#garfield Logs
Nov 10 2020
#garfield Calendar
08:57 AM polprog: nice, the parts came in
09:07 AM polprog: 4/6 on the second part of the ee exam
09:08 AM polprog: i screwed up a superposition rule (-1p) and did not write kirchoffs rules for the circuit
09:08 AM polprog: it was said in the problem to do so but i didnt need them
09:08 AM polprog: to solve the problem
09:08 AM polprog: screw them
09:08 AM polprog: i passed that half anyway
12:34 PM rue_mohr: in this morning lecture we learned the term ADC!
12:35 PM rue_mohr: good show polprog
12:35 PM polprog: :)
12:35 PM polprog: now gotta solve a weird mechanics problem
12:35 PM polprog: well,, its a cool problem but unusual
12:37 PM polprog: knoblaucheks' lynx would switch to https and your https site is borked
12:37 PM polprog: not only it 404s, mozilla shows a warning about untrustworty cert provider
12:45 PM rue_mohr: it has no https
12:45 PM rue_mohr: hmmm
12:47 PM Tom_L: oh noes
01:04 PM rue_mohr: because if we want to download a pic of a cat doing something silly it needs to be secure?
01:05 PM rue_mohr: because the service provider of the person who is looking at the page can insert ads?
01:09 PM rue_mohr: hahha, these vga people got the front and back porches mixed up
01:10 PM rue_mohr: I'm sure of it
01:10 PM rue_mohr: and everyone must be copying everyones work
01:12 PM rue_mohr: huh or I'm mixed up, why do I know this to be the other way around
01:42 PM polprog: example?
01:42 PM polprog: i know front proch is on the end of the line i think
01:43 PM rue_mohr: I was pretty sure that front porcha nd back porch were relative to the video signal
01:43 PM polprog: its weird
01:43 PM rue_mohr: sync front porch video back porch
01:44 PM rue_mohr: but apparently its front_porch sync back_porch video
01:44 PM polprog: yeah
01:44 PM polprog: https://i.stack.imgur.com/Cyrkp.gif
01:44 PM polprog: (this is cvbs, disregard that colorburst)
01:44 PM rue_mohr: sure, but, why did I specifically have it locked in my head the wrong way around?
01:44 PM polprog: idk
01:45 PM polprog: because its the other way around than id expect it :D
01:45 PM rue_mohr: yea
01:45 PM polprog: for me the front of the signal is on the left...
01:45 PM rue_mohr: should I buy a 4M sram chip and make a vga thing?
01:45 PM polprog: do you want to
01:45 PM rue_mohr: if I'm gonna it doesn't matter :)
01:46 PM polprog: :P
01:46 PM rue_mohr: I did a quick test for a 22V10D
01:46 PM polprog: show them how it's done, bitbang some HDMI from an eprom
01:46 PM rue_mohr: it compiled ok, and flashed ok, I have to test the chip
01:46 PM rue_mohr: hah
01:46 PM rue_mohr: the bitrate on that might be a tad high
01:47 PM polprog: i wonder what's the *lowest* scan rate and resolution that can be carried over HDMI
01:47 PM rue_mohr: class in 15 mins
01:47 PM rue_mohr: ooo
01:47 PM polprog: i gotta do the programming assignment for a moment now
01:47 PM polprog: i mean, i gotta do it on friday but i have to take a break from mech
01:47 PM rue_mohr: send yuor answer in FSM format
01:47 PM polprog: lol
01:48 PM polprog: they proffesor has a collection of doom screenshots on his homepage, as well as some APL code
01:48 PM rue_mohr: State = FSM[State];
01:48 PM rue_mohr: now just write the FSM table
01:48 PM polprog: he probably would have cracked any FSM thing i threw at him
01:48 PM polprog: mail him some eproms with the answers programmed in
01:49 PM rue_mohr: :)
01:50 PM rue_mohr: oh I was going to write a program to test the fsm for that SAR
01:50 PM rue_mohr: it might be handy for fsm tests in general
01:51 PM rue_mohr: 10 mins till class
01:51 PM rue_mohr: I'm quite sure the 22V10 I got from china were pulled
01:51 PM rue_mohr: they have an ID on them
01:52 PM rue_mohr: and the pins look pulled (not splayed as per new chip, actually slightly leaned as if the chip was stressed sideways in the socket)
01:52 PM rue_mohr: I'm not worried about it
01:53 PM rue_mohr: it looks to me like the number of terms you can put on one of these is kinda limited
01:54 PM rue_mohr: a 16V8 couldn't be used for a full 8x8 memory it looks like
01:54 PM polprog: yeah, most stuff from china is pulled
01:54 PM rue_mohr: and a 22V10 couldn't be used for a 14x8 memory
01:54 PM polprog: my 6502s were probably pulled too
01:55 PM polprog: the 27c512 were pulled for sure, there was code on them
01:55 PM rue_mohr: :)
01:55 PM rue_mohr: it'd be polite for them to atleast erase them for ya
01:55 PM polprog: haha
01:59 PM aandrew: whelp, that took a lot more concrete than I expected
01:59 PM aandrew: ~100kg
01:59 PM aandrew: and of course I started to get a feel for how much water to add on the last bag or so
02:00 PM rue_mohr: heh
02:00 PM rue_mohr: were you able to keep it wet between mixes?
