3 ms·

I saw this video yesterday -- cool to see it pop up here today. I also watched this Defcon breakdown on hardware hacking where the presenter goes over the spec

I saw this video yesterday -- cool to see it pop up here today. I also watched this Defcon breakdown on hardware hacking where the presenter goes over the specialized tools to both mount the chips as well as the software to pull firmware at rest to reverse engineer it. Video: https://www.youtube.com/watch?v=Kxvpbu9STU4 https://www.youtube.com/watch?v=Kxvpbu9STU4 YouTube channels like My Mate Vince, Northridge Fix, Necroware, and others show folks without any electronics degrees troubleshooting and fixing hardware and I just can't help but want to watch them and think about how fun it would be to get involved...unfortunately I already am at a limit to my hobbies. The older I get, the more I enjoy thinking about EE topics and wondering where I would be today if I would have found out more about this discipline vs going a more traditional administration/IT route.

jve 4y ago
> The older I get, the more I enjoy thinking about EE topics and wondering where I would be today if I would have found out more about this discipline vs going a more traditional administration/IT route. I wonder if the EE folks wonder other way around?
- poleguy 4y ago
  EE here. Not really about admin/IT. But I do wonder sometimes if I should have done software as a career. Mostly it's a grass is greener feeling when I hear about the high end of the salaries. But I do very well financially, the jobs are stable, and I write cool software to solve my own problems, rather than grinding on someone else's Jira tasks. I do lots of IT and computer admin stuff too, but also on my own terms. Setting up simulation clusters, remote hardware automation, development environments, storage, etc... so much so that our department hired a guy to take on some of that work.
  - Tangurena2 4y ago
    Another EE here. I suspect that there is a lot of "greener grass" going on (oh look! I spy with my little eye, grass beginning with G). I've been doing software dev for most of the past quarter century. My current hobby is Eurorack synthesis and I build/solder about half of the items from DIY kits. The ones that I'm mostly into are built around STM32 microcontrollers (I threw out all my C/C++ books about 15 years ago - oops!).
    - orangepurple 4y ago
      > (I threw out all my C/C++ books about 15 years ago - oops!). The future is here for STM32: https://github.com/stm32-rs/stm32-rs https://github.com/stm32-rs/stm32-rs
- pkaye 4y ago
  I was a firmware engineer (now doing web development) and knew plenty of ASIC designers who wished they did software. They were good at what they did but felt the pay and opportunities were better in software development.
_Microft 4y ago
The projects I mention here are not clones themselves but ever cheaper PCB prototypes and microcontrollers make it a fantastic time to create hardware for old systems (e.g. ISA). Enjoy these examples: there is PicoGUS [0], a Gravis Ultrasound emulation, a Raspi Pico based PCMCIA WLAN card [1], EDO/FPM RAM modules [2], "Snark Barker", a soundblaster 1.0 clone [3]... If you want more electronics stuff, I suggest you follow the creators of these on Twitter. [0] https://github.com/polpo/picogus https://github.com/polpo/picogus [1] https://www.yyzkevin.com/pcmcia-pico-w-card/ https://www.yyzkevin.com/pcmcia-pico-w-card/ [2] https://twitter.com/0xCats/status/1524708654913662977 https://twitter.com/0xCats/status/1524708654913662977 [3] https://github.com/schlae/snark-barker https://github.com/schlae/snark-barker
doublerabbit 4y ago
Same thought I have. 15 years of SysAdmin and wanting to move on to something else. What boggles my mind in this video as example. Do the DOS drivers as an example actually communicate with the chips on the PCB, or are there firmware/data burnt on to the chips? I'm confused in that what it looks like is all you need to do is design a schematic and place chips down on to a PCB and then write software drivers. I thought chips had to be coded too.
- _Microft 4y ago
  The card for the CD drive actually uses only stock logic chips but no microcontrollers at all and therefore need no ROM dump of an existing card to work. You can find the list on the Github page of the project: https://github.com/AkBKukU/CM153-Repro https://github.com/AkBKukU/CM153-Repro
- xani_ 4y ago
  ISA bus is basically just quasi-memory bus with some extras. You get address, you either get data or send data, done. You just set an address and write/read data to it. All you need to do simple input/output on it is to * wait for "right" address on bus and "right" pins to light up * either latch the data, or output the data * bang the right control bit. It is entirely doable using basic logic chips, because back then any kind of programmable logic was super expensive. Compare it to usb-c where you need tens of thousands of lines of code to even negotiate some power stuff (or use specialized chips... that have those tens of thousands of lines of code baked in).
- mmastrac 4y ago
  In this case it is effectively a fancy serial port, which means all the chips do all the work for you, no programming needed.
- tenebrisalietum 4y ago
  The ISA bus is basically connected directly to the CPU. See its pinout - site is in German but there is a nice diagram. https://de-academic.com/dic.nsf/dewiki/643647 https://de-academic.com/dic.nsf/dewiki/643647 The lines marked Addr are address lines, and Data are data lines. Things that appear readable/writeable to the CPU as RAM manipulate the MEMW/MEMR lines, and things that appear readable/writeable to the CPU using its "I/O port" instructions manipulate the IOW/IOR lines. I don't know the low level protocols used there but there is definitely one and why some simple logic chips had to often be added I guess (74SNxxxxx chips very common). RAM basically does the same thing, as well as anything built in like the clock chip, keyboard controller, serial ports, floppy controller, etc. One big disadvantage of ISA is that the bus when it was introduced had to run at the same speed as the CPU. When CPUs started getting faster than 8Mhz something was done to keep the ISA bus as 8Mhz, but not sure what. The other big disadvantage is that this was intimately tied to the 8088 Intel CPU signal lines. PCI (and other things like NuBus and MCA) were CPU independent and required a chipset or other controller to interface, and that's where the complexity started to take off.

