PC AT mainboard with both 16-bit ISA and 32-bit PCI slots. (Credit: htomari, Flickr)

How Intel Gave Us The PCI Bus While Burying VESA’s VL-Bus

Gigabyte GA486IM mainboard from 1994 with ISA, VLB and PCI slots. (Credit: Rjluna2, Wikimedia)
Gigabyte GA486IM mainboard from 1994 with ISA, VLB and PCI slots. (Credit: Rjluna2, Wikimedia)

The early days of home computing were quite a jungle of different standards and convoluted solutions to make one piece of hardware work on as many different platforms as possible. IBM’s PC was an unexpected shift here, as with its expansion card-based system (retroactively called the ISA bus) it inspired a new evolution in computers. Of course, by the early 1990s the ISA bus couldn’t keep up with hardware demands, and a successor was needed. Many expected this to be VESA’s VLB, but as [Ernie Smith] regales us in a recent article in Tedium, Intel came out of left field with its PCI standard after initially backing VLB.

IBM, of course, wanted to see its own proprietary MCA standard used, while VLB was an open standard. One big issue with VLB is that it isn’t a new bus as such, but rather an additional slot tacked onto the existing ISA bus, as it was then called. While the reasoning for PCI was sound, with it being a compact, 32-bit (also 64-bit) design with plug and play and more complex but also more powerful PCI controller, its announcement came right before VLB was supposed to be announced.

Although there was some worry that having both VLB and PCI in the market competing would be bad, ultimately few mainboards ended up supporting VLB, and VLB quietly vanished. Later on PCI was extended into the Accelerated Graphics Port (AGP) that enabled the GPU revolution of the late 90s and still coexists with its PCIe successor. We covered making your own ISA and PCI cards a while ago, which shows that although PCI is more complex than ISA, it’s still well within the reach of today’s hobbyist, unlike PCIe which ramps up the hardware requirements.

Top image: PC AT mainboard with both 16-bit ISA and 32-bit PCI slots. (Credit: htomari, Flickr)

Harbor Freight And LEGO PCB Vise Is Cheap And Effective

It doesn’t take much chasing things around the bench with a soldering iron to appreciate the value of good work holding. And don’t get us started on those cheap “helping hands” alligator clip thingies; they’re somehow worse than no work holding. Isn’t there a better way?

Maybe, judging by [Paul Bryson]’s idea for a dirt cheap PCB vise. It’s a pretty clever design that’ll have you heading to Harbor Freight, or whatever the moral equivalent is in your location, where you’ll pick up a small ratcheting bar clamp. [Paul] used a 4″ (10 cm) clamp; that which looks fine for a wide range of boards, but we suppose you could go bigger if you like. You could also stop there and just clamp your PCBs in the plastic jaws, but [Paul] adorned the jaws with swiveling arms made from LEGO Technic pieces, of all things. Rubber grommets slipped onto Technic pegs go into the holes on the beam to hold the PCB edges firmly, while the swiveling action adapts to odd-shaped boards.

To our mind, the biggest advantage to this design other than cost is how low it holds the PCB — a decided advantage while working under the microscope. Don’t have any Technics parts close to hand? No worries, 3D printed parts could easily stand in, and maybe even improve the design. [Paul] also shows off a substitute for the Technics beam rendered in PCB material, which would reduce the height of the workpiece over the bench even more.

We’ve seen a lot of PCB vises come and go, using everything from scrap wood to 3D printed compliant mechanisms. But we doubt you’ll find anything more cost-effective than [Paul]’s design.

Fail Of The Week: PCB LED Cube Fails Successfully

Remember LED cubes? We sure do — they were all the rage for a while, and then it seemed like everyone just sort of lost interest in them. There are probably a lot of reasons for that, not least of which is likely the amount of work it takes to put one together from discrete LEDs and separate pieces of wire. Could there be a better way?

Of course there could, and [Sasa Karanovic] thought he had it all figured out with this PCB-based LED cube. At first glance, it seems to make perfect sense; after all, weren’t PCBs invented to take the place of all that pesky point-to-point wiring in the early days of electronics? The boards [Sasa] designed are pretty cool, actually. They’ve each got room for 16 addressable WS2812 LEDs in 5 mm packages, with every possible bit of substrate removed to block the minimum amount of light. That left very little room for traces on the 2-mm-wide arms, so the PCBs had to have four layers, which raised eyebrows at the PCB house when [Sasa] submitted the design.

Such an airy and open design obviously has the potential for mechanical issues, which [Sasa] addressed by adding pads at three corners of each board; a vertical PCB connects to each LED board to provide mechanical support and distribute signals to the LEDs. The cube seems solid enough as a result, and even when handled the LED boards don’t really flop around too much. See the cube in action in the video below.

