Hacker News new | past | comments | ask | show | jobs | submit login

Very good point -- you'd have these fast, expensive transistors (however they are packaged or not packaged) cache their data, then transfer it back to a slower, cheaper, main microprocessor, which would be doing the rest of the device's work.

Do you know where these 10 GHz transistors can be found?




Laser rangefinders don't switch the light source at 10GHz. They "chop" the laser on-off at (say) 10kHz. The measure the difference between the reference (onboard) and the return using (typically) a PLL that can measure time differences down to picoseconds (these are usually used in high-frequency RF equipment). Using a microcontroller counter would be too slow.

Cheaper LRF's can use triangulation with a known baseline. That's how Neato robotic managed their $10 spinning version for the vacuum: http://www.hizook.com/blog/2009/12/20/ultra-low-cost-laser-r...

To wit... I like to disassemble laser rangefinders!

http://www.hizook.com/blog/2009/01/04/velodyne-hdl-64e-laser...

http://www.hizook.com/blog/2009/02/14/omron-sti-optoshield-o...

http://www.hizook.com/blog/2008/12/15/sick-laser-rangefinder...


Wow, amazing!




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: