Ralford.net

  [Hardware]

After my lengthy complaint about floppy drives, someone at work brought up a situation where floppy may be a necessity: for the case that bootable removable media is required, but bootable USB is not supported by the OS.

Comments (0)      Post a Comment  

  [Hardware]

Found a website this morning that has an archive of just about every CPU and co processor that ever existed: www.cpu-collection.de. It has pictures as well as blurbs about what makes each chip unique in the history of computers.

Comments (0)      Post a Comment  

  [Hardware]

Here is a great applet showing how enhancement mode N-channel MOSFETs work:
N-channel enhancement MOSFET

Comments (0)      Post a Comment  

  [Hardware]

As we move forward in 21st century, I'm continually baffled that the world is still supporting floppy controllers. I associate floppy disks with old 486 problems - they remind me of resorting to my 5-1/4" boot disk to get me out of weird situations. They also bring back memories of shareware games that came on one or two 3.5" disks, like the old 1st persons Wolfenstein 3D and Doom, and even old Apogee Software games like the side scrolling Duke Nukem. A friend of mine still has the original Commander Keen on a 5-1/4" disk.

Where I work, there was once an engineer that used floppy disks regularly. He was a BIOS developer, and would throw binaries on them and give them to other engineers for BIOS testing. I've occasionally used them for DOS boot disks, but nine times out of ten, they don't have the capacity for the files I need, so I just use a bootable USB key, or a PCMCIA ATA flash card. I've been using the same USB key and ATA flash card for at least three years now, and haven't had them "go bad" on me. But I lost track of the number of floppies I've thrown away. Flash technology can be used on every PC these days. Even my ancient desktop PC that runs at 733MHz has USB ports, and the PCI bus is always expandable. It has a floppy drive, but I just don't trust it, as whenever I've put a floppy in it, it comes out covered in dust.

As we approach 2008, we still don't leave the floppy behind us. In fact, there are designs in progress where I work that have floppy controllers on boards next to USB 2.0 and IDE flash chip sockets. The only explanation I see for this is that it doesn't take much effort to design a floppy controller into hardware. It's on almost every super I/O chip out there, and what could go wrong when connecting a handful of low speed signals to a 0.1" header? Especially when it makes that one customer happy. I haven't used Vista yet, but the save icon in Microsoft Word for XP is still a picture of a floppy disk.

I was digging through a bin of old computer hardware in my closet, and found a pile of old hardware I've never parted with: eight 30-pin SIMM RAM (FPRAM) modules from my old 486, an old modem, a Pentium 75MHz processor with the P54 core, a 5-1/4" floppy drive from my old 486, and about six 3.5" floppy drives. I also found an old add-in board with some funny looking chips. One chip, in a 44-pin PLCC package, was the Western Digital WD37C65C. I looked up this guy and found out it was the first PC/AT floppy controller, designed back in 1987! Digging around some more, I found a scanned copy of the WD37C65C datasheet.

So here we are twenty years later, and the computer hardware world just won't let go of the floppy. We've been through ZIP drives with parallel port interfaces (ooooh, so enhanced!), minidiscs, and even USB floppy drives. ZIP disks were cool for backup for some of us here and there. I got my hands on one back during my freshman year in college. By that time, email was a more convenient way to get my files between my dorm room and lab. Minidisc technology is my favorite - with the shudder door having "don't forget the floppy!" written all over it. The USB floppy controller is an excuse to have another USB product. If the interface is there, use a disk-on-key already!

I don't think I have a good solution for how to move the world away from the floppy. Unless... One day I design a USB key that's shaped like a floppy disk, complete with write protection as it was on 3.5" disks. The shutter door will also be there, and when you open it, you'll see some flash memory.

Comments (0)      Post a Comment  

  [Hardware]

I was working on some EPLD code at work today, and had to hop on the internet to learn how to setup constraints in verilog. I was quite surpised to find the following example on the Xilinx website: Yes, the "//" (double front-slash) in front of the constraint would typically indicate a single line comment in the code. Could this perhaps be a typo?

I rooted around a somewhat out-of-date verilog syntax specification for a hint of proof that comments truely were parsed in the verilog grammar. A few more google searches later and I stumbled into a Xilinx page about Verilog Meta Comments.

Apparently, the Verilog parser actually parses comments in search of these meta comments. The Xilinx page mentioned that Verilog does not have a direct attribute definition while VHDL does. Without doing more research on this, it almost looks like the meta comments were a work-around to allow Verilog users to port their design constraints to VHDL.

