#garfield Logs
May 19 2020
#garfield Calendar
01:01 AM rue_mohr: heh
01:07 AM rue_mohr: Tom_L, aka why I'm shimming with twitter
02:14 AM zhanx_ is now known as zhanx
03:21 AM polprog: morning
09:32 AM rue_mohr: hi
09:32 AM rue_mohr: bye
09:32 AM rue_mohr: off to work
09:33 AM polprog: same
09:33 AM polprog: todays labs will be about transistors and digital electronics
09:33 AM polprog: streaming for the kids i teach as a side gig
01:37 PM polprog: done..
02:00 PM Tom_itx is now known as Tom_L
02:02 PM rue_mohr: home for lunch
02:03 PM rue_mohr: two of the cores I'v been boiling to dissassemble came apart, not broken!
02:03 PM Tom_L: cores?
02:03 PM rue_mohr: some small ferrite transformers off printer pcbs I was dissassembling last night
02:05 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1160212.jpg
02:05 PM rue_mohr: its made for high voltage
02:05 PM rue_mohr: but I think its good for 2-4W
02:05 PM rue_mohr: zhanx, your back!
02:06 PM rue_mohr: I had a link for you
02:06 PM rue_mohr: a long time ago
02:07 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1160213.jpg
02:09 PM rue_mohr: I'm still trying to find a pattern to twitter likes
02:48 PM rue_mohr: hmm, wonder if that should have been "I can see a bug in the hardware"..... :)
02:49 PM rue_mohr: cant think of a comeback with somehting about a software workaround
03:16 PM aandrew: how do you depopulate the boards? I always like the blowtorch and hit the board against something hard method
03:17 PM aandrew: for bulk removal anyway
03:59 PM aandrew: what do you do with the resulting "broth" from boiling transformers?
04:22 PM polprog: rue eats cores for lunch.. you have to boil them for 6h for the cores to soften
04:22 PM polprog: very rich in iron!
04:23 PM polprog: rue, i can see you liked my DOA 486 board
04:25 PM polprog: what are you driving the VFD with? avr or something moar?
04:30 PM polprog: rotated the scope a bit and it seems to be more comfy to use now
05:54 PM Tom_L: https://www.youtube.com/watch?v=g0OLIth_cOI&feature=youtu.be
05:55 PM polprog: built another of these psus
05:55 PM polprog: i soldered a slightly useless vfd onto a protoboard.. ill be making this a testbed
05:55 PM polprog: the filament voltage seems too low for *some* VFDs, or something
05:55 PM polprog: some are very faint
06:01 PM rue_mohr: aandrew, combination
06:01 PM rue_mohr: some things a heat gun (patiently!)
06:02 PM rue_mohr: some things a soldering gun and an air nozzel
06:02 PM rue_mohr: other things (2 and some 3 pin) just by iron
06:02 PM rue_mohr: heh, the broth has shelac of some sort in it
06:03 PM rue_mohr: just a mega 8
06:03 PM rue_mohr: polprog, ^
06:03 PM polprog: k
06:03 PM rue_mohr: want a firmware copy?
