Designing your own integrated circuits as a one-person operation from your home workshop sounds like science fiction. But 20 years ago, so did rolling your own circuit boards to host a 600 MHz microcontroller with firmware you wrote yourself. Turns out silicon design isn’t nearly as out of reach as it used to be and Matt Venn shows us the ropes in his Zero to ASIC workshop.
Held during the 2020 Hackaday Remoticon, this is a guided tour of the tools used in the Skywater PDK — the Process Design Kit that is an open-source ASIC toolkit produced in a partnership between Google and SkyWater Technology. We covered the news when first announced back in June, but this the most comprehensive look we’ve seen into the actual design process.
Matt builds up the demo starting from the very simple design of an N-channel MOSFET with click-and-drag tools similar to graphics editing software. The good news it that although you can draw your own structures like this, for digital designs you won’t have to. A wide variety of IP has been contributed to the open source project allowing basic building blocks to be pulled in using HDL. However, the power of drawing structures will certainly be the playground for those needing analog design as part of their projects.
As with EDA software used for circuit boards, the PDK includes design rule checks to ensure you aren’t violating the limits of the 130 nm chip fab. There’s some other black magic in there too, as Matt specifically mentions an antenna rules check to safeguard your design from being fried by induced current on “large” (microscopically so) metalized runs during the fabrication process.
The current workflow involves grinding through a large number of configuration files, something Matt admits took him a long time to wrap his head around. However, what’s available for proofing your design is very impressing. He demonstrates SPICE simulation to calculate timings, and shows numerous examples of verification drawings generated by the compilation process, either in the form of seeing the structures as they will be laid out, or as logical flow charts. This is crucial as a single run will take 2-3 months to come back from fab — you want to get things right before buttoning up the project. Incidentally, that’s know as “tapeout”, a term you’ve likely heard before and he says it comes from reels of magnetic tape containing the design being removed from the computer and sent to production. Who knew? (This tidbit in strikethrough appears to be incorrect).
But wait, there’s more to this than just designing the things. Part of the intrigue of the Skywater-PDK project is that Google bought into covering a group run about once per quarter so that open-source designs can be ganged onto a multi-project wafer free of charge to the people submitting them. That’s pretty awesome and we’re giddy to hear news of people getting their wafer-level chip scale devices — also known as flip chips — back for testing. Matt is planning a more in-depth paid course on the topic. For now, get a taste of what’s involved from this excellent workshop found after the break.
For anyone serious about mining cryptocurrency such as Bitcoin, we’re well past the point where a standard desktop computer is of much use. While an array of high-end GPUs is still viable for some currencies, the real heavy hitters are using custom mining hardware that makes use of application-specific integrated circuits (ASICs) to crunch the numbers. But eventually even the most powerful mining farm will start to show its age, and many end up selling on the second hand market for pennies on the dollar.
According to [xjtuecho], it takes a little bit of work to get the EBAZ4205 ready for experimentation. For one thing, you may have to solder on your own micro SD slot depending on where you got the board from. You’ll also need to add a couple diodes to configure which storage device to boot from and to select where the board pulls power from.
Once you’re done, you’ll have a dual core Cortex A9 Linux board with 256 MB DDR3 and a Artix-7 FPGA featuring 28K logic elements to play with. Where you go from there is up to you.
This isn’t the first time we’ve seen FPGA boards hit the surplus market at rock bottom prices. When IT departments started dumping their stock of Pano Logic thin clients back in 2013, a whole community of dedicated FPGA hackers sprouted up around it. We’re not sure the if the EBAZ4205 will enjoy the same kind of popularity in its second life, but the price is certainly right.
[Jean-Francois Debroux] spent 35 years designing analog ASICs. He’s started a book and while it isn’t finished — indeed he says it may never be — the 180 pages he posted on LinkedIn are a pretty good read.
The 46 sections are well organized, although some are placeholders. There are sections on design flow and the technical aspects of design. Examples range from a square root circuit to a sigma-delta modulator, although some of them are not complete yet. There are also sections on math, physics, common electronics, materials, and tools.
You might have caught Maya Posch’s article about the first open-source ASIC tools from Google and SkyWater Technology. It envisions increased access to make custom chips — Application Specific Integrated Circuits — designed using open-source tools, and made real through existing chip fabrication facilities. My first thought? How much does it cost to tape out? That is, how do I take the design on my screen and get actual parts in my hands? I asked Google’s Tim Ansel to explain some more about the project’s goals and how I was going to get my parts.
The goals are pretty straightforward. Tim and his collaborators would like to see hardware open up in the same way software has. The model where teams of people build on each other’s work either in direct collaboration or indirectly has led to many very powerful pieces of software. Tim’s had some success getting people interested in FPGA development and helped produce open tools for doing so. Custom ASICs are the next logical step.
