He didn't, really. He made an inexpensive clone of the military radio.
Where I work, we make complex industrial robots that incorporate multiple SBCs. The SBCs we use cost a bit more than $2K each, although you can buy functionally equivalent ones off of Amazon for about $200.
The reason that we go with the expensive ones over the cheap ones is that the expensive ones are hardened for use in very rough environments (electrically noisy, dirty, high temperature, high-vibration, subject to frequent physical shocks, etc.). The cost isn't for the computational abilities, it's for the ruggedness required to operate constantly in adverse environments for years on end. My guess is that the same thing is true for these military radios.
keepamovin•2mo ago
What types of things differentiate a normal component from a 'ruggedized' component?
JohnFen•2mo ago
I'm not entirely certain at the component level (I'm a software dev, not a hardware guy), aside from the components having a much higher heat tolerance than consumer components. Most of the difference is in construction of the board to make it tolerant of vibration and shock.
freedomben•2mo ago
I'm also a software guy but started in military grade hardware.
You're correct. Stress testing alone on components and integrated modules will typically wreck consumer grade devices in a hurry. Engineering hardware to operate in extreme conditions takes significant effort and iteration. Even little things like manufacturing solder joints with no weak spots is a big task. Plus there's a lot of cost for certification, which you need at all levels.
The actual circuit design is usually the easy part
JohnFen•2mo ago
Where I work, we make complex industrial robots that incorporate multiple SBCs. The SBCs we use cost a bit more than $2K each, although you can buy functionally equivalent ones off of Amazon for about $200.
The reason that we go with the expensive ones over the cheap ones is that the expensive ones are hardened for use in very rough environments (electrically noisy, dirty, high temperature, high-vibration, subject to frequent physical shocks, etc.). The cost isn't for the computational abilities, it's for the ruggedness required to operate constantly in adverse environments for years on end. My guess is that the same thing is true for these military radios.
keepamovin•2mo ago
JohnFen•2mo ago
freedomben•2mo ago
You're correct. Stress testing alone on components and integrated modules will typically wreck consumer grade devices in a hurry. Engineering hardware to operate in extreme conditions takes significant effort and iteration. Even little things like manufacturing solder joints with no weak spots is a big task. Plus there's a lot of cost for certification, which you need at all levels.
The actual circuit design is usually the easy part