Commander X16

codewar65 wrote: ↑Sat Apr 01, 2023 3:05 am I got the weirdest vibs from David's latest video. Not sure how to read this. Neither one seem happy, but both seem to need to address the we/us.

The Devil is in the details. Pray for final perseverence. Most of us know the first 90% of a project is EZ. It's that last 90% that's the killer. Finish stronk.

BruceRMcF wrote: ↑Sun Apr 02, 2023 1:09 am And when the game console version comes out, with mostly surface mount parts ... indeed, given that OPM parts are not as abundant as some in the design team seems to have thought, even fewer than had previously been though ... $150-$175 seems to me to be a reasonable price.

Yeah, I can see that price if they insist on keeping the "real CPU" but at least ditch some of the support chips in favor of an FPGA that also serves as the Yamaha sound chip. And considering the total beating they're taking just trying to get those... I think the FPGA replacement argument will fly. There is an old saying... "You can't change someone's mind in real time." You need an event like being sold a hundred blatantly fake chips to do that.

As to the 50 dollar price: Yeah, that was a mistake. But hey, anytime you try to do new stuff you make mistakes. You just have to dust yourself off, unapologetically own your errors, and move on.

At the same time, I can't blame the people that believed the 50 dollar price... we live in a world where it IS possible to get Linux computers based on a SOC for that or even less. But that price doesn't exist in a vacuum, it exists because of demand (Millions of units.) and the modern zeitgeist of contract manufacturers like TSMC and the commodification of high volume manufacturing. These people do a simple capability comparison and then just assume that something primitive from the 80's just HAS to be cheap to make now.

What they don't see is that primitive stuff from the 80's was manufactured on primitive equipment that only existed in small amounts due to the relatively low volume of demand in the 80's (As compared to now.) No one is going to manufacture that obsolete equipment now, and even old fabs are being pressed to make what they can, as the new fabs that come on line focus on new, denser designs that are more profitable. The result? You can't make old chips like the Yamaha sound chip even if you wanted to. You have to make new chips... and everything in audio is digital now.

This is why I think FPGAs are the future. Even if ... say... the Lattice Ice40up5K used in the VERA goes the way of the Dodo and is no longer manufactured, another FPGA will pop in to fill that niche in the FPGA food chain. And! As a bonus, it will probably be better and faster. Sure, you have to fix the board to use the new hotness, but the HDL that drives it will be largely if not completely compatible.

Daedalus wrote: ↑Sun Apr 02, 2023 2:14 am ... Yeah, I can see that price if they insist on keeping the "real CPU" but at least ditch some of the support chips in favor of an FPGA that also serves as the Yamaha sound chip. ...

They were already alluding to consideration of replacing the glue logic that provides the chip select logic with a CPLD ... those are parts that add more substantially more to the build cost than to the BOM ... and extending that to an FPGA including the OPM chip implementation seems a straightforward step.

Wavicle wrote: ↑Sat Apr 01, 2023 9:07 pm

Daedalus wrote: ↑Sat Apr 01, 2023 4:16 pm Back to the "Vintage Through Hole vs Surface Mount" debate: Surface mount isn't even that hard to do for the hobbyist. Sure, you need to upgrade your gear... in addition to a soldering iron, you need a hot air rework station and a hot plate of any stripe (I've done simple boards with an electric skillet I bought at Target for 35 bucks.) If you want to do development for FPGAs, there are open source solutions you can use on a PC (I recommend a Linux based computer for that.) that you need to download and learn, but it's no harder than learning to program. ... And you needed to know how to program anyway just to be able to implement new ideas on ANY platform.

I would go so far as to say as long as you aren't doing the tiny components, and don't have significant issues with steady hands, surface mount is faster and easier for hobbyists. There are a few skills and techniques to learn, and you'll probably never long for THT again. I taught my 20-year-old daughter to assemble both my Mini-ITX all-through-hole X16 and the SMT VERAs 6 months ago. She prefers surface mount assembly by a wide margin. When I had the new YM2151 FPGA boards, she asked if I needed help assembling them. She didn't ask that same question for my last few through hole boards.

The only thing that I prefer through hole for is board rework. It isn't much fun trying to solder a bodge wire to a 0.25mm pad which is 0.25mm away from another 0.25mm pad. I can do it, but it takes a few attempts and often leaves me frustrated.

Sorry for off-topic replie. I can't send PMs yet.
but how do you get the ym2151 FPGA boards?

GTR3QQ wrote: ↑Wed Apr 05, 2023 10:18 pm Sorry for off-topic replie. I can't send PMs yet.
but how do you get the ym2151 FPGA boards?

I designed them. I ordered them from a manufacturer in China. You can see a video of it running here: https://www.youtube.com/watch?v=5i5-B9j3mMU. It's the board on the left side of the larger "Hendrix" board with yellow and orange jumpers connecting its audio out to the rest of the audio mixer. In this case, I'm using an I2S DAC, not a YM3012. I'm still debugging a few issues when running at speed, but I'm close.