#garfield Logs

May 26 2019

12:00 AM rue_mohr: :)
12:04 AM zhanx: thursday I have off so hopefully i can put it all together now
12:04 AM rue_mohr: ?...
12:04 AM rue_mohr: I wonder if I can get the cnc machine to bend wire jumpers for me...
12:04 AM zhanx: i am book solid for time till then, i need a few hours to play with it minus kids
12:04 AM rue_mohr: next month we have a large apartment and a mall to do fire annual on
12:04 AM zhanx: fun
12:04 AM rue_mohr: er, was it two large appartemtns
12:06 AM zhanx: not completely that is not needed i would thing
12:06 AM zhanx: think
12:06 AM zhanx: you can trigger it
12:06 AM zhanx: oh update on the mech?
12:07 AM rue_mohr: I have more grinding on the last ring to do
12:07 AM rue_mohr: then I can weld them togethor
12:07 AM zhanx: weight test it yet?
12:07 AM rue_mohr: heh, the double-rings are solid
12:07 AM rue_mohr: I need to figure out bearings
12:08 AM zhanx: give me the id and od and i can help with that
12:08 AM rue_mohr: the failure with this kinda thing in the past has been centering the rings/bearings so they align properly when spun
12:08 AM zhanx: axial loads of course
12:08 AM rue_mohr: where do I have that....
12:09 AM rue_mohr: the outter two rings are
12:09 AM zhanx: axial load double race with outer lobs
12:09 AM rue_mohr: outtermost "R3": ID 2822mm OD 2938mm
12:09 AM rue_mohr: middle "R2": ID 2554mm OD 2670mm
12:10 AM rue_mohr: I'm pondering bushings
12:10 AM rue_mohr: nylon
12:10 AM rue_mohr: to say there is next to no speed is an understatment
12:11 AM rue_mohr: I was thinking of a nylon bushing about 3" dia
12:11 AM rue_mohr: I think aligning them side-to-side has been lots of the challange
12:11 AM zhanx: you are in pinion bear sizes
12:12 AM rue_mohr: and they have to be hollow
12:12 AM rue_mohr: I need to run an optic data link down the middle
12:13 AM rue_mohr: bushings are less peices tho, I'v generally decided when soemthing needs to be mean, you make it out of the largest peices you can
12:13 AM zhanx: remember dental floss and a small hole with a vacuum cleaner to start the wire pull :P
12:13 AM rue_mohr: :)
12:14 AM rue_mohr: if its bad I usually start with fishing line and the corner of a sandwich bag made into a self-expanding mouse
12:14 AM zhanx: tell rue_mohr a trick is like teaching a jester to juggle
12:16 AM rue_mohr: I'm also wondering if I should always use the /1 counter clock, if I count rollovers, its only 244Hz interrupt overhead
12:16 AM rue_mohr: at 1Hz, thats not even 8 bits extra
12:16 AM zhanx: rue_mohr, what is helpinh my business, i get odd calls with people not know what exactly is wrong and i can still fix it
12:16 AM rue_mohr: :)
12:21 AM rue_mohr: "the nack"
12:21 AM zhanx: due the /1 first and if that is not enough change it up
12:21 AM rue_mohr: if I do the /256 first, I can know, by how many counts exactly what rate to change to for best resolution
12:21 AM rue_mohr: OR I just count overflows
12:21 AM zhanx: i also stupidly took a bank contract
12:21 AM rue_mohr: BUT
12:21 AM zhanx: I have to clean pc's of malware etc
12:21 AM rue_mohr: if its over about 3kHz, your better off counting cycles for 250ms
12:21 AM rue_mohr: and doing *4
12:21 AM zhanx: see now you are losing me
12:21 AM rue_mohr: well
12:21 AM rue_mohr: instead of measuring the time of a full cycle
12:21 AM rue_mohr: you count cycles over 250ms, and then *4 to get Hz
12:21 AM zhanx: ok i understand that
12:21 AM zhanx: but the reason
12:21 AM rue_mohr: my precision goal is 0.1%, I'm at about 0.6%
12:21 AM rue_mohr: the range is 1Hz to 40kHz
12:21 AM zhanx: i know so much but so little at the same time
12:21 AM rue_mohr: to ~0.1% freq
12:22 AM rue_mohr: no green grass over here tho.
12:23 AM rue_mohr: hmm, if I'm cycle counting, how the hell do I measure phase angle...
12:24 AM rue_mohr: I think the resolution is gonna plummet to crap.
12:24 AM rue_mohr: think on this...
12:24 AM rue_mohr: 40Khz, is a cycle time of
12:24 AM rue_mohr: 25000 ns
12:25 AM rue_mohr: 16Mhz is 62.5 ns
12:25 AM rue_mohr: that only gives me 400 counts
12:25 AM rue_mohr: BUT, thats still better than 1 degree
12:26 AM rue_mohr: so I'm ok with that
12:28 AM rue_mohr: it would be easier if /1 was the only method I was using
12:28 AM rue_mohr: or that and cycle counting
12:28 AM rue_mohr: ok
12:28 AM rue_mohr: a) stop comparator from ringing
12:28 AM rue_mohr: b) rewrite all the code
12:29 AM rue_mohr: a.m) make and eat supper
01:54 AM rue_mohr: so it takes about 50ms for my profiler circuit to settle
06:32 PM Tom_L: kindof a sad day. boxing up the sherline and taking the enclosure apart
07:47 PM rue_mohr: hmm
07:48 PM rue_mohr: is it just missing the motor?
07:48 PM rue_mohr: er, motor + spindle?
08:07 PM Tom_L: yes
08:08 PM Tom_L: and steppers now
08:09 PM Tom_L: boxed em up and labelled em for later
08:14 PM rue_mohr: was feed rate or chatter the major issue
09:19 PM Tom_L: on the sherline?
09:19 PM Tom_L: chatter
09:19 PM Tom_L: which limits feedrate
09:24 PM rue_mohr: mmm
09:25 PM rue_mohr: wonder how much buss fair is..
11:07 PM Tom_L: so how long does a package take to skip across canada?
11:10 PM Tom_L: if it went straight up from Chicago, it still has a fair ways to travel
11:30 PM rue_mohr: packages move across canada about 1week/km