02:00 PM rue_mohr: bang it up good to shake it all in?
02:01 PM aandrew: and my hands are sore now from a combination of the rough work and the concrete giving off calcium ...carbonate?
02:02 PM aandrew: kind of -- the two bags were no issue (and buying the quick set fence post one was a mistake I think) but I had to run out to Home Depot to grab two more bags and what I'd poured was more or less set already by then. I tried ramming a rebar rod down but noticed that was going to be bad so stopped
02:04 PM aandrew: hindsight being 20/20, I would make sure I have more than enough on hand, make sure it's "normal" concrete mix, not "sets in 20m", get some rebar down there before pouring, and overall I would have made the thickness of my walls min 1"
02:04 PM aandrew: although I learned a trick to kind of bow out the plywood forms, find some rocks that are about the right width and jam them down between the two plywood boards :-)
02:04 PM aandrew: it means my concrete will be () shaped a bit but that's fine for my very first attempt
02:05 PM aandrew: and I'm still 99% sure I will not be able to get the outer plywood form removed because there's no fucking way I'm gonna dig any more to get enough room to get a drill down there
02:38 PM polprog: weird
02:38 PM polprog: got two cpp classes that are identical
02:38 PM polprog: they just have a different number of fields
02:38 PM polprog: one doesnt want to take a specific istream overload
02:38 PM polprog: grr
02:38 PM polprog: ill tackle that later
03:26 PM polprog: ok, done
04:51 PM polprog: ok, ive managed to stuff the laser somewhere safe
04:52 PM polprog: midnight in 10 mins
04:52 PM polprog: hmm, should i try to make the triangle gen
04:52 PM polprog: i am pretty sure i dont need a 555 for it
04:52 PM polprog: i can use a....
04:52 PM polprog: ....HCT14
04:52 PM polprog: ;)
05:13 PM rue_mohr: aandrew, yea, that fencepost stuff can start setting up really fast, not much time to work anything
05:13 PM rue_mohr: also, they say the normal concrete sets harder
05:14 PM rue_mohr: polprog, the margins might be a bit tight for a 7414, go for it tho!
05:15 PM polprog: i need to make it 50% duty fist though
05:15 PM polprog: hmm i didnt order flip flops damnit
05:15 PM polprog: ahh no there they are
05:15 PM polprog: 4013
05:17 PM rue_mohr: did you see my 74hc14 flipflop?
05:17 PM polprog: i think i did
05:23 PM rue_mohr: I can post later if you want
05:23 PM Tom_L: any more on your schematic book?
05:23 PM polprog: hmm this is a d flip flop, how can i make it an sr
05:23 PM polprog: hmmmm
05:23 PM polprog: i know
05:23 PM rue_mohr: use the set and reset inputs
05:23 PM polprog: wait ni
05:23 PM polprog: no
05:23 PM polprog: i want it to alternate every clk
05:24 PM polprog: yes
05:24 PM rue_mohr: ok
05:24 PM rue_mohr: its old, whatever
05:26 PM polprog: ok, downloaded it and put on the workshop machine
05:26 PM polprog: so i can search it easier
05:26 PM polprog: well
05:26 PM polprog: browse
05:26 PM rue_mohr: sure,
05:26 PM rue_mohr: I'm gonna try to do a new scan of the whole thing
05:26 PM polprog: how come i didnt come up with the idea of a local webserver earlier
05:26 PM polprog: well i did but not on this machine
05:27 PM polprog: the older box runs it for floppy access
05:27 PM polprog: ultramafic
05:27 PM rue_mohr: like 1am there now?
