Could Seven Gamers Play Quake On Just One 1996 SGI Machine?

May 1, 2021 by Lewin Day 46 Comments

[Linus Tech Tips] undertook a fun experiment a few years back. By running multiple virtual machines on a single tower PC with tons of RAM and GPUs, it was possible to let seven gamers play on a single rig at once. [CelGenStudios] found the idea intriguing, and has theorised that the same feat could be possible on mid-1990s Silicon Graphics hardware.

The idea is to use the Origin 2000 server as the base. These didn’t ship with any form of video output or even a keyboard and mouse interface. However, by substituting in the IO6G module from the Onyx2 machine, and SI graphics cards from the Octane, it’s possible to get graphics and input up and running. With multiple graphics cards and a few CAD DUO boards installed via a PCI adapter called the “shoebox”, there’s provisions for up to four separate monitors, keyboards and mice. With all this hardware, it’s theoretically possible for four users to login to the X server running in the IRIX OS on the Origin 2000 machine. Then, it’s a simple matter of firing up four instances of Quake and a dedicated server and you’re up and gaming.

[CelGenStudios] goes so far as exploring the limits of the supercomputer-grade hardware, suggesting that 7 players or more could be possible. Unfortunately, SGI hardware isn’t easy to come by, nor is it cheap, even decades after release — so thus far, the concept remains untested. We’d dearly love to see such a setup happen at QuakeCon or a hacker con, though, so if you pull it off, you know how to call. We note there’s a few Octane 2000s at the Jim Austin Computer Collection, so perhaps they might be the ones to achieve the feat.

In the meantime, check out a practical exploration of the concept on modern hardware with the original [Linus Tech Tips] project. The basic theory is simple – create a hugely powerful PC, with a beefy CPU, plenty of RAM, and one graphics cards for each of the seven players. They run multiple virtual machines and managed to deliver a full 7 player experience running off just one CPU.

Continue reading “Could Seven Gamers Play Quake On Just One 1996 SGI Machine?” →

If You Can Measure It, You Must Display It

May 1, 2021 by Elliot Williams 16 Comments

When can you be sure that you’re logging enough data? When you’re logging all of the data! Of course there are exceptions to the above tongue-in-cheek maxim, but it’s certainly a good start. Especially since data storage on, for instance, an SD card is so easy and cheap these days, there’s almost no reason to not record most every little bit of data that your project can produce. Even without an SD card, many microcontrollers have enough onboard flash, or heck even RAM, to handle whatever you throw at them. The trick, then, is to make sense out of that data, and for me at least, that often means drawing pretty pictures.

I was impressed this week by a simple but elegant stepper motor diagnosis tool hacked together by [Zapta]. Essentially, it’s a simple device: it’s a “Black Pill” dev board, two current sensors, an EEPROM for storing settings, and a touchscreen. Given that most of us with 3D printers rely on stepper motors to get the job done, it’s certainly interesting to do some diagnostics.

By logging voltage and current measurement on each phase of a stepper motor, you can learn a lot about what’s going on, at least if you can visualize all that data. And that’s where [Zapta]’s tool shines. It plots current vs motor speed to detect impedance problems. Tuning the current in the first place is a snap with Lissajous patterns, and it’ll track your extruder’s progress or look out for skipped steps for you across an entire print job. It does all this with many carefully targeted graphs.

I was talking to [Niklas Roy] about this, and he said “oh check out my hoverboard battery logger“. Here we go again! It sits inline with the battery and logs current and voltage, charging or discharging. Graphs let you visualize power usage over time, and a real-time-clock lets you sync it with video of using the hoverboard to help make even more sense of the data.

So what are you waiting for? Sensors are cheap, storage is cheap, and utilities to graph your data after the fact are plentiful. If you’re not logging all the relevant data, you’re missing out on some valuable insights. And if you are, we’d love to see your projects! (Hint, hint.)

3D Printer Control Board Packs A Raspberry Pi Compute Module 4

May 1, 2021 by Tom Nardi 51 Comments

Traditionally, 3D printer control boards have used simplistic 8-bit microcontrollers to command the stepper drivers and ultimately move the machine where it needs to go. Newer boards have switched over to 32-bit microcontrollers, but they’re still relatively limited computationally. Because of this, a Raspberry Pi running OctoPrint is usually used to provide more complex features such as remote management and live video.

Looking to combine these different devices into a single all-in-one board, [pkElectronics] developed the Sigmoid S7P. With an STM32 microcontroller, TMC2209 stepper drivers, a Raspberry Pi Compute Module 4, and plenty of room for expansion, it promises to be a drop-in upgrade for essentially any 3D printer running on an open source firmware that could be ported over.

According to [pkElectronics], the idea for the Sigmoid had been floating around for several years, but never got off the ground due to the difficulties in dealing with the SO-DIMM interface used by previous iterations of the Compute Module. But with the switch to smaller and denser connector for the CM4, the board finally started to take shape.

Whether you just used it as a convenient way to integrate OctoPrint into your printer, or want to get into something more advanced like Klipper, the Sigmoid S7P looks like a very exciting project. [pkElectronics] says they are considering producing the board commercially if there’s interest, so if you want one of these for your own custom 3D printer build, let them know.

Apple-Picking Robot Stems From Labor Shortage

May 1, 2021 by Kristina Panos 35 Comments

Among all the job-related problems wrought by the pandemic, here is another one that comes as the result of people generally staying home: there are hardly any backpackers to do traditional transient backpacker jobs like picking apples. Researchers at Monash University’s Department of Mechanical and Aerospace engineering found a way to fill in the gap by building a pneumatic robot arm that can harvest an apple every seven seconds at top speed.

