

Celebrated Navajo (Diné) artist Marilou Schultz recently ...

2025-09-06T15:27:04Z

Celebrated Navajo (Diné) artist Marilou Schultz recently completed a striking weaving. Although this rug may appear abstract, it is a representation of the wiring inside an integrated circuit. It shows the 555 timer, said at one point to be the world's most popular IC. Let's take a closer look...

That QR code is the coolest thing. Nice how the black parts show ...

2025-08-19T14:08:50Z

In reply to nevent1q…zfn5
_________________________

That QR code is the coolest thing. Nice how the black parts show the underlying structure, including the alternating squares of the timing pattern, while the colored parts have the data

The latest issue of the journal Science mentions a magnetic ...

2025-08-18T18:49:18Z

The latest issue of the journal Science mentions a magnetic compound Cr2Gr2Te6. The element Gr confused me, but it turned out to be a typo for Ge, germanium. Strangely, I found multiple papers with the same typo in the same context, so I wrote a short blog post about it.

https://www.righto.com/2025/08/Cr2Ge2Te6-not-Cr2Gr2Te6.html

How did Intel's 386 processor connect to the outside world? ...

2025-08-17T14:46:00Z

How did Intel's 386 processor connect to the outside world? Tiny wires connect to the bond pads around the edges. Special input/output circuits—visible in zoomed-in blocks—send and receive signals while protecting the chip from dangers that could destroy it. Let's take a look... 1/N

Here's one more scan of the 386 for your viewing enjoyment... ...

2025-08-09T21:29:33Z

In reply to nevent1q…xfgr
_________________________

Here's one more scan of the 386 for your viewing enjoyment...

For more photos and information on the 386 package, see my latest ...

2025-08-09T16:49:07Z

In reply to nevent1q…m375
_________________________

For more photos and information on the 386 package, see my latest blog post. Thanks to Jon Bruner and Lumafield for the CT scans.
https://www.righto.com/2025/08/intel-386-package-ct-scan.html

Here are the remaining four layers inside the package: one signal ...

2025-08-09T16:48:24Z

In reply to nevent1q…yxw9
_________________________

Here are the remaining four layers inside the package: one signal layer and three power layers. There are four power layers altogether so the chip's I/O circuitry and logic circuitry can be powered separately. That way, current surges from I/O don't affect the CPU.

One of the six power layers inside the Intel 386 processor's ...

2025-08-09T16:45:03Z

In reply to nevent1q…4znv
_________________________

One of the six power layers inside the Intel 386 processor's packaging. It's almost a solid conductive plane; the grid of holes joins the ceramic above and below. Most of the pins are connected to lower layers, so only a few pins (red dots) reach this layer.

The CT scan of a signal layer inside the 386's package. The ...

2025-08-09T16:43:59Z

In reply to nevent1q…9ryw
_________________________

The CT scan of a signal layer inside the 386's package. The wide abstract-looking regions connect the pins (red dots) to the chip in the middle. Unexpectedly, I spotted thin wires running to the edges of the chip. These are used during manufacturing to electroplate the pins with gold.

The silicon die of the 386 processor is mounted in the middle of ...

2025-08-09T16:42:20Z

In reply to nevent1q…r9ts
_________________________

The silicon die of the 386 processor is mounted in the middle of the ceramic package. Tiny bond wires connect the die to two tiers of contacts that surround it. Intricate but hidden wiring inside the ceramic connects these contacts to the 132 gold-plated pins. 2/N

Intel's 386 processor was popular in late 1980s ...

2025-08-09T16:41:24Z

Intel's 386 processor was popular in late 1980s microcomputers. From the outside, the 386 chip is a boring ceramic square. But I obtained a 3-dimensional CT X-ray scan from Lumafield, revealing six layers of complex wiring hidden inside the ceramic package. Let's take a closer look... 1/N

The chip uses these components to turn an FM radio signal into ...

