#garfield Logs

Nov 24 2020

12:49 AM rue_bed: its not morning
01:40 AM rue_shop1: damnit, a plain old NPN transistor works just fine for the sample and hold
01:40 AM rue_shop1: I bought like $30 of junk from aliexpress
04:35 AM rue_bed: I think to do the timing for all this I need a 4017 and diode rom
04:35 AM rue_bed: which means I need an oscillator too
04:35 AM rue_bed: osc -> 4017 ->
04:35 AM rue_bed: triangle wave gen
04:35 AM rue_bed: dc-dc
04:35 AM rue_bed: sample & hold
04:36 AM rue_bed: the triangle wave gen can be a bipolar current source and cap
04:36 AM rue_bed: the dc-dc will just be annoying
04:37 AM rue_bed: probably 2 transistors, 4 diodes, and 4 caps
04:37 AM rue_bed: the sample and hold is just a transistor and cap
04:37 AM rue_bed: ... and resistor
04:38 AM rue_bed: this isn't being a nice small cirucit..
04:39 AM rue_bed: if I divide 20khz into 4
04:39 AM rue_bed: no 8
04:40 AM rue_bed: if I divide 40kHz by 8
04:40 AM rue_bed: with the first 4 going to charge the triangle gen
04:40 AM rue_bed: and the second two triggering the s&h
04:41 AM rue_bed: and then I can use every second one for running a dc-dc
04:41 AM rue_bed: or just take the dc-dc off the input osc
04:41 AM rue_bed: meaning 6 diodes
04:41 AM rue_bed: the osc can be a transistor/inductor osc
04:43 AM rue_bed: that said, a 555 is a 3 component osc
04:44 AM rue_bed: but I think I'm sitll only at 2 op-amps then
04:45 AM rue_bed: lm555 is 10c ea
04:45 AM rue_bed: 4017 is 14c
04:45 AM rue_bed: damn for needing lots of these
04:46 AM rue_bed: I should check out the capacitor version before I go tooo crazy
06:54 AM aandrew: rue_mohr: it's your time to shine: https://hackaday.com/2020/11/23/a-cpu-less-computer-with-a-single-nor-gate-alu/
06:55 AM aandrew: they're still too high tech, you need to show them the SRAM state machine
07:32 AM Tom_itx is now known as Tom_L
08:31 AM rue_mohr: heh
08:38 AM rue_mohr: intersting
08:40 AM rue_mohr: wow thats a lot of work into softare
09:37 AM aandrew: I'm sure you could replace that NOR gate with a 555
09:37 AM aandrew: might have to connect pin 4 though
09:37 AM rue_mohr: :)
09:37 AM rue_mohr: pin 4 is great for a startup delay
09:38 AM rue_mohr: its something like 10ms/0.1uF
12:59 PM rue_mohr: polprog,
12:59 PM rue_mohr: cross check
12:59 PM rue_mohr: if I apply 5V to 5mH, then at 100us I will have 100mA, right?
01:40 PM polprog: oh
01:40 PM polprog: thats not my strong side
01:40 PM polprog: :(
01:40 PM polprog: u =di/dt
01:40 PM polprog: Ldi/dt
01:41 PM polprog: LI=Ut
01:41 PM polprog: I=Ut/L
01:42 PM polprog: I=5*100u/5m=100u/m =100 e-6/1e-3 =100 e-3
01:42 PM polprog: yes
01:42 PM polprog: i just finished solving all 3 peculiar AC problms from the past exams
01:43 PM polprog: now ill be explaining them to friends...
01:43 PM rue_mohr: yea, I figured you would have been on something close
01:43 PM rue_mohr: cant recall if your program does that
01:43 PM polprog: i think it doesnt.. hmm
01:43 PM rue_mohr: did mine? was I just doing capacitors?
01:43 PM rue_mohr: hmmm
01:44 PM polprog: and trying to figure out how does one make so many mistakes in the answers
01:44 PM polprog: like, its as if he didnt even read that
01:44 PM polprog: hes dyslexic
01:44 PM polprog: he said so
01:44 PM polprog: so get someone else to read it once
01:44 PM polprog: jesus
01:45 PM rue_mohr: one of the people being helped?
01:45 PM polprog: dividing 10k by sqrt 3 and getting 500 as a result
01:45 PM polprog: no, the proffesor
01:45 PM rue_mohr: oh
01:45 PM rue_mohr: you would expect that to fail a sanity check
01:46 PM polprog: im still not sure if id pass the exam
01:46 PM polprog: yes, thats the thing!
01:46 PM rue_mohr: 10000/~1.4 ?
01:46 PM polprog: yeah
01:46 PM polprog: 1.6 iirc
03:05 PM Tom_L: profs don't give a shirt about anything but their pay
03:07 PM Tom_L: they should be paid based on your ability to learn from them
04:33 PM rue_mohr: yea, I think performance based pay would bea good idea for doctors too
04:35 PM Tom_L: any of the professional 'arts'
04:35 PM Tom_L: dr lawyer, teacher etc
05:34 PM polprog: damn
05:34 PM polprog: done
05:34 PM polprog: you know what
05:34 PM polprog: i found another error
05:35 PM polprog: i think he just copied and pasted this and that from his other notes
05:35 PM polprog: somehow he got RMS voltage to be 3.5mA
05:35 PM polprog: lol
05:38 PM rue_mohr: hahahaha
05:38 PM rue_mohr: that could be a LOT of voltage
05:38 PM polprog: idk what is the source of errors like
05:39 PM polprog: Irms = Irms/sqrt(3) = ...