A suite of cameras and algorithms look for fruit amongst the foliage and carefully remove it by gripping it gently and twisting, much like a human would. In order to do this, the robot must consider the shape of the fruit, the way it’s hanging, and where to separate it from the tree while keeping damage to a minimum. A suction system helps pull the apple into the soft, four-fingered grip and then the arm twists and turns to deposit the apple into the bin.

There are a lot of upsides to this robot, including the fact that it works in any lighting and weather conditions and can ID an apple in less than 200 milliseconds. The only problem is that this operation results in the occasional missing stem — a cosmetic problem that sounds nit-picky, but would definitely prevent some stores from buying the fruit. Well, that, and there only seems to be one of these robots so far.

There are two videos after the break — a short one that gives you the gist, and a much longer one that offers a view of the suction cup, which emerges from the middle of the fingers like a xenomorph’s little mouth.

Some readers may be wondering why apples are still picked individually when shaking harvesters exists. “Shake-and-catch” tends to bruise apples, making them undesirable for produce sellers, however, apples destined for juicing have no issue with being handled roughly by the harvesters as shown in this fascinating harvest video. Robot grippers are gentle and we’ve seen all shapes and sizes that are suited to a particular need. When the needs are more general, rollers or squishy spheres might be the answer.

Continue reading “Apple-Picking Robot Stems From Labor Shortage” →

Modding A Casio W800-H With A Countdown Timer

April 30, 2021 by Lewin Day 28 Comments

Stock, the Casio W800-H wristwatch ships with dual time modes, multiple alarms, and a stopwatch – useful features for some. However, more is possible if you just know where to look. [Ian] decided to dive under the hood and enable a countdown timer feature hidden from the factory.

The hack involves popping open the case of the watch and exposing the back of the main PCB. There, a series of jumpers control various features. [Ian]’s theory is that this allows Casio to save on manufacturing costs by sharing one basic PCB between a variety of watches and enabling features via the jumper selection. With a little solder wick, a jumper pad can be disconnected, enabling the hidden countdown feature. Other features, such as the multiple alarms, can be disabled in the same way with other jumpers, suggesting lower-feature models use this same board too.

It’s a useful trick that means [Ian] now always has a countdown timer on his wrist when he needs it. Excuses for over-boiling the eggs will now be much harder to come by, but we’re sure he can deal. Of course, watch hacks don’t have to be electronic – as this custom transparent case for an Apple Watch demonstrates. Video after the break.

Continue reading “Modding A Casio W800-H With A Countdown Timer” →

VGA Graphics Card In 74xx Logic

April 30, 2021 by Chris Lott 17 Comments

Feeling nostalgic we presume, [Glen Kleinschmidt] set out to build a 640x480x64 VGA controller card from discrete logic chips. If we ignore the 512Kx8 Cypress SRAM video memory, he succeeds, too — and on a very readable, single page A3 schematic. The goal is to interface some of his older 8-bit machines, like the TRS-80 Model 1 and the BBC Micro, but for now he’s running a demo using a 20+ year old PIC16F877 micro.

[Glen] provides all the schematics, Gerbers, and C source code on his website should you be inclined to reproduce one for yourself. He has three versions in the works, with various capabilities (there’s a table on his website). As an alternative, one could always use an FPGA or a custom-built chip such as the SSD1963 to generate video for these micros, but sometimes the urge to go retro is too great to resist. We get the feeling that for [Glen], this is a project unto itself, and being able to interface it to his 8-bit computers is just a convenient excuse.

This isn’t [Glen]’s first retro project, either. Check out his analog computer “bouncing ball” project we covered back in 2017. Have you struggled with the build vs. buy decision, and how do you decide?

Continue reading “VGA Graphics Card In 74xx Logic” →

Look Out Below! China’s Heavy-Lift Rocket Due For Uncontrolled Reentry Within Days

April 30, 2021 by Tom Nardi 57 Comments

On April 28th, China successfully put the core module of their Tianhe space station into orbit with the latest version of the Long March 5B heavy-lift booster. This rocket, designed for launching large objects into low Earth orbit, is unique in that the 33.16 m (108.8 ft) first stage carries the payload all the way to orbit rather than separating at a lower altitude. Unfortunately, despite an international effort to limit unnecessary space debris, the first stage of the Long March 5B booster is now tumbling through space and is expected to make an uncontrolled reentry sometime in the next few days.

The massive booster has been given the COSPAR ID 2021-035-B, and ground tracking stations are currently watching it closely to try and determine when and where it will reenter the Earth’s atmosphere. As of this writing it’s in a relatively low orbit of 169 x 363 km, which should decay rapidly given the object’s large surface area. Due to the variables involved it’s impossible to pinpoint where the booster will reenter this far out, but the concern is that should it happen over a populated area, debris from the 21 metric ton (46,000 pound) booster could hit the ground.

This is the second launch for the Long March 5B, the first taking place on May 5th of 2020. That booster was also left in a low orbit, and made an uncontrolled reentry six days later. During a meeting of the NASA Advisory Council’s Regulatory and Policy Committee, Administrator Jim Bridenstine claimed that had the rocket reentered just 30 minutes prior, debris could have come down over the continental United States. Objects which were suspected of being remnants of the Long March 5B were discovered in Africa, though no injuries were reported.

China’s first space station, Tiangong-1, made an uncontrolled reentry of its own back in 2018. It’s believed that most of the 8,500 kg (18,700 lb) burned up as it streaked through the atmosphere, and anything that was left fell harmlessly into the South Pacific Ocean. While small satellites are increasingly designed to safely disintegrate upon reentry, large objects such as these pose a more complex problem as we expand our presence in low Earth orbit.