Hackaday Links: February 10, 2019

Last month was NAMM, the National Association of Musical Something that begins with ‘M’, which means we’re synthed and guitarded out for the year. The synth news? Behringer are making cheap reproductions and clones of vintage gear. There’s something you need to know about vintage gear: more than half of everything produced today has a Roland 808 or 909 drum machine (or sample), a 303 bass synth, or a 101 mono synth in it. Put an 808, 909, 303, and 101 on the same table, connected to a mixer, and you can make most of the electronic music from the ’80s and ’90s. And Behringer is cloning these synths. Neat times. But there’s a problem: Roland is trademarking these drum machines and synths, with trademark filings in the US and Germany. These are ‘trade dress’, or basically the beautiful red, orange, yellow, and white buttons of the 808 and the digital cyber silver plastic aesthetic of the 303, but there you go. It’s round one in the Roland v. Behringer match, may the first person to give me an 808, 909, 303, and 101 for a thousand dollars win.

Synths? Sure thing. Here’s a stash of New Old Stock 8580 SID chips, the ‘synth on a chip’ found in the Commodore 64. The price? $50. [ben] bought one of these, and the card that came with it said,  “We purchased these chips in 2006 and they’ve been stored in our climate-controlled storage area ever since. Even still, we found a handful of them that didn’t pass testing. Treat them with care!” Yes, a bunch of SID chips for sale is noteworthy, but at $50 a piece for 1980s technology, can someone explain why a chip fab isn’t cranking these things out? If there’s one ancient piece of silicon where the demand meets what it would cost to spin up the silicon line, the SID is it. Where are the modern reproductions?

Excited about making an electronic badge this year? Seeed is offering badge sponsorships for 2019, with an offer of a 5% discount on PCBA, and a 10% discount if you put the Seeed logo on the board. I might be a little biased, but Seeed is a place where you can just ask, “hey, you guys do clear soldermask?” and they find a way to do it.

The best way to tell if someone is rich isn’t by seeing if they have an i8 parked outside their mansion, or just a piece of junk with an M badge. It isn’t whether or not their filet mignon is wagyu or just Kobe, and it isn’t if they’re cruising the skies in a G650 or just puttering around in a Cessna Citation. No, the best way to tell if someone is rich is to notice their AirPods. Yes, Apple’s wireless headphones (which are actually pretty good!) are the best foundation of a class division these days. The best class signal since private railroad cars now has a problem: people are printing their own AirPods. [Brady32] over on Thingiverse has modeled AirPods, and now the design is being given away for free. The horror. Now anyone can print out their own little bits of white plastic, stick them in their ears, and tell the world, ‘I’m better than you. Don’t bother talking back, because I obviously can’t hear you.’

Raspberry Pi has a store! Yes, everyone’s favorite single board computer now has an ‘experimental space’ in Cambridge’s Grand Arcade. The Beeb is saying this store is ‘bucking the retail trend’, yes, but any retail trend doesn’t really apply here; brands have storefronts, and it’s not about revenue per square foot. Makerbot had a store, and it wasn’t about selling printers. Microsoft has stores. Sony built a mall to advertise the original PlayStation. While the Raspberry Pi brick and mortar store will probably never make any money, it is an indication the Raspberry Pi foundation has built a valuable brand worthy of celebration. Here’s some pics of the store itself.

Did you know Hackaday has a retro edition? It’s true! retro.hackaday.com is a lo-fi version of Hackaday without CSS or Javascript or any other cruft. It’s hand-written HTML (assembled by a script) of the first ten thousand or so Hackaday posts. The idea is that old computers could load the retro site, just to prove they could. [Matthias Koch] has an Atari PC3 — Atari’s PC compatible with an 8088 running at 8Mhz, 640k of RAM and a 20 MB hard drive — and got this thing to pull up the retro site. Good work!