05:39 PM polprog: i think this was the same like like a signal generator output being 5.7kV
05:39 PM rue_mohr: haha
05:39 PM polprog: like
05:39 PM rue_mohr: so, he steps over reasonable
05:39 PM polprog: how the hell that doesnt even turn a red light on in your mind
05:40 PM rue_mohr: his mind isn't on it
05:41 PM polprog: anyway
05:41 PM polprog: ive solved all 3 problems from the previous exams he posted on the site
05:42 PM polprog: with some help from my grandpa
05:42 PM polprog: then i solved them again with my friends
05:42 PM rue_mohr: ? yu got a granda into stuff?
05:42 PM polprog: yes
05:42 PM polprog: didnt i tell you?
05:42 PM rue_mohr: I dont think so
05:42 PM polprog: he finished control theory iirc
05:43 PM polprog: then went onto supervise building foundries in southern europe
05:43 PM polprog: huge electric arc furnances
05:43 PM rue_mohr: aluminum
05:43 PM polprog: hmm, i dont remember
05:44 PM polprog: but he didnt wanna talk about your smelting :/
05:44 PM rue_mohr: hahaha
05:44 PM polprog: he says that if theres oxygen access to the aluminum, its gonna be shit not metal
05:44 PM rue_mohr: yea, see, hes familiar with aluminum
05:44 PM polprog: i mean, i wouldnt build a scaffold out of your molten alu :) but i guess its more than enough for some small handles and jazz
05:45 PM rue_mohr: aparenlty hydrogen is the kicker
05:45 PM polprog: no offence :D
05:45 PM rue_mohr: no he's right, I'v seen well cast metal
05:46 PM rue_mohr: degassing is a bitch
05:46 PM rue_mohr: ok I need to stop printing these lvdt spools
05:47 PM rue_mohr: I can wind one tonight
05:51 PM polprog: is it like with regular printers
05:51 PM polprog: you click print, it doesnt do anything, so you click over and over again and suddenly bam, 50 lvdt spools?
05:57 PM polprog: ive got a POS keyboard, it's programmable
05:57 PM polprog: PS/2
05:57 PM polprog: also got an ibm pos computer and a touchscreen for it
05:57 PM polprog: pretty neat
05:57 PM polprog: works in linux
05:58 PM polprog: doesnt work in dos sadly
05:58 PM polprog: the nice thing is that it's a pretty large screen 14" at least
05:58 PM polprog: it runs off a DC jack *or* a poweredUSB 12V
05:58 PM polprog: and VGA
05:59 PM polprog: the programing software didnt detect the keyboard though (DOS)
05:59 PM polprog: maybe i have to try the older computer
05:59 PM polprog: i have a pentium pro system sitting there, ill try to use that
06:00 PM polprog: they keyboard has got enough 1u switches for the whole alphabet, i could program it to be a nice compact board
06:00 PM polprog: 84 keys
06:00 PM polprog: not all populated
06:00 PM polprog: ill have to set up an alert on ebay if someone sells a similar one, or sells keys
06:02 PM polprog: or see of any of the other keyboards' keycaps have stems that fit...
06:02 PM polprog: it's a Preh MC 84
06:32 PM rue_mohr: :)
06:33 PM rue_mohr: did you see the industrial keybaord I have?
06:33 PM rue_mohr: 1 broken key :(
06:35 PM rue_mohr: ok, 5mH charge time
06:54 PM rue_shop1: polprog, we were out by a factor of 2 on the current
06:54 PM rue_shop1: its only peaking at 50ma
06:54 PM rue_shop1: how is that even possible?
06:54 PM polprog: rue_shop1: the keyboard you use at the PC?
06:54 PM polprog: hmm, weird
06:54 PM polprog: oh
06:55 PM polprog: hmm
06:55 PM polprog: no
06:55 PM polprog: weird
06:55 PM polprog: oh maybe we didnt integrate
06:55 PM polprog: but thats Ldi = Udt
06:55 PM polprog: so we get LI = Ut okay...
06:56 PM polprog: no, we must have integrated wrong
06:56 PM polprog: it should be calculated from the conservation of energy law
06:56 PM polprog: that has the 1/2 factor in it
06:56 PM rue_shop1: its got an RC curve to it tho
06:56 PM polprog: huh
06:56 PM rue_shop1: which might be the error
06:56 PM polprog: or that
06:56 PM rue_shop1: it should be a linear ramp
06:57 PM rue_shop1: maybe the series resistnace of these coils is too high
06:57 PM polprog: maybe
06:57 PM rue_shop1: measuring
06:58 PM rue_shop1: hahah damn near 80 ohms
06:58 PM rue_shop1: ok ok
06:58 PM rue_shop1: I need a "502"
06:58 PM polprog: oh boy, thats a lot
06:58 PM polprog: i need to shower...
07:33 PM rue_shop1: ok I made one
07:33 PM rue_shop1: testing
07:33 PM rue_shop1: 6mH
07:33 PM rue_shop1: whatever
07:35 PM rue_shop1: ok I do need to back it down
07:37 PM rue_shop1: ok taking off 34 turns
07:41 PM rue_shop1: uhm..,..
07:43 PM rue_shop1: 7.8 ohms 5mH
07:43 PM rue_shop1: hmm
07:54 PM rue_shop1: ok, there!
07:54 PM rue_shop1: and its sinking 12mA into the flyback suppression...
08:01 PM rue_shop1: ok it works
08:01 PM rue_shop1: I had to adjust the duty properly (I was giving the inductor more resset time) and get the flyback dealt with
11:39 PM rue_mohr: I'd tried a toroid and another closed-magnetic-loop thing, but they both saturate
11:39 PM rue_mohr: I suppose it IS only 5kHz