#robotics Logs

Jul 21 2018

03:24 AM Jak_o_Shadows: I think I am reasonably happy with this link for the stewart platform
03:27 AM Jak_o_Shadows: Is it too long? Probably. Is the lower link not long enough? Possibly.
04:01 AM rue_mohr: image?
05:50 AM Jak_o_Shadows: when I have it assembled
05:50 AM Jak_o_Shadows: So probably tomorrow (my time)
07:36 AM Jak_o_Shadows: Here it is with both ends fixed
07:36 AM Jak_o_Shadows: One end will have servos w. Lever arms to change the link length
07:36 AM Jak_o_Shadows: http://pasteall.org/pic/show.php?id=e46bf4ace72172bc4531f4675c0b34ba
07:36 AM Jak_o_Shadows: But I was having trouble visualising the angles those servos had to be on. So I joined it all solid
10:19 AM rue_mohr: dont people usually just use rods threaded at each end?
10:19 AM rue_mohr: can it move it is it ridgid? cant mentally parse that for some reason
10:20 AM rue_mohr: oh its cause 6 hours sleep
02:58 PM kave: what is a keyframe?
02:59 PM kave: Im trying to understand the concept
03:01 PM kave: my question is related to Pose Graph Optimization and SLAM
03:55 PM rue_shop4: hmm
03:55 PM rue_shop4: usually a reference
07:18 PM Jak_o_Shadows: You are correct most people use threaded rod a.
07:19 PM Jak_o_Shadows: But my printing it, I have the option of putting force sensors in it later.
07:19 PM Jak_o_Shadows: It is solid because I was having trouble visualising the link end locations
07:20 PM Jak_o_Shadows: Gotta have it not straight up/down for better force handling
07:41 PM Rez is now known as LoRez
08:21 PM rue_shop3: hahahah, its not oscillatirng because its not a transisitpr
08:21 PM rue_shop3: damn AM reciever chip I think
08:31 PM mrdata-: eh?
08:35 PM justan0theruser is now known as justanotheruser
09:03 PM rue_shop4: odd, I got this reciever to oscillate, but its 6.5Mhz
09:04 PM rue_shop4: its a toy car, is there a 6.5Mhz band for toy cars?
09:13 PM rue_shop3: quiz, who knows what a 9N74 is
09:13 PM Tom_L: a piece of sand
09:14 PM Tom_L: part for a VFD
09:15 PM rue_shop3: its a OLD 7474
09:15 PM rue_shop3: 9N was the pre-ttl prefix
09:15 PM Tom_L: i even forgot what was pre ttl or rtl
09:15 PM Tom_L: what came before rtl?
09:15 PM rue_shop3: and the LM9014 is a 9014
09:16 PM rue_shop3: not sure on how that all went
09:16 PM Tom_L: i had some old dividers that had offset legs
09:16 PM rue_shop3: looks like there was 9Nxx 54--xx and then later 74--xx
09:16 PM Tom_L: 4 & 6 leg iirc
09:16 PM rue_shop3: then s then ls
09:17 PM Tom_L: http://www.armory.com/~rstevew/Public/SoundSynth/TopOctave/topdividers.html#Fig2
09:17 PM Tom_L: figure 9 & 10
09:19 PM Tom_L: ever seen those?
09:21 PM rue_shop3: no
09:21 PM rue_shop3: looks pretty specialized
09:21 PM rue_shop3: tho M2556 rings a bell
09:23 PM rue_shop3: wow, what a mess to try to make replacements for
09:24 PM rue_shop3: surprised they dont just have whole replacement boards for the keybaords
09:26 PM Tom_L: just old
09:26 PM Tom_L: they were single sided phenolic boards back then
09:27 PM Tom_L: traces were as wide as your finger :)
09:27 PM Tom_L: all tin coated, no mask as i remember
09:28 PM rue_shop3: hmm, ok so, most of what I'm gonna snag up on are the rails
09:29 PM Tom_L: rails for what?
09:29 PM rue_shop4: new cnc
09:29 PM Tom_L: oh, you're back working on that now?
09:30 PM rue_shop4: do you ever recall an RC toy being marked 6.5Mhz?
09:30 PM Tom_L: no
09:30 PM rue_shop4: there are about 4 threads going
09:30 PM Tom_L: i got little to nothing done today
09:31 PM rue_shop4: the gyro cage rings, this stupid signal generator, new cnc...