The only other program I know that parses comments is Doxygen. Doxygen is a piece of software that parses comments and uses them automatically generate documentation for your code. It parses the code (comments) separately from the compiler, so doesn't really fall into the unusual category of meta comments.

Comments (0)      Post a Comment  

  [Hardware]

My girlfriend recently picked up an iPod mini. One of the requirements on the box was USB 2.0 (or FireWire), so we also picked up a 4-port USB 2.0 card. Up until now, neither of us had any USB 2.0 ports on our desktop machines. In fact, the iPod is the first device I've run into that requires USB 2.0 over USB 1.x.

Most of our newer CPU designs at my workplace support USB. When I first started as a full time employee, I assumed the technology to be USB 2.0. Turns out that they are all USB 1.1. Our product line didn't support USB 2.0 until we came out with a dedicated peripheral card to support it.

So what's the fuss with 2.0? USB 2.0 supports three data rates: Low Speed, Full Speed, and High Speed. Before USB 2.0 came into existence, only Low Speed and Full Speed were supported. The corresponding transfer speeds are:

• Low Speed - 1.5 Mbit/s
• Full Speed - 12 Mbit/s
• High Speed - 480 Mbit/s

Fortunately, USB 2.0 supports USB 1.x devices, but you won't get 480 Mbit/s with an old 1.x USB card. If you get an iPod mini, chances are that it might still work with your old card, but transferring MP3s to it will be 48 times faster with USB 2.0. Better go upgrade.

Comments (0)      Post a Comment  

  [Hardware]

On April 19th of this year, Moore's Law turned 40. A few weeks ago, an interview with Gordon Moore appeared on Techworld.com. In it they asked Moore how long he thinks Moore's Law will hold true - how long the number of ICs on a chip will exponentially grow with respect to time. As the size of transistors we know them are approaching their size limitation in nanometers and even picometers, Moore says eventually we will reach the point where transistors will need to be at the atomic level for the law to hold true.

A copy of the orginal article from the April 1965 issue of Electronics Magazine can be found at Intel's webpage.

Comments (0)      Post a Comment  

  [Hardware]

I was browsing the web for info on open collector outputs and stumbled into a project promoting open hardware. It doesn't seem popular, but the idea behind it consists of three rules. In the words of OpenCollector.org, for open hardware to exist...

1. Information on using the hardware must be available
2. The design of the hardware must be available
3. Design software for the hardware must be availble

Elaborating on these bullets, OpenCollector.org explains:

1. The interface to the hardware must be explicitly made public, so the hardware can be used freely.
2. The design of the hardware must be made public, so that others can implement it and learn from it.
3. The tools used to create the design should be free, so that others can develop and improve the design.

I'm a fan of numerous open source projects. I think a big reason that open source works is because of the nature of the internet. Without the internet, it would be more difficult (maybe impossible?) to develop and distribute an open source software project. Schematics and VHDL code could easily be made available on the internet, but a physical module or system would be difficult to build without a lump sum of money.

If you're interested, another similar (and more recent) source for open hardware can be found here: OpenHardware.de

Comments (0)      Post a Comment  

  [Hardware]

An article over at EETimes.com mentions that Intel is making a push for a common cell phone bus - specifically aimed towards cell phone memory. One complaint from the article mentions that "too many buses, too many pins, too many interfaces" are all "[making] it difficult for designers of memory controllers".

According to the manager of Intel, the multi-media card (MMC) interface is a good bus to put the focus on. Multi-media cards are typically found in various cell phones, digital cameras, and PDAs.

Comments (0)      Post a Comment  

  [Hardware]

VIA is in the news again this week. This time, they've released a new mini-ITX board that has dual EDEN-N CPUs. The board's chipset uses a VIA CN400 north bridge and a VT8237R south bridge.

Besides the dual processors, there are a few other differences between this board and VIA's previous mini-ITX boards. The new board supports up to two 2GB of memory with it's two memory slots. There is one standard PCI slot, but the board has an additional mini PCI slot on the bottom side. There are 3 ethernet connections on the board, one of which is a 10/100/1000 Base-T connection.

For more information on the EPIA DP-310, check out the board's specifications at VIA's website. Or go here to see what slashdotters have to say about the mini-ITX form factor.

Comments (0)      Post a Comment  

<< 1 of 2 >>