05:28 PM polprog: 0027
05:28 PM polprog: not late by my standards, i have an hour left or so
05:28 PM polprog: ohh
05:28 PM polprog: right
05:28 PM polprog: long exp photo
05:28 PM rue_mohr: heh
05:28 PM polprog: hmm, can i do it
05:28 PM polprog: sure
05:29 PM polprog: ok, now need to find a way to make a list of thumbnails with lighthttpd
05:29 PM rue_mohr: bah, I clicked on of the images and my browser is locked up
05:29 PM polprog: but thats later
05:29 PM polprog: rip
05:29 PM polprog: heh, random thing opened
05:29 PM polprog: 4069 amplifier
05:29 PM polprog: i did that with hc04, now that i have 4069 too i should compare them
05:29 PM rue_mohr: :)
05:30 PM rue_mohr: the 22V10 test passed
05:46 PM polprog: hmm, ill have to rethink it or get a tripod
05:46 PM polprog: back to the tri gen
05:54 PM polprog: hmm
05:55 PM polprog: it.. sorta works but it doesnt make a nice wfm
05:56 PM polprog: bloody pnps
05:58 PM polprog: ok, this one makes a cv charge circuit
05:59 PM polprog: this looks better, bigger base Rs
06:01 PM polprog: hmm, this is good
06:03 PM polprog: this breadboad is so effing noisy
06:10 PM Tom_L: input P
06:10 PM Tom_L: 0.181+(0.0482*p)+(0.00323*p*p)+(0.0000629*p*p*p)
06:10 PM Tom_L: pwm % to led pwm %
06:10 PM Tom_L: rpi
06:16 PM polprog: going to bed
06:17 PM polprog: ill figure your that 4 diode current source later
06:17 PM polprog: for now i have a simple current source/sink thingy
06:17 PM polprog: nn
06:30 PM rue_mohr: back
06:30 PM rue_mohr: I have the... "cement"
06:30 PM rue_mohr: apparently, 1093c
06:30 PM Tom_L: gonna cook some metal?
06:31 PM rue_mohr: polprog, oh, the 74hc14 would pull too much current and bend it!
06:31 PM polprog: no, im driving it thru the 100k Rs
06:31 PM polprog: sec
06:32 PM rue_mohr: the diodes just allow one transistor to do the cap while the other is shorted out by the flipflop
06:32 PM polprog: hmm
06:32 PM polprog: https://i.ibb.co/vHjZSDq/IMG-20201111-013107.jpg
06:32 PM polprog: ive got this
06:32 PM polprog: not the most perfect
06:33 PM polprog: ill have to play with the diode approach
06:33 PM rue_mohr: hu?
06:33 PM rue_mohr: no no thats still RC
06:33 PM polprog: ideally cap voltage would go into the 14 and the 14 output would drive the current source
06:33 PM polprog: hmm
06:33 PM rue_mohr: try this,
06:33 PM polprog: right
06:33 PM rue_mohr: remove the 22k
06:33 PM rue_mohr: and bump the base resistors up to about 500k
06:34 PM rue_mohr: or higher
06:34 PM rue_mohr: 470k to 1M
06:34 PM polprog: i tried without the 22k, it was squareish
06:34 PM polprog: actually
06:34 PM polprog: that makes sense, ill have to test with a bigger cap
06:34 PM polprog: but ill check the diode thing you made
06:34 PM polprog: hmm
06:35 PM polprog: tomorrow
06:35 PM rue_mohr: isit tommorow,
06:35 PM rue_mohr: hurry up
06:35 PM rue_mohr: it is tommorow
06:36 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1170986.jpg
06:36 PM rue_mohr: ^^ tom
06:36 PM rue_mohr: should be good for 660c then
06:36 PM rue_mohr: ?
06:36 PM polprog: its 1:34, so yeah next evening
06:37 PM polprog: gotta eat something and go to sleep
06:37 PM polprog: nighters
06:37 PM rue_mohr: while its cooking re-arrange those resisors
06:42 PM polprog: ive put it in the drawer already :(
06:53 PM rue_mohr: {muffled} beep beep.... change my resistors... please beep beep....
07:11 PM TermMoon: !
07:12 PM rue_mohr: !
07:12 PM TermMoon: ;)
07:13 PM rue_mohr: yea I'v got soemthing
07:13 PM rue_mohr: and the casts are apparently ill
07:13 PM rue_mohr: cast
07:13 PM rue_mohr: cats
07:13 PM TermMoon: HA
07:14 PM TermMoon: i mean :(
07:14 PM rue_mohr: well, they puke when their ill
07:14 PM rue_mohr: so whats up?
07:14 PM TermMoon: not much
07:15 PM rue_mohr: so, FSM, tell me what you know
07:15 PM rue_mohr: got what it stands for?