2025-08-02T15:48:17Z

In reply to nevent1q…qn3c
_________________________

The chip uses these components to turn an FM radio signal into audio by combining an oscillator, mixers, filters, amplifiers, and other circuits. If you want to know more about this FM radio chip, I go into great detail in my blog post: https://www.righto.com/2025/08/reverse-engineering-analog-TDA7000.html

The most unusual components in the chip are the "varicap ...

2025-08-02T15:48:10Z

In reply to nevent1q…qj0k
_________________________

The most unusual components in the chip are the "varicap diodes", which help adjust the radio's frequency. These special diodes act as capacitors with a value that can be controlled by changing the voltage.

Another important part of the chip is the resistors, which impede ...

2025-08-02T15:47:30Z

In reply to nevent1q…sf0n
_________________________

Another important part of the chip is the resistors, which impede the flow of electricity. They are typically long and zig-zagging. The ends of the resistors are connected to the metal wiring. The arrows indicate the holes in the insulating oxide layer where these contacts take place.

Transistors are the main components of the chip, amplifying ...

2025-08-02T15:46:58Z

In reply to nevent1q…sq0z
_________________________

Transistors are the main components of the chip, amplifying signals and performing other functions. Modern chips use CMOS transistors, but this chip uses NPN and PNP transistors. These diagrams show how the transistors look on the die, along with cross sections showing the silicon layers.

This close-up of the die shows how the metal wiring (white lines) ...

2025-08-02T15:45:41Z

In reply to nevent1q…epe5
_________________________

This close-up of the die shows how the metal wiring (white lines) connects the underlying silicon structures (gray and green). Unlike modern chips, this chip only has one layer of metal, so the wiring needs to twist and turn when other wires get in the way.

Here's a die photo of the FM radio chip, the TDA7000. You can ...

2025-08-02T15:45:07Z

In reply to nevent1q…0sss
_________________________

Here's a die photo of the FM radio chip, the TDA7000. You can see the greenish silicon, with white metal wiring on top. The 18 white squares around the edges are the bond pads that connect the die to the chip's pins, with gold wires attached. I've labeled the main functional blocks of the chip. 2/N

In 1983, Philips produced the first FM radio receiver on a chip, ...

2025-08-02T15:44:12Z

In 1983, Philips produced the first FM radio receiver on a chip, leading to products such as the FM radio wristwatch. Let's look at the tiny silicon die inside this chip and see how it works. 1/N

This box has thirteen orange digits at the top, which look like ...

2025-07-21T16:38:46Z

In reply to nevent1q…nyc9
_________________________

This box has thirteen orange digits at the top, which look like Nixie tubes. But they are a different technology called edge-lit lightguide display. Each digit has ten plastic sheets and ten lightbulbs. Each sheet has dots etched in the shape of a number. Lighting a sheet lights up that number.

We obtained a mysterious box of 1960s electronics. I ...

2025-07-21T16:37:44Z

We obtained a mysterious box of 1960s electronics. I reverse-engineered it and with much effort, we got it running. It turns out to be a test unit for testing NASA's Up-Data Link, a system from the Apollo moon landing to control the spacecraft from the ground. Let's take a look inside... 1/n

Texas Instruments introduced LED-based opto-isolators in 1964. ...

2025-07-12T02:30:18Z

In reply to nevent1q…q0j0
_________________________

Texas Instruments introduced LED-based opto-isolators in 1964. They used an infrared GaAs LED shining onto a silicon photo-transistor. This device had the curious name of "molecular multiplex switch/chopper"; making small devices through "molecular electronics" was the hot thing back then. The device cost $275 (almost $3000 in current dollars; nowadays, an opto-isolator costs a few pennies. 2/3

NASA used this "Up-Data Link Test Set" to test one of the ...

