Random Access Memory From A Rotating Drum In A Bendix G15

When it’s the 1950s and you are tasked to design a computer system that features not only CPU registers but also a certain amount of RAM, you do not have a lot of options. At this point in time, discrete logic was the rule, and magnetic core memory still fairly new and rather expensive. This is where the rotating drum comes in, which is somewhat like a cross between an old-style cylinder record and a hard drive. In a recent [Usagi Electric] video, a 1950s Bendix G15 system is demonstrated, which features such a rotating drum device, alongside both tube-based circuits and newfangled diode-based circuitry.

Simplified diagram of a rotating drum random access memory unit, showing the read-erase-write process as the drum spins.
Simplified diagram of a rotating drum random access memory unit, showing the read-erase-write process as the drum spins.

This particular unit was borrowed from the System Source museum, with the intent to restore it to a working condition. Part of this process involved figuring out the circuitry, which was made easy by the circuit schematic drawings that came with the original machine. According to the official brochure by the manufacturer, the ‘short lines’ that are intended for the CPU registers, the access time was less than 1 millisecond, which is pretty darn fast considering the era and the discrete CPU’s clock speed.

For the drum itself, however, popping the cover off the unit showed that it had suffered some damage that had resulted in the multiple heads contacting the surface. Despite this disappointment, it’s not the end of the restoration, however. The museum has one more Bendix G15 standing around, with a rotating drum unit that looks to be in mint condition. The damaged magnetic coating on the other rotating drum may conceivably be resurfaced, which if successful could provide new hope to a lot of retro systems out there that also use magnetic media, whether in drum or disk format.

Continue reading “Random Access Memory From A Rotating Drum In A Bendix G15”

Debian Buzz (1.1) running under Bochs. (Credit: Thomas Stewart)

Looking Back On 30 Years Of Debian

The early history of Linux is a rather murky period to most, long before the era of glitzy marketing and proclamations of ‘the Linux desktop’ being the next hot thing. This was also the era when the first Linux distributions were born, as the Linux kernel never came as a whole OS package – unlike the BSDs – which necessitated others to package it with the elements that make up kernel and user space, such as the GNU tools.

One of these original distributions was Debian, which this month celebrates its 30th birthday. Its entire history, starting with the initial 0.01 release is covered in great detail on the Debian website. After the first release of the Linux kernel in 1991, it would take until August of 1993 when [Ian Murdock] embarked on the Debian project, sponsored by the GNU Project of the Free Software Foundation. This was a pretty rough period, with much of 1994 spent figuring out the basics of the system, the package manager and establishing a release system. Continue reading “Looking Back On 30 Years Of Debian”

A Soyuz-2.1b rocket booster with a Fregat upper stage and the Luna 25 lunar lander blasts off from a launchpad at the Vostochny Cosmodrome in Amur Oblast, Russia.

Luna 25’s Demise: Raising Fundamental Questions About Russia’s Space Program

The recent news that Russia’s Luna 25 Moon lander had made an unexpected lithobraking detour into the Moon’s surface, rather than the expected soft touchdown was met by a variety of responses, ranging from dismay to outright glee, much of it on account of current geopolitical considerations. Yet politics aside, the failure of this mission casts another shadow on the prospects of Russia’s attempts to revive the Soviet space program after a string of failures, including its ill-fated Mars 96 and Fobos-Grunt Mars missions, the latter of which also destroyed China’s first Mars orbiter (Yinghuo-1) and ignited China’s independent Mars program.

To this day, only three nations have managed to land on the Moon in a controlled fashion: the US, China, and the Soviet Union. India may soon join this illustrious list if its Chandrayaan-3 mission’s Vikram lander dodges the many pitfalls of soft touchdowns on the Moon’s surface. While Roscosmos has already started internal investigation, it does cast significant doubt on the viability of the Russian Luna-Glob (‘Lunar Sphere’) lunar exploration program.

Will Russia manage to pick up where the Soviet Union left off in 1976 with the Luna 24 lunar sample return mission?

