Back [03:58:18]<MarkX> any links or tips for creating a PID loop control without floating point values?
[04:00:41]<z64555> research fixed point computing
[04:00:52]<MarkX> z64555, kk
[04:01:27]<z64555> What's your CPU's word length? please tell me it's higher than 8 bits...
[04:03:18]<MarkX> aren't words always 16bits :S
[04:04:52]<MarkX> my int is 2 bytes
[04:05:09]<MarkX> long/float/double are 4 bytes
[04:06:25]<z64555> ok so your word length is most likely 16 bits, and your double is a lie
[04:06:58]<z64555> doubles are always float64's
[04:07:02]<MarkX> yes it is
[04:07:07]<MarkX> hahaha
[04:07:12]<MarkX> 8bit avr
[04:07:17]<MarkX> atmega32u4
[04:13:24]<z64555> hm. wonder why they call it an 8-bit. perhaps because their GP registers are all 8-bit?
[04:13:37]<z64555> but then again it's a RISC
[04:14:55]<z64555> Anyway, Look into fixed point data structures, you'll likely have to make one yourself, although if you're lucky you might find a generalized library that allows you to set the whole and fractional sizes
[04:16:19]<z64555> another common method is to store the values as "fractionals" and then divide the value by the fractional ratio to get the whole value
[04:17:09]<z64555> well, no, not fractionals, accumulated type is what I think they call it
[17:01:05]<SpeedEvil>http://imgur.com/gallery/zkP5lQM [17:02:01]<robopal> λολ
[17:02:04]<robopal> lol
[17:02:53]<robopal> you know what is more funny? ...I didn't see the umbrella, I was staring at the robotic arm XD
[19:07:29]<z64555> so regenerative suspension is now a thing
[21:42:36]<rue_house> ok
[21:50:28]<z64555> ok? ok.
[21:50:32]<z64555> everything's ok.
[21:50:33]<z64555> :)
[23:50:12]<rue_house>http://www.thingiverse.com/thing:650446 [23:50:16]<rue_house> hehe
[23:55:17]<z64555> oh look! It's a Schrodinger's box
[23:56:13]<rue_house> :)