What is the current state of 3D printing? What is the current state of 3D printing videos on YouTube? Oh boy we’re going to find out. [Potent Printables] did an ‘analysis’ of 3D printing videos published to YouTube, and found the category riddled with ‘clickbait’, without giving an operational definition of what ‘clickbait’ actually is, or how it’s different from any other content (because who would make a video that doesn’t have the purpose of attracting viewers) Anyway, there’s a problem with the YouTube algorithm, and 3D printing blogs are copying it, filling the entire hobby with disillusioned beginners, or something. After defining ‘The Most Viewed’ as not being a news or documentary footage (okay, that’s fair), having at least three printing videos, not clickbait, and gives the designer proper attribution, [Potent Printables] found a list filled with [Maker’s Muse], [3D Print Guy] and other channels who do 3D printing work, but don’t put 3D printing in the title. This is great; 3D printing isn’t a fascinating new technology that’s the first step towards Star Trek replicators; we’ve slid down the trough of disillusionment and now 3D printers are just tools. It’s great, and in 2018 things are as they should be.

20 thoughts on “Hackaday Links: February 10, 2019

    1. Because it costs about 100k$ to get a run of silicon made. A little cheaper if you’re just filling out a mask ROM, a little more if the silicon’s entirely custom. A lot more if you have to start running multiple iterations for bug fixes. (The cost is mostly the mask and the setup, not the lithography)

      Even assuming the SID masks are found, they won’t work as-is on a modern fab – we’d have to find an old 6µ facility because everything’s changed. Even if we have the logical “masters” from which the masks are made and scale them, we can’t just scale them to a modern process node because silicon behaves differently at different sizes.

      1. 100k divided by 50 is 2000. If the assumptions are correct, it takes two thousand SID chips to break even.

        We don’t have a whole lot of data for SID chips in products, in the modern era, but we did have the Elektron SIDstation. They sold more than 2000 of those (maybe ten times that), and production of that ended in 2006.

        It blows. my. mind. no one has cloned the SID silicon yet.

        1. Regarding the SID 8580 and reproduction:
          100k divided by 50 = 2000 that part is correct. But it doesn’t stop there… there are a lot of other factors that come in to account. But practically, the first 2000 doesn’t make you any profit at all.

          The technology of the design does not apply to the technology of modern manufacturing… huge problem!
          But besides all that, even if it was all possible with the greatest ease, then I still think that the main problem is that the chip is not interesting enough for big players (considering the expected quantities in which they are to be sold) so the relatively small investment (small for a big player) doesn’t return very much profit. For a small player the investment is simply too large and with to much risk.

          What we see is that many small single person companies are making reproductions, very advanced emulators using modern parts. Perhaps technologically much more complex but cheaper to produce in small quantities and the sound quality is very good and it’s sometimes hard to hear the difference between those and an original one.

          Regarding the SID 8580, it’s older brother 6581 is much more interesting, as it’s imperfections created the most interesting sound effects. Recreating these imperfections could be a challenge.

      1. Not even close. There’s many of these systems still in active use. Some manufacturers even have OEM rebuild programs for these systems as they are still quite effective for a lot of products.

        I’m pretty sure a new run could be made. But the problem is that the price doesn’t stay at $50 a piece if you have to dump 4 to 6 thousand chips on the market. Suddenly supply outruns all demand and prices drop to $0.50 a piece. Good luck making a profit.

        Prices on these chips are high because of low supply, not because of high demand. Nobody is making new chips because there is not enough demand to warrant a full silicon run.

  1. “I made a youtube video about youtube’s algorithm pointing users to videos designed to make money”- Potent printables

    Suprise suprise, people have learnt how to manipulate youtube’s algorithm and basic psychology of human beings in order to get millions of views and thus money from ad revenue. People have been gaming algorithms since the very first search engine. People have also always tried to do the least amount of work for the most amount of money since the concept of money was invented (and the same concept existed before that except it was a matter of trying to get the best deal using the bartering system)

    I will give him credit for coming up with a list of good videos to watch but its a moot point if he also isnt trying to game the algorithms like everyone else is. The problem is that he expects youtube to better their algorithm, which will never happen as there is no incentive for youtube to do so (it would be prohibitively expensive as youtube would have to curate videos for every possible category of video). Rather than rag on youtube, he should be working with his fellow creators to understand how to game the algorithm. In the end its a battle of people who spend time to create great content versus people who spend time learning how to game youtube’s algorithm

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.