07:15 PM TermMoon: still getting used to my keyboard
07:15 PM rue_mohr: yea, those things are crazy!
07:16 PM TermMoon: b;lamk keys
07:16 PM rue_mohr: sure
07:16 PM rue_mohr: FSM...
07:16 PM TermMoon: ok
07:16 PM TermMoon: 1 sec
07:16 PM TermMoon: fsm -> fuck sakes man!
07:17 PM rue_mohr: nope I'm reffering to finite state machines
07:17 PM TermMoon: oh! :P
07:17 PM rue_mohr: ok, how about LUT
07:17 PM TermMoon: ladys under thumb?
07:18 PM rue_mohr: no, its not even a high class slut, its a Look Up Table
07:18 PM TermMoon: JHAHA
07:18 PM rue_mohr: ok, finite state machines, what do you know
07:18 PM TermMoon: ok sure
07:18 PM TermMoon: ok
07:19 PM rue_mohr: tell me something that a finite state machines has a finite number of
07:19 PM TermMoon: address based, weiged againsed input
07:19 PM rue_mohr: sure
07:19 PM TermMoon: states
07:19 PM rue_mohr: yes
07:19 PM rue_mohr: so, is a 4 bit counter a FSM?
07:20 PM TermMoon: ies
07:20 PM rue_mohr: what if its an up/down counter?
07:20 PM TermMoon: debatable
07:20 PM rue_mohr: expalin the debate
07:21 PM TermMoon: well i would say:
07:22 PM rue_mohr: so, how many states does a 4 bit counter have then?
07:22 PM TermMoon: any machine based in world of physical constraints, like havug an address limit and a fixed unstructiom set,
07:22 PM rue_mohr: approximitly
07:22 PM TermMoon: hiold om
07:23 PM rue_mohr: I can read keyboard mash, stop correcting and just type
07:23 PM TermMoon: --technicjly has a HYGE but fiimnite amoint oif stsdrs
07:23 PM rue_mohr: a 4 bit counter?
07:23 PM TermMoon: anuthing in rhwe workd
07:23 PM rue_mohr: dont confuse states with transitions
07:24 PM TermMoon: no states
07:24 PM TermMoon: !
07:24 PM TermMoon: in the world
07:24 PM rue_mohr: so, the correct answer is 16
07:24 PM TermMoon: !@SMDQOME!!!
07:24 PM TermMoon: :|
07:24 PM rue_mohr: can you identify 3 of the states a 4 bit counter could be in
07:25 PM TermMoon: 1 10 11 100 101 110 111
07:25 PM rue_mohr: well, thats more than 3, but yes
07:25 PM rue_mohr: now, lets say we have a 3 bit up counter with no reset, how many transitions are there?
07:26 PM TermMoon: 7
07:26 PM TermMoon: 8
07:26 PM TermMoon: firgot 0
07:26 PM rue_mohr: there ya go
07:27 PM rue_mohr: ok, do the number of transitions change if the 3 bit counter has a reset line?
07:28 PM TermMoon: thats a fuvjimg dood quesrtiuon
07:28 PM TermMoon: i,m,a rhiink 1 sec
07:29 PM TermMoon: yes to squared?
07:29 PM rue_mohr: not quite,
07:30 PM TermMoon: :|
07:30 PM rue_mohr: you can go from any of the 8 states, back to zero
07:30 PM TermMoon: doiub;e??
07:30 PM rue_mohr: so we have the orig 8, plus 7 more (one of them already went back to zero)
07:30 PM TermMoon: oj
07:30 PM TermMoon: olk
07:31 PM rue_mohr: well, you might say there is one from 0 back to 0, so, +-
07:31 PM TermMoon: :}
07:31 PM rue_mohr: ok, so, lets see
07:32 PM rue_mohr: yea, its hard to grasp everything you can do without a target app
07:33 PM TermMoon: ind33d
07:33 PM rue_mohr: so, state machines are event based
07:33 PM rue_mohr: it does not mean that each event has to cause a change to a new state tho
07:33 PM TermMoon: isnt everything
07:33 PM TermMoon: ?
07:34 PM rue_mohr: no
07:34 PM rue_mohr: a logic gate isn't changing because of an event
07:34 PM TermMoon: on gard
07:34 PM rue_mohr: its sort ajust holding an equilibriam
07:35 PM TermMoon: oh sure so when an electron comes knocking on youir dppr its not an evernt??
07:36 PM TermMoon: teasing
07:36 PM rue_mohr: lots of state mcahines are clocked
07:36 PM TermMoon: questiom!