06:03 PM polprog: i guess
06:03 PM rue_mohr: yea, your filament should be about 5V off ground for most vfds
06:04 PM rue_mohr: then about 24-30V on your drive
06:04 PM rue_mohr: the software does the same thing as for an led array
06:04 PM rue_mohr: walk thru the grids and light up the segmetns you want
06:05 PM polprog: that means my drive is on the low side.. just 20
06:05 PM polprog: i think.. that depends on the core
06:05 PM polprog: getting 30 from 5 is hard
06:05 PM polprog: it coult probably run off 9V
06:08 PM rue_mohr: gee twitter is going crazy latley
06:08 PM rue_mohr: usually you generate the votlage by turns ratio
06:08 PM rue_mohr: so, if you have 10 turns/5V on your input, and you want 30V output, you put on 40 turns for the output
06:09 PM rue_mohr: I cheat, I start with the input votlage
06:09 PM rue_mohr: so
06:09 PM rue_mohr: I'm converting 12 to 36
06:09 PM rue_mohr: I have a centertapped winding for input
06:09 PM rue_mohr: that, by its nature will give me 24
06:10 PM rue_mohr: so on each side I added another winding, same turns as the input, which gives me the input votlage again
06:10 PM rue_mohr: 12+12+12 = 36
06:10 PM rue_mohr: input -mmmm-+-mmmm-
06:10 PM rue_mohr: where + is the center tap
06:11 PM rue_mohr: center tap at 12V
06:11 PM rue_mohr: when you ground the left side, the right goes up to 24V
06:11 PM rue_mohr: so
06:12 PM rue_mohr: add boost windings -mmmm-R-mmmm-+-mmmm-L-mmmm-
06:12 PM rue_mohr: + goes to 12V
06:12 PM rue_mohr: R and L go to the transistors
06:12 PM rue_mohr: the far left and far right are my voltage output
06:12 PM polprog: i mean with my toplogy (flyback)
06:40 PM aandrew: rue_mohr: yes, how do you dispose of it
06:44 PM aandrew: wait what
06:44 PM aandrew: er no
06:44 PM aandrew: 1-mmmmm-R-mmmmm-+-mmmmm-L-mmmmm-2
06:45 PM aandrew: + is 12V, R/L go to transistors which ground those points alternately, and 1/2 are your outputs? That's not flyback is it? there's no isolation there
06:45 PM aandrew: ugh. nevermind. I'm confusing polprog and rue's lines in the chat
06:45 PM polprog: deosnt hurt..
06:46 PM polprog: we do the same things most of the time so you get a 24h coverage
06:46 PM aandrew: I don't understand rue's topology though, it's like a dual boost or something
06:46 PM aandrew: lol
06:46 PM polprog: me neither..
06:46 PM polprog: my topology is
06:46 PM polprog: https://polprog.net/rozne1/ircjunk/vfd_dcdc/IMG_20200518_234759.jpg
06:47 PM aandrew: hm your transformer has 12V on both center taps too
06:48 PM polprog: yeah
06:48 PM polprog: well, 5
06:48 PM polprog: not 12
06:48 PM polprog: hmm. theres a mistake
06:48 PM polprog: that should be 5, it runs fine off 5V :P
06:48 PM aandrew: :-)
06:49 PM polprog: im not really sure why. the same thing on p55nf06 doesnt generate that much flyback
06:49 PM polprog: must be the quirky nature of uln2003
06:50 PM polprog: they are darlington arrays, maybe i should get 4 mosfets and make a darlington too
06:50 PM polprog: and 4 power BJTs
06:50 PM polprog: see what happens
06:50 PM polprog: the ULN here runs a tad warm
06:50 PM polprog: ill glue a heatsink to it, and call it a day
06:54 PM rue_mohr: aandrew, yea thats pushpull
06:54 PM rue_mohr: polprog, has a strange pushpull flyback going on
06:54 PM rue_mohr: its a pushpull boost
06:54 PM rue_mohr: (mine)
06:55 PM rue_mohr: yea I use 12V on my transformers, cause I regard 5V as logic supply
06:55 PM rue_mohr: polprog, fets drive differently thean bipolars tho
06:56 PM rue_mohr: a bipolar turns on at 0.6V, the fet is around 2V
06:56 PM polprog: yeah, thats why i wanna see if i get that pushpull flyback effect on bjts
06:56 PM rue_mohr: polprog, most of the uln heat is probably the diodes
06:56 PM rue_mohr: it was with mine, give a try to some external doides and almost all the heat might go away
06:56 PM polprog: im using an ext diode
06:57 PM polprog: for some reason - maybe core related
06:57 PM polprog: it takes 300mA
06:57 PM polprog: the other core took 150
06:57 PM polprog: the core is a bit warm as well
06:57 PM polprog: maybe ill just wind more turns on the next unit
06:58 PM aandrew: rue_mohr: it's weird to see the transistor in the "middle" of the inductor like that
06:58 PM polprog: so i got 24 turns each on the primary
06:59 PM rue_mohr: my peaks were 1A
06:59 PM polprog: anf heck knows how much on the sec
06:59 PM polprog: ill try to add a cap later
06:59 PM polprog: the sec took about 1/2 the length of wire than the pri
06:59 PM rue_mohr: becase it runs on core saturation, the core will be warm
06:59 PM polprog: allright
06:59 PM polprog: glue
06:59 PM polprog: sleep
06:59 PM polprog: in that order
07:00 PM aandrew: if you sniff hte glue it'll help you sleep
07:00 PM polprog: not this kinda glue
07:00 PM rue_mohr: https://threeneurons.wordpress.com/miscellaneous-projects/
07:00 PM rue_mohr: there are a few good ones in there
07:01 PM rue_mohr: https://threeneurons.files.wordpress.com/2014/12/royer01s.jpg?w=500
07:01 PM rue_mohr: that is the one I was going to point to
07:01 PM aandrew: haha
07:02 PM aandrew: "simple toy which makes a noise very similar to a phasor firing, on the orignal Star Trek series. The real one, with Capt. Kirk, not that woosy Picard"
07:03 PM aandrew: rue_mohr: right, that's a more traditional push-pull oscillator
07:04 PM polprog: lol
07:05 PM polprog: makes sense
07:05 PM aandrew: hm, so when one transistor turns OFF, the voltage reversal in that half of hte coil goes to yank the OTHER transistor's base low, turning it off?
