[HN Gopher] No Time Like MEMS Time
       ___________________________________________________________________
        
       No Time Like MEMS Time
        
       Author : rbanffy
       Score  : 14 points
       Date   : 2024-10-09 10:57 UTC (1 days ago)
        
 (HTM) web link (morethanmoore.substack.com)
 (TXT) w3m dump (morethanmoore.substack.com)
        
       | theideaofcoffee wrote:
       | MEMS devices absolutely blew my mind the first time I read about
       | them. Thinking: "yeah, this just shouldn't exist". Like, how can
       | you make something with moving parts out of something that
       | normally doesn't move (polycrystalline silicon), and is very
       | fragile at that, all the while doing it at a scale that you
       | require similar processes as 'normal' semiconductors. You can
       | make physical oscillators like the peeps do in this article and
       | have surrounding circuitry to drive and measure them, levers and
       | actuators to mount actual tools onto: probes to manipulate single
       | cells, mirrors to control light (looking at you, DLP) and so many
       | other things. Super interesting!
        
       | nvader wrote:
       | > The beauty of using MEMS over quartz is multi-faceted.
       | 
       | That is crystal-clear to me.
        
       | CamperBob2 wrote:
       | Speaking from (independent) experience, the SiTime parts live up
       | to the hype.
       | 
       | You do _not_ want to be in the quartz crystal business going
       | forward; it 's almost as dead as vacuum tubes, even if the
       | manufacturers don't know it yet. Nothing will be left to fight
       | over but the very cheapest commodity parts.
        
       | amelius wrote:
       | > Ian: There was a story a while ago with Tesla's AI chip - they
       | had to deliver power at a thousand amps per square centimetre.
       | 
       | Where can I read more about that?
        
       | kurthr wrote:
       | One thing to realize about MEMS is that it is mostly used for
       | sensors (accelerometers, gyros, pressure, magnetometers) and
       | actuators. There's a whole journal by that name.
       | 
       | Fundamentally, an oscillator/resonator is the sensing of
       | _nothing_.
       | 
       | You want the frequency of output to be completely independent of
       | acceleration, strain, rotation, pressure, magnetic/electric
       | field, etc). That's really hard to do and involves a combination
       | of building very robust silicon packaging, minimizing (making
       | symmetric) all contacts to the outside to shield it, and
       | compensating for every possible effect you can measure.
        
       ___________________________________________________________________
       (page generated 2024-10-10 23:00 UTC)