#robotics Logs

Jul 16 2018

12:59 AM rue_mohr: mrdata, Si5351A
01:25 AM mrdata: rue_mohr, aha
01:25 AM mrdata: Low output period jitter: < 70 ps p2p, typical
01:25 AM mrdata: that block diagram looks quite familiar, eh
01:57 AM mrdata: the thought occurred to me that the lion's share of the power line noise on your board could be ringing due to the parasitic inductances.
01:57 AM mrdata: i calculated the ripple current on the 1.1 uF as about 12 A
01:59 AM mrdata: because I = q / tau
02:16 AM rue_bed: I cant tell with my scope how it is
02:16 AM rue_bed: I think its ground noise
02:16 AM mrdata: this tells me, replace your red V+ wires with 1 ohm 1 watt Rs to dissipate the ringing
02:17 AM mrdata: and add a cap at the pin headers
02:17 AM rue_bed: I had an oops with it a long time ago, I kinda clipped the ground to 270VDC, ever since that flash, its had a bit of noise on it
02:17 AM rue_bed: I could try that tho
02:17 AM rue_bed: if that noise is real anyhow
02:18 AM rue_bed: the 5351 does fractional divide, with a pll
02:18 AM mrdata: neat
02:18 AM rue_bed: so, it probably has the resolution I'm after
02:18 AM mrdata: what is its frequency resolution
02:18 AM rue_bed: its another fork in the road, I got one, waiting for itnow too
02:19 AM mrdata: they brag about 0 ppm error; i guess any error is the drift from the crystal
02:19 AM rue_bed: its 1/(20bits).(20bits)
02:19 AM rue_bed: AND, if needed I can use two outputs with that mixer
02:20 AM rue_bed: dunno, lots of options so far
02:20 AM mrdata: yup
02:20 AM rue_bed: arg, its a work night its 12:20 and I'm still not asleep
02:20 AM rue_bed: I need to stop browing the web all night
02:21 AM rue_bed: but I'm not tired, so, whatever
02:21 AM mrdata: yeah, i have to sleep and then wake up in like 6 hrs
02:21 AM mrdata: so i'm going to call it a night
02:21 AM rue_bed: yes, almost 6 hours exactly now
02:21 AM rue_bed: I'll lay on my side for a bit and not go to sleep that way for a while
02:22 AM rue_bed: that chip, on a board was $6 CAD
02:22 AM rue_bed: with 25.000Mhz crystal
03:12 AM Jak_o_Shadows: satuff takes a minimum of 3 weeks to get here
03:12 AM Jak_o_Shadows: i juST ASSUME A MONTH
07:45 AM Elleo: 16
07:45 AM Elleo: oops
09:14 PM theBear: aarg! i closed my peepers for to rest them a little while sitting in the chair here... now i'm all perma-chilled to my poor forest animal core, and lacking in energy and elbow-lubricants to warm it, 'cos the chilly make the weak will
09:15 PM theBear: speaking of unexpected outbursts... 16 Elleo ?!!?!? my word ! i say ! woops indeed <cheeky grin>
10:09 PM rue_mohr: hmm
10:11 PM Tom_L: not much done today
10:11 PM Tom_L: tweaked the gcode a bit
10:12 PM rue_mohr: mh
10:12 PM rue_mohr: damn hot
10:14 PM Tom_L: i didn't know it got hot up there
10:16 PM Tom_L: that reminds me.. i still gotta program the hex head recess on the bottom side of the part
10:19 PM rue_mohr: mrdata, I was thinking about why the osc woul dbe do unstable
10:20 PM rue_mohr: and I recon its par for the way it works
10:20 PM rue_mohr: the low pass turn the square wave edges into a slope, a steep one
10:20 PM rue_mohr: any slight variation on the trip-points of that edge make a huge timing difference
10:20 PM rue_mohr: so, this all makes snse