2025-07-12T02:15:40Z

NASA used this "Up-Data Link Test Set" to test one of the electronic systems onboard the Apollo spacecraft. I've been reverse-engineering this box and I found that it uses LED-based opto-isolators. I didn't realize that they had LEDs and optoelectronics in the mid-1960s, so I investigated... It turns out that opto-isolators existed back then but were extraordinarily expensive. 1/N

And if you'd rather see inside the Globus visually, check out ...

2025-07-11T00:57:12Z

In reply to nevent1q…3pqv
_________________________

And if you'd rather see inside the Globus visually, check out my blog post that examines its gears, cams, differentials, and other mechanisms in detail. It's essentially a mechanical analog computer. Thanks to Marcel for providing the Globus. 4/4
https://www.righto.com/2023/01/inside-globus-ink-mechanical-navigation.html

Lumafield's scanner is very cool if you want ...

2025-07-11T00:56:52Z

In reply to nevent1q…dwmj
_________________________

Lumafield's scanner is very cool if you want three-dimensional scans of objects for failure analysis, reverse engineering, or inspection. You can play with the Globus scan for yourself at this link, zooming, rotating, and slicing the Globus. 3/4
https://voyager.lumafield.com/project/d848dd54-d594-479f-a723-a463547ea7aa

This photo shows the Globus with its rotating globe and a ...

2025-07-11T00:56:27Z

In reply to nevent1q…l57a
_________________________

This photo shows the Globus with its rotating globe and a crosshair showing the spacecraft's location. Dials next to the globe indicate latitude and longitude, while the dial at the bottom indicates when the spacecraft is in light or darkness. The knobs control orbital settings. 2/4

The Globus is a navigational instrument that uses a rotating ...

2025-07-11T00:55:22Z

The Globus is a navigational instrument that uses a rotating globe to show the position of the Soyuz spacecraft above the Earth. Inside the Globus, a complicated system of gears and motors positions the globe. Jon Bruner from Lumafield (npub1jsp…ggr4) created a three-dimensional X-ray scan for us. 1/4

In the 1960s, this absurdly large camera was used to make masks ...

2025-06-25T17:34:03Z

In the 1960s, this absurdly large camera was used to make masks for integrated circuits. The layers of the integrated circuit were drawn at large scale and then optically shrunk to make the glass masks that were used in manufacturing.

It turned out that the computer's START button had a strand ...

2025-06-20T03:02:53Z

In reply to nevent1q…y6tq
_________________________

It turned out that the computer's START button had a strand of wire that shorted one of its contacts. As a result, it would keep re-starting as long as you held the button, blowing through the HALT or single-step until it crashed. The contacts bounced a bit, causing the random gaps we saw.

You push the green START button on the console to start the ...

2025-06-20T03:01:09Z

In reply to nevent1q…7zpt
_________________________

You push the green START button on the console to start the computer after a HALT, and also to single-step it. Hmmm. For convenience, there's also a START button on the card reader and one on the printer. We found that everything worked fine with those buttons. Hmmm.

I hooked up an oscilloscope and discovered that the computer ...

2025-06-20T02:59:46Z

In reply to nevent1q…haec
_________________________

I hooked up an oscilloscope and discovered that the computer wasn't running a single instruction cycle. Instead, it ran a few cycles (the pulses below), stopped briefly, ran a few more, irregularly stopping and starting, "building up steam" until it ran thousands of cycles. But why this pattern?

The IBM 1401 computer lets you run code a single cycle at a time, ...

2025-06-20T02:58:19Z

In reply to nevent1q…nkaa
_________________________

The IBM 1401 computer lets you run code a single cycle at a time, so you can see what happens at each step. (The arrow on the control panel.) But when I tried to single-step through the bad HALT instruction, the computer immediately died in a distant memory location. Now we had two problems. 2/N

Last week, the vintage IBM 1401 computer at the Computer History ...

2025-06-20T02:57:39Z

Last week, the vintage IBM 1401 computer at the Computer History Museum started behaving strangely: it wouldn't halt. More specifically, if you had two HALT instructions in a row, it would halt for the first, but when you continued, it crashed mysteriously. Here's how we fixed it.... 1/N

How thin is spider silk compared to the wiring on a chip? ...

2025-05-17T19:02:38Z

How thin is spider silk compared to the wiring on a chip? Here's a close-up of a spiderweb strand diagonally across an Intel 386 processor. The spiderweb is 4 µm thick, just a bit thinner than this 1985 chip's 5 µm metal wiring. Wiring in modern chips is 100 times thinner.

Transistors in a matchbook? We considered replacing the bad ...

2025-05-16T22:00:30Z

In reply to nevent1q…jljf
_________________________

Transistors in a matchbook? We considered replacing the bad transistor with one of these vintage Texas Instruments transistors that come in a cool package. But
@curiousmarc.bsky.social
found a different transistor that worked better. Next week I'll verify that it fixes the tape drive. 5/5

We tracked down the fault to this circuit board and found the ...

2025-05-16T21:58:40Z

In reply to nevent1q…42gm
_________________________

We tracked down the fault to this circuit board and found the transistor in the lower left had failed.
CuriousMarc (npub1vgw…46mg) replaced the bad transistor with one from his extensive collection. This board is a "hex inverter". For comparison, a 1970s-era hex inverter IC is at the right.

Unfortunately, the computer sat in an unheated garage in Germany ...

2025-05-16T21:57:58Z

In reply to nevent1q…5zxw
_________________________

Unfortunately, the computer sat in an unheated garage in Germany for many years, so many of the parts suffer from corrosion. You can see green corrosion around the base of some of these transistors. Every so often, one of the transistors will fail, causing a computer malfunction.

The IBM 1401 was the most popular computer of the early 1960s. It ...

2025-05-16T21:56:59Z

In reply to nevent1q…0shl
_________________________

The IBM 1401 was the most popular computer of the early 1960s. It has rows and rows of circuit boards—each the size of a playing card—to implement its logic. This computer predates silicon, so it is implemented with germanium transistors.

Tape drive repair at the Computer History Museum. I helped figure ...

2025-05-16T21:55:49Z

Tape drive repair at the Computer History Museum. I helped figure out why the tape drives wouldn't backspace. The problem turned out to be a bad germanium transistor on an inverter card. 1/5

The Kindle 11 uses an MT8113 processor. This is a 1 GHz dual-core ...

2025-05-10T16:46:32Z

In reply to nevent1q…vx3r
_________________________

The Kindle 11 uses an MT8113 processor. This is a 1 GHz dual-core Cortex A-7, which is 32-bit like the 386. The 386 ran at up to 40 MHz. That would put your Kindle at roughly 50 times as fast. (Apples and oranges blah blah blah.)

If this isn't enough about the 386 processor's prefetch ...

2025-05-10T16:31:50Z

In reply to nevent1q…9k0d
_________________________

If this isn't enough about the 386 processor's prefetch unit, see my latest blog post. It has more photos, schematic diagrams, and lots of explanation:
https://www.righto.com/2025/05/386-prefetch-circuitry-reverse-engineered.html

This close-up of the 386 processor's die shows the different ...

2025-05-10T16:28:35Z

In reply to nevent1q…hpuj
_________________________

This close-up of the 386 processor's die shows the different parts of the prefetch unit. There's a lot going on just to hold 16 bytes of instructions.

Another part of the prefetcher converts an 8-bit value to a 16- ...

2025-05-10T16:27:51Z

In reply to nevent1q…4kuj
_________________________

Another part of the prefetcher converts an 8-bit value to a 16- or 32-bit value. This is tricky if the value is negative, since the upper bits must be filled with 1's. Here's a close-up of the sign extension circuitry that fills upper bits with the correct value.

How hard can it be to add 1 to a number? The problem is that ...