jve 4y ago

> The older I get, the more I enjoy thinking about EE topics and wondering where I would be today if I would have found out more about this discipline vs going a more traditional administration/IT route. I wonder if the EE folks wonder other way around?

poleguy 4y ago
EE here. Not really about admin/IT. But I do wonder sometimes if I should have done software as a career. Mostly it's a grass is greener feeling when I hear about the high end of the salaries. But I do very well financially, the jobs are stable, and I write cool software to solve my own problems, rather than grinding on someone else's Jira tasks. I do lots of IT and computer admin stuff too, but also on my own terms. Setting up simulation clusters, remote hardware automation, development environments, storage, etc... so much so that our department hired a guy to take on some of that work.
- Tangurena2 4y ago
  Another EE here. I suspect that there is a lot of "greener grass" going on (oh look! I spy with my little eye, grass beginning with G). I've been doing software dev for most of the past quarter century. My current hobby is Eurorack synthesis and I build/solder about half of the items from DIY kits. The ones that I'm mostly into are built around STM32 microcontrollers (I threw out all my C/C++ books about 15 years ago - oops!).
  - orangepurple 4y ago
    > (I threw out all my C/C++ books about 15 years ago - oops!). The future is here for STM32: https://github.com/stm32-rs/stm32-rs https://github.com/stm32-rs/stm32-rs
pkaye 4y ago
I was a firmware engineer (now doing web development) and knew plenty of ASIC designers who wished they did software. They were good at what they did but felt the pay and opportunities were better in software development.

_Microft 4y ago

The projects I mention here are not clones themselves but ever cheaper PCB prototypes and microcontrollers make it a fantastic time to create hardware for old systems (e.g. ISA). Enjoy these examples: there is PicoGUS [0], a Gravis Ultrasound emulation, a Raspi Pico based PCMCIA WLAN card [1], EDO/FPM RAM modules [2], "Snark Barker", a soundblaster 1.0 clone [3]... If you want more electronics stuff, I suggest you follow the creators of these on Twitter. [0] https://github.com/polpo/picogus https://github.com/polpo/picogus [1] https://www.yyzkevin.com/pcmcia-pico-w-card/ https://www.yyzkevin.com/pcmcia-pico-w-card/ [2] https://twitter.com/0xCats/status/1524708654913662977 https://twitter.com/0xCats/status/1524708654913662977 [3] https://github.com/schlae/snark-barker https://github.com/schlae/snark-barker

doublerabbit 4y ago

Same thought I have. 15 years of SysAdmin and wanting to move on to something else. What boggles my mind in this video as example. Do the DOS drivers as an example actually communicate with the chips on the PCB, or are there firmware/data burnt on to the chips? I'm confused in that what it looks like is all you need to do is design a schematic and place chips down on to a PCB and then write software drivers. I thought chips had to be coded too.

_Microft 4y ago
The card for the CD drive actually uses only stock logic chips but no microcontrollers at all and therefore need no ROM dump of an existing card to work. You can find the list on the Github page of the project: https://github.com/AkBKukU/CM153-Repro https://github.com/AkBKukU/CM153-Repro
xani_ 4y ago
ISA bus is basically just quasi-memory bus with some extras. You get address, you either get data or send data, done. You just set an address and write/read data to it. All you need to do simple input/output on it is to * wait for "right" address on bus and "right" pins to light up * either latch the data, or output the data * bang the right control bit. It is entirely doable using basic logic chips, because back then any kind of programmable logic was super expensive. Compare it to usb-c where you need tens of thousands of lines of code to even negotiate some power stuff (or use specialized chips... that have those tens of thousands of lines of code baked in).
mmastrac 4y ago
In this case it is effectively a fancy serial port, which means all the chips do all the work for you, no programming needed.
tenebrisalietum 4y ago
The ISA bus is basically connected directly to the CPU. See its pinout - site is in German but there is a nice diagram. https://de-academic.com/dic.nsf/dewiki/643647 https://de-academic.com/dic.nsf/dewiki/643647 The lines marked Addr are address lines, and Data are data lines. Things that appear readable/writeable to the CPU as RAM manipulate the MEMW/MEMR lines, and things that appear readable/writeable to the CPU using its "I/O port" instructions manipulate the IOW/IOR lines. I don't know the low level protocols used there but there is definitely one and why some simple logic chips had to often be added I guess (74SNxxxxx chips very common). RAM basically does the same thing, as well as anything built in like the clock chip, keyboard controller, serial ports, floppy controller, etc. One big disadvantage of ISA is that the bus when it was introduced had to run at the same speed as the CPU. When CPUs started getting faster than 8Mhz something was done to keep the ISA bus as 8Mhz, but not sure what. The other big disadvantage is that this was intimately tied to the 8088 Intel CPU signal lines. PCI (and other things like NuBus and MCA) were CPU independent and required a chipset or other controller to interface, and that's where the complexity started to take off.