09:31 PM Tom_L: was gonna write some more cad cam
09:31 PM Tom_L: need to make some spacers next too
09:34 PM rue_shop4: oh and the oven
09:35 PM rue_shop4: where is the todo list for this vaccation
09:35 PM Tom_L: you got another vacation now?
09:35 PM Tom_L: i've a mind to make a taller Z
09:36 PM Tom_L: but i probably won't
09:41 PM rue_shop4: yea took them a minth apart
09:42 PM Tom_L: gonna use the C channel on this one?
09:43 PM Tom_L: i think i started a cad model using C channel, dunno if i kept it
09:43 PM rue_shop4: yea, base will be /----\
09:44 PM rue_shop4: I have an I beam I think is for an upright, but they dont perform well with twist
09:44 PM Tom_L: find some square tube
09:44 PM rue_shop4: so maybe squarestock
09:44 PM rue_shop4: THAT SAID
09:44 PM Tom_L: this 4" worked out great
09:44 PM rue_shop4: I'm not loading it
09:44 PM rue_shop4: so, in reality it dosn't matter
09:45 PM Tom_L: but if you could you might
09:46 PM Tom_L: i'll show you one i started
09:48 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Channel_design.jpg
09:51 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Bad_idea_design.jpg
09:51 PM Tom_L: and another one
10:03 PM rue_shop4: oh channel upright
10:03 PM rue_shop4: I forgot about that
10:03 PM Tom_L: i was just playing with ideas at that point
10:04 PM Tom_L: the upright channel would need some sort of brace behind it
10:13 PM rue_shop4: mrdata, I'm also having problems with an amp that can deal with the 10:1 range
10:15 PM rue_mohr: oh right, my scope is only 25Mhz
10:16 PM rue_mohr: on top of everything else, I'm gonna be hitting my head on that
10:16 PM rue_mohr: hmmmm
10:16 PM rue_mohr: time to make a mixer
10:19 PM rue_mohr: I found an interseting trick in an app note
10:20 PM rue_mohr: it was for a pll, but the phase detector couldn't take the main freq, so they put a mixer in to bring it down
10:28 PM Tom_L: what's all this frequency divider stuff for?
10:28 PM Tom_L: maybe i asked that already...
11:12 PM mrdata: a 25 MHz scope will be rough at the top end, if you need to see 10 MHz
11:12 PM mrdata: i simulated a ring modulator
11:13 PM mrdata: in ltspice; it was interesting
11:13 PM mrdata: so let's say you were to mix a 1 MHz signal with a 900 kHz one, to try to get 100 kHz
11:14 PM mrdata: well, you get the 100 kHz but you also get 200, 300, 400, 500, 600, 700, 800, 1100, 1200, 1300, 1400, ...
11:14 PM mrdata: and the original frequencies
11:20 PM rue_shop4: oof, logic chips do NOT like breadboarding 26Mhz
11:20 PM rue_shop4: oof, mess of noise then eh?
11:22 PM rue_shop4: mrdata, that kinda makes sense tho, you want the two freqs to be quite a ways off from your result so you can filter
11:23 PM rue_shop4: I dont have a 74xx74 that can accept this signal
11:36 PM rue_shop4: Tom_L, how important do you think table stability needs to be for pulling up on the bed?
11:36 PM rue_shop4: this is something I have argued with myself a number of times
11:37 PM rue_shop4: I suppose helix cutters can pull up on a job, a bit
11:42 PM rue_shop4: odd
11:43 PM rue_shop4: a 4069 ring osc is slower than a 4011 with 1 input of each gate tied to vcc
11:49 PM rue_shop4: arg, so many bias voltages!
11:49 PM mrdata: breadboards have huge parasitics
11:50 PM mrdata: at 26 MHz, prototype with manhattan construction
11:50 PM mrdata: follow rules of radio
11:54 PM rue_shop4: ugh, 74hc14 output is just horrid ringing
11:54 PM rue_shop4: the inter-row of a breadboard is 2pf
11:54 PM rue_shop4: EEVblog
11:54 PM rue_shop4: whats the impedence of 2pf at 26Mhz?
11:55 PM rue_shop4: bet its about 1k
11:55 PM rue_shop4: 3k
11:55 PM rue_shop4: so, yea, that would make a mess
11:56 PM rue_shop4: esp since this 26Mhz is square wave }:)
11:57 PM rue_shop4: I cant understand why a 4011 is faster than a 4069