Continue reading “Luna 25’s Demise: Raising Fundamental Questions About Russia’s Space Program”

Resin Printer Temperature Mods And Continuous IPA Filtration

Two essential parts to producing successful resin 3D prints: keeping resin at its optimal temperature and lots and lots of IPA to clean the printed parts with after printing. Unfortunately, most consumer MSLA printers do not come with a resin tray heater, and tossing out IPA after every cleaning session because of some resin contamination is both wasteful and somewhat expensive. These are two things that can be fixed in a number of ways, with [Nick Wilson] going for the ‘crank it to 11’ option, using a high-tech, fully integrated solution for both problems.

The vat with IPA is kept clean through the use of a diaphragm pump that circulates the alcohol through two filter stages, one for larger — up to 5 micrometer — particulates and one for smaller 0.5-micrometer junk. A 405 nm LED lighting section before the filters is intended to cure any resin in the IPA, theoretically leaving the IPA squeaky clean by the time it’s returned to the vat.

For the resin tray heater, a more straightforward 12V 150 Watt silicone heater strip is stuck to the outside edge of the metal resin tray, along with a temperature-controlled relay that toggles the heater strip on and off until the resin reaches the desired temperature. None of these are necessarily expensive solutions, but they can be incredibly useful if you do a fair amount of resin printing.

Continue reading “Resin Printer Temperature Mods And Continuous IPA Filtration”

The Clathrate Gun Hypothesis: Unearthing Puzzles Of Warming Events Past

As the Earth continues to warm at a worrying rate, scientists continue to work to understand the processes and mechanisms at play. Amidst the myriad of climate-related theories and discussions, the clathrate gun hypothesis stands out not only for its intriguing name but for the profound implications it might have on our understanding of global warming events.

Delving into this hypothesis is akin to reading a detective novel written by Mother Earth, with clues hidden deep beneath the ocean and Arctic ice. It’s a great example of how scientists attempt to predict the future by unpicking the mysteries of the past.

Continue reading “The Clathrate Gun Hypothesis: Unearthing Puzzles Of Warming Events Past”

An Odd Home Computer From The 1980s

If you were around when the Altair 8800 was king, you might remember the name Cromemco. They were an early vendor of add-ons for the Altair, along with companies like Godbout and Morrow. The company was mostly famous for a very crude digital camera for the Altair and a similarly-crude graphics interface card. They graduated into building S-100 bus computers. Like many similar companies, they could taste the upcoming home PC market, and they wanted a piece of it. Their answer? The $1,800 C-10 Cromemco Personal Computer, and you can see [Vintage Geek’s] thoughts on the odd machine in the video below.

The system ran CP/M and, like many similar systems, got lost in the rush to get the IBM PC. Compared to other computers of the time, the C-10 was compact. The keyboard layout seems odd today, but there wasn’t really much standardization in those days.

Continue reading “An Odd Home Computer From The 1980s”

A 6502 Overlay Debugger

Retired hardware engineer [Plasmode] recently took on the challenge of building a debugger for the 6502 designed to sit atop the microprocessor while seated in a solder less breadboard. The result is the Diagnostic Overlay for W65C02 Breadboard, consisting of 128 kB SRAM and a 1250-gate CPLD. Except being 0.8 in wide, the overlay debugger is otherwise the same size as the 6502’s 40-pin DIP package, so it doesn’t overhang other portions of your circuit.

Being an initial concept prototype, [Plasmode] mounted the chips dead-bug style on perf board — a process he himself found tiring. If he builds additional debuggers, presumably he will consider making a PCB.

The prototype was constructed using point-to-point soldering with 30-ga wire wrap wire.  It was all done under the inspection microscope.  There are not many connections, but they are rather tedious so I can only do a dozen or so wires per session.  It took me 2 days and several hours total to finish the prototype board.

This design is based on the CRC65 Frugal 6502 Single Board Computer, of course omitting the 6502 itself. Instead of a physical ROM memory chip, he implemented a 64-byte boot loader inside the CPLD and a serial port. This lets him to bootstrap the system over the serial port. He plans on expanding this to include other DIP-packaged retro microprocessors in the future. Check out his Hackaday.io project page ( above ). If you want to dig deeper, he posted the schematics here.