07:37 PM rue_mohr: a clock edge makes takes the input status and flips it over to the next state
07:37 PM TermMoon: question.
07:37 PM rue_mohr: uh hu
07:37 PM TermMoon: ok:
07:38 PM TermMoon: if you pipe ins into outs of an FSM you get normal computer jazz !FSM >>????
07:39 PM rue_mohr: you can
07:39 PM TermMoon: :)_
07:39 PM rue_mohr: there is really advanced stuff that I'v forgotten
07:39 PM rue_mohr: I left me a few notes
07:40 PM rue_mohr: ok, lets say you have a task
07:41 PM TermMoon: sweepimg trhe gflopor
07:41 PM rue_mohr: to make an FSM that outputs the sequence 00 01 11 10
07:41 PM rue_mohr: and then repeats
07:41 PM TermMoon: kim
07:41 PM rue_mohr: your to make it with a LUT with feedback
07:42 PM TermMoon: address 10 dricrrvrs the reset?
07:42 PM TermMoon: data
07:42 PM rue_mohr: the goal is to use memory and a feedback latch
07:42 PM rue_mohr: D0 -> latch ->A0
07:42 PM rue_mohr: D1 -> latch ->A1
07:43 PM rue_mohr: each clock pulse, the data at the memory locations is sent to the address lines
07:44 PM rue_mohr: its edge triggered, so only 1 lookup happens at a time
07:44 PM Tom_L: https://www.youtube.com/watch?v=ypBunFtrOrE&feature=youtu.be
07:45 PM Tom_L: you got one of those in your tool belt?
07:45 PM rue_mohr: start from address 00
07:46 PM rue_mohr: if the state is the address
07:46 PM rue_mohr: at address 00 what should you data be
07:46 PM rue_mohr: address data
07:46 PM rue_mohr: 00 01
07:47 PM rue_mohr: 01
07:47 PM rue_mohr: 01 11
07:47 PM rue_mohr: 10
07:49 PM rue_mohr: 11 10
07:49 PM rue_mohr: 10 ?
07:49 PM rue_mohr: 10 00
07:49 PM rue_mohr: summary
07:49 PM rue_mohr: address data
07:49 PM rue_mohr: --------------
07:49 PM rue_mohr: 00 01
07:49 PM rue_mohr: 01 11
07:49 PM rue_mohr: 10 00
07:49 PM rue_mohr: 11 10
07:51 PM rue_mohr: and it'll loop
07:51 PM rue_mohr: Tom_L, :/ no
08:05 PM Tom_L: https://i.imgur.com/pta5Vuw.jpg
08:06 PM Tom_L: you'd have fun with that fiber fuser!
08:07 PM rue_mohr: yea looking for one at work
08:07 PM Tom_L: that guy just bought that one to do some for his business
08:07 PM Tom_L: cheaper than hiring it done
08:08 PM Tom_L: single mode
08:09 PM Tom_L: ai-9 splicer $1000 if they wanna know
08:10 PM Tom_L: https://www.amazon.com/Signal-Fire-Splicing-Alignment-splicer/dp/B07SDNVP3T
08:10 PM rue_mohr: rather find a used one
08:12 PM Tom_L: https://www.ebay.com/itm/AI-9-Automatic-FTTH-Fiber-Optic-Welding-Splicing-Machine-Fiber-Fusion-Splicer/274568620496?hash=item3fed90a9d0:g:cnUAAOSw~hNfqcxj
08:14 PM aandrew: Tom_L: oh my, that pic
08:14 PM Tom_L: US shipping only :/
08:15 PM Tom_L: aandrew, like that?
08:15 PM Tom_L: you might like this then: http://tom-itx.no-ip.biz:81/~webpage/women_vs_men/whymendie_index.php
08:15 PM rue_mohr: Shipping:
08:15 PM rue_mohr: May not ship to Canada
08:17 PM Tom_L: are there any restrictions on shipping things like that across?
08:19 PM aandrew: there shouldn't be
08:20 PM aandrew: most ebay shippers who don't want to fuck around just ship to the continental US
08:20 PM aandrew: Tom_L: that ladder isn't too bad, the upside down jackstand though just makes me cringe
09:39 PM aandrew: wtf I just had an LM386 fall into my lap. like literally
09:40 PM aandrew: https://imgur.com/a/Elmnc88
10:24 PM rue_mohr: from nowhere?
10:25 PM rue_mohr: imugr will not allow firefox
10:42 PM Tom_L: sure will