2025-05-10T16:27:15Z

In reply to nevent1q…qvfa
_________________________

How hard can it be to add 1 to a number? The problem is that incrementing a 30-bit number can require carry the 1, carry the 1, repeated 30 times: slow. In the 386's prefetcher, Manchester carry chain and carry skip circuits make this fast, implemented with complicated dynamic logic.

The 386 processor is a 32-bit processor, so it works in 4-byte ...

2025-05-10T16:25:28Z

In reply to nevent1q…exgp
_________________________

The 386 processor is a 32-bit processor, so it works in 4-byte chunks. This network shifts bytes around so the chunk is aligned correctly. Bits go down the vertical wires, across horizontal wires, and then out the bottom. Energizing different wires provides the desired shift and lines up the bytes.

Anyone use Intel's 386 processor from 1985, the first 32-bit ...

2025-05-10T16:24:11Z

Anyone use Intel's 386 processor from 1985, the first 32-bit x86? To improve performance, it preloaded instructions into a 16-byte prefetch queue (highlighted in red). Why does such a small queue take up so much of the chip? Let's look under the microscope at its complicated circuitry... 1/N

Intel's 386 processor (1985) moved the x86 architecture to 32 ...

2025-05-01T17:42:53Z

Intel's 386 processor (1985) moved the x86 architecture to 32 bits, but it needed to be backward compatible with earlier 16 and 8-bit processors. As a result, it needed complicated circuitry for its internal registers: six different circuits for 30 registers. Let's look at the silicon circuits. 1/N

Yes, the Retro Chip Tester generated the right checksum for the ...

2025-04-13T16:42:29Z

In reply to nevent1q…32hd
_________________________

Yes, the Retro Chip Tester generated the right checksum for the ROM, but the chip failed in use. I suspect that the badly-programmed bit didn't settle on the output fast enough, causing a timing issue. We also saw one ROM address where the logic analyzer showed the right byte on the bus but the 6502 clearly read the wrong byte. This also suggests a timing issue (or maybe a voltage level issue).

For more details on this Commodore PET repair, see my blog post: ...

2025-04-13T16:23:12Z

In reply to nevent1q…wjs4
_________________________

For more details on this Commodore PET repair, see my blog post:
https://www.righto.com/2025/04/commodore-pet-repair.html
Thanks to CuriousMarc (npub1vgw…46mg)
and Tube🌞 Time (npub1xvg…62t2) for their help with this project. Thanks to Mike Naberezny for providing the PET.

To make a long story short(er), we had programmed the replacement ...

2025-04-13T16:22:17Z

In reply to nevent1q…ye43
_________________________

To make a long story short(er), we had programmed the replacement ROM with not enough voltage, so it had a couple of bad bits, enough to make the computer crash.
Finally, we got the computer running and I ran a simple graphics program, similar to ones I made when I was a kid.

Now the failures were very strange: the computer booted up and I ...

2025-04-13T16:19:29Z

In reply to nevent1q…tfzs
_________________________

Now the failures were very strange: the computer booted up and I could type in a simple "hello world" program. But when I ran it, the program printed random floating point numbers and then crashed. There was also a mysterious 2 by 2 checkerboard character near the middle of the screen.

Even with the replacement ROMs, the PET still wouldn't boot. ...

2025-04-13T16:18:44Z

In reply to nevent1q…u2ct
_________________________

Even with the replacement ROMs, the PET still wouldn't boot. Marc attached his vintage logic analyzer to the 6502 so we could trace what the processor did, byte by byte. I found that the memory test was failing on the first byte. We found three bad memory chips, but the system still failed.

We used a "Retro Chip Tester" to test the ROM chips and ...

2025-04-13T16:18:06Z

In reply to nevent1q…t82v
_________________________

We used a "Retro Chip Tester" to test the ROM chips and found two bad ROMs were producing bad signals. Unfortunately, the PET's ROMs are weird, unavailable chips, so I got two adapter boards and used normal ROMs. The adapters are in the third row of chips from the front.