A process design kit (PDK) is a by now fairly standard part of any transformation of a new chip design into silicon. A PDK describes how a design maps to a foundry’s tools, which itself are described by a DRM, or design rule manual. The FOSSi foundation now reports on a new, open PDK project launched by Google and SkyWater Technology. Although the OpenPDK project has been around for a while, it is a closed and highly proprietary system, aimed at manufacturers and foundries.
The SkyWater Open Source PDK on Github is listed as a collaboration between Google and SkyWater Technology Foundry to provide a fully open source PDK and related sources. This so that one can create manufacturable designs at the SkyWater foundry, that target the 130 nm node. Open tools here should mean a far lower cost of entry than is usually the case.
Although a quite old process node at this point (~19 years), it should nevertheless still be quite useful for a range of applications, especially those that merge digital and analog circuitry. SkyWater lists their SKY130 node technology stack as:
Support for internal 1.8V with 5.0V I/Os (operable at 2.5V)
1 level of local interconnect
5 levels of metal
High sheet rho poly resistor
Optional MiM capacitors
Includes SONOS shrunken cell
Supports 10V regulated supply
HV extended-drain NMOS and PMOS
It should be noted that use of this open source PDK is deemed experimental at this point in time, and should not be used for any commercial or otherwise sensitive applications.
Typically when we select a project for “Fail to the Week” honors, it’s because something went wrong with the technology of the project. But the tech of [Leo Fernekes]’ innovative LED sign system was never the problem; it was the realities of scaling up to production as well as the broken patent process that put a nail in this promising project’s coffin, which [Leo] sums up succinctly as “The Inventor’s Paradox” in the video below.
The idea [Leo] had a few years back was pretty smart. He noticed that there was no middle ground between cheap, pre-made LED signs and expensive programmable signboards, so he sought to fill the gap. The result was an ingenious “LED pin”, a tiny module with an RGB LED and a microcontroller along with a small number of support components. The big idea is that each pin would store its own part of a display-wide animation in flash memory. Each pin has two terminals that connect to metal cladding on either side of the board they attach to. These two conductors supply not only power but synchronization for all the pins with a low-frequency square wave. [Leo]’s method for programming the animations — using a light sensor on each pin to receive signals from a video projector — is perhaps even more ingenious than the pins themselves.
[Leo]’s idea seemed destined for greatness, but alas, the cruel realities of scaling up struck hard. Each prototype pin had a low part count, but to be manufactured economically, the entire BOM would have to be reduced to almost nothing. That means an ASIC, but the time and expense involved in tooling up for that were too much to bear. [Leo] has nothing good to say about the patent game, either, which his business partners in this venture insisted on playing. There’s plenty of detail in the video, but he sums it up with a pithy proclamation: “Patents suck.”
Watching this video, it’s hard not to feel sorry for [Leo] for all the time he spent getting the tech right only to have no feasible way to get a return on that investment. It’s a sobering tale for those of us who fancy ourselves to be inventors, and a cautionary tale about the perils of participating in a patent system that clearly operates for the benefit of the corporations rather than the solo inventor. It’s not impossible to win at this game, as our own [Bob Baddeley] shows us, but it is easy to fail.
Join us on Wednesday, December 18 at noon Pacific for the Weird World of Microwaves Hack Chat with Shahriar Shahramian! We’ve been following him on The Signal Path for years and are excited to pick his brain on what is often considered one of the dark arts of electronics.
No matter how much you learn about electronics, there always seems to be another door to open. You think you know a thing or two once you learn about basic circuits, and then you discover RF circuits. Things start to get a little strange there, and stranger still as the wavelengths decrease and you start getting into the microwave bands. That’s where you see feed lines become waveguides, PCB traces act as components, and antennas that look more like musical instruments.
Shahriar is no stranger to this land. He’s been studying millimeter-wave systems for decades, and his day job is researching millimeter-wave ASICs for Nokia Bell Labs in New Jersey, the birthplace of the transistor. In his spare time, Shahriar runs The Signal Path, a popular blog and YouTube channel where he dives tear-downs, explanations, and repairs of incredibly sophisticated and often outrageously expensive equipment.
We’ll be sitting down with Shahriar this week for the last Hack Chat of 2019 with a peek inside his weird, wonderful world of microwaves. Join us with your questions about RF systems, microwaves in the communication industry, and perhaps even how he manages to find the gear featured on his channel.
Click that speech bubble to the right, and you’ll be taken directly to the Hack Chat group on Hackaday.io. You don’t have to wait until Wednesday; join whenever you want and you can see what the community is talking about.