What’s nice about this design is the perfect spacing between the LEDs in all three dimensions, and the way everything lines up nice and straight. That would be really hard to do with wire, even for the most practiced of circuit sculptors. [Sasa] seems to agree, but still deems the build a failure because the PCBs block too much of the view. We suppose he’s got a point, and we’re not sure how well this would scale to an 8×8 cube. We’re not sure how we’d feel about paying for PCBs that are mostly air either, but as failures go, this one still manages to be pretty successful. Continue reading “Fail Of The Week: PCB LED Cube Fails Successfully”

Cheap Hack Gets PCI-X Card Working In PCI Slot

PCI and PCI-X are not directly compatible, and you’d be forgiven for thinking that means you’re out of luck if you need to use a PCI-X card in a machine that only has basic PCI slots. And yet, that needn’t be the case. As [Peter] shows us, you can work around this with a cheap hacky hack. Our favorite kind!

[Peter] had a PCI-X RAID card that he wanted to use on his Socket 7-based computer. The 3ware 9550SX PCI-X card is 3.3 V only, and doesn’t fit in a typical PCI slot. It’s not compatible mechanically or electrically. Enter a PCI-X riser, which gets around the missing notch that would normally not let the card sit in the slot. Other than that, it just took masking off some pins to avoid damage from the 5 V rail. Throughput is good, too, reportedly sitting at roughly 60-70 MB/s.

The hard part is probably finding a PCI-X riser; PCI-Express stuff is far more common. Few of us need to deal with PCI-X anymore, but if you’re working on some ancient industrial hardware or something, this hack might just save your beans from the roast pot one day.

Continue reading “Cheap Hack Gets PCI-X Card Working In PCI Slot”

High Caliber Engineering On A Low Torque PCB Servo Motor

Building a 3D motor printed motor is one thing, but creating a completely custom servo motor with encoder requires some significant engineering. In the video after the break [365 Robots] takes us through the build process of a closed-loop motor with a custom optical encoder.

The motor, an axial flux design, uses a stack of 0.2mm PCBs with wedge shaped coils clamped in a 3D printed body. It’s similar to some of the other PCB motors we’ve featured, but what really sets this build apart is its custom optical encoder, which was a project in its own right. The 4-bit absolute position encoder uses IR LEDs to shine through an PCB disc with concentric gray code copper encoder rings onto IR receivers. This works because FR4, the composite material used in PCBs doesn’t block IR light.

The motor’s body was printed from ABS to withstand the heat during operation. [365 Robots] didn’t skimp on the testing either, creating a 3D printed closed-loop test stand with load cell and Arduino. Like other PCB motors it produces very little torque, roughly 2% of a typical NEMA17 stepper motor. Even so, the engineering behind this project remains impressive.

Continue reading “High Caliber Engineering On A Low Torque PCB Servo Motor”

3D Mouse With 3D Printed Flexures And PCB Coils

3D mice with six degrees of freedom (6DOF) motion are highly valued by professional CAD users. However, the entry-level versions typically cost upwards of $150 and are produced by a single manufacturer. [Colton Baldridge] has created the OS3M Mouse — an open source alternative using PCB coils and 3D printed flexures.

The primary challenges in creating a 6DOF input device, similar to the 3Dconnexion Space Mouse, lie in developing a mechanical coupling that enables full range motion, and electronics capable of precisely and consistently measuring this motion. After several iterations of printed flexure combinations and trip down the finite element analysis (FEA) rabbit hole, [Colton] had a working single-piece mechanical solution.

To measure the knob’s movement accurately, [Colton] employs inductive sensing. Inductance to Digital Converters (LDCs) assess the inductive alterations across three pairs of PCB coils, each having an opposing metal disk mounted on the knob. This setup allows [Colton] to use a Stewart platform‘s kinematic model calculate theĀ  knob’s relative position. The calculation are done on an STM32 which also acts USB HID send the position data to a computer. For the demo [Colton] created a simple C++ app to translate the position data to Solidworks API calls.

Continue reading “3D Mouse With 3D Printed Flexures And PCB Coils”

Brand-New PCB Makes Replica TRS-80 Possible

If like us, you missed out on the TRS-80 Model I back when it first came out, relax .With this brand-new PCB that’s a trace-for-trace replica of the original and a bunch of vintage parts, you can build your own from scratch.

Now, obviously, there are easier ways to enjoy the retro goodness that is the 46-year-old machine that in many ways brought the 8-bit hobby computing revolution to the general public’s attention. Sadly, though, original TRS-80s are getting hard to come by, and those that are in decent enough shape to do anything interesting are commanding top dollar. [RetroStack]’s obvious labor of love project provides the foundation upon which to build a brand new TRS-80 as close as possible to the original.

The PCB is revision G and recreates the original in every detail — component layout, connectors, silkscreen, and even trace routing. [RetroStack] even replicated obvious mistakes in the original board, like through-holes that were originally used to fixture the boards for stuffing, and some weird unused vias. There are even wrong components, or at least ones that appear on production assemblies that don’t show up in the schematics. And if you’re going to go through with a build, you’ll want to check out the collection of 3D printable parts that are otherwise unobtainium, such as the bracket for rear panel connectors and miscellaneous keyboard parts.

While we love the devotion to accuracy that [RetroStack] shows with this project, we know that not everyone is of a similar bent. Luckily there are emulators and clones you can build instead. And if you’re wondering why anyone would devote so much effort to half-century-old technology — well, when you know, you know.

Thanks to [Stephen Walters] for the tip.

Feature image: Dave Jones, CC BY-SA 4.0, via Wikimedia Commons