07:05 PM polprog: so the current setup runs fine at 5v on most of the displays.. makes the test VFD filament glow (!) at 12V
07:06 PM rue_mohr: oh I'm not 100%
07:06 PM rue_mohr: I have it on my list to set one up and analize the hell out of it
07:06 PM polprog: more like one transistor turning on causes the voltage on the other half to go low, turning itself off?
07:06 PM polprog: ugh
07:06 PM aandrew: so, for example, R1/R3 bias Q1's gate so the transitor is on.
07:06 PM polprog: i stop concluding at this point
07:07 PM aandrew: when the voltage on Q1's drain falls below R1/R3 - diode drop, it starts pulling Q2 gate down
07:07 PM aandrew: slowly turning it off... hhm
07:07 PM aandrew: no it's nto the negative spike, it's the low Ron and the bias point
07:07 PM aandrew: C1 seems critical from the notes
07:08 PM polprog: hmm
07:08 PM polprog: ill check tomorrow if you got it figured out
07:08 PM polprog: :P
07:08 PM polprog: nighters
07:08 PM aandrew: night polprog
07:09 PM aandrew: he's got another Royer there farhter down with non-ct transformers
07:19 PM rue_mohr: thats got a few good ones
07:30 PM rue_mohr: https://encrypted-tbn0.gstatic.com/images?q=tbn%3AANd9GcSDfUVy2Nm7hUZn1a_hzR6ssMrG-oCRMuFbCPnY23TSrDTVHEwZ&usqp=CAU
07:32 PM rue_mohr: ok software for polprog
07:32 PM rue_mohr: thats the next goal
07:32 PM rue_mohr: wake up rue
07:38 PM aandrew: hm that too uses a single cap between the transistors
07:38 PM aandrew: tiny pic though
07:38 PM rue_mohr: I know
07:38 PM aandrew: I gotta work on software tongiht too
07:38 PM aandrew: supposed to have done this like 2.5 weeks ago
07:38 PM aandrew: I've been "percolating" but now I gotta just crank it out
07:39 PM rue_mohr: I'd help, but I'm falling asleep
07:40 PM rue_mohr: hmm how do I github
07:41 PM aandrew: hm?
07:41 PM rue_mohr: damn it doesn't remember me, how the hell am I supposed to know who I am
07:41 PM Tom_L: i can find you
07:42 PM Tom_L: https://github.com/ruenahcmohr
07:42 PM Tom_L: just like that
07:42 PM rue_mohr: ah, it does have the middle name
07:42 PM rue_mohr: hey I got the password on the second try!
07:43 PM aandrew: lol
07:43 PM Tom_L: good thing or i'd have to tell you that too
07:43 PM Tom_L: :)
07:43 PM rue_mohr: :)
07:44 PM Tom_L: got the pi running the mill tonight
07:44 PM rue_mohr: ah cool
07:44 PM rue_mohr: which interface?
07:44 PM Tom_L: eth
07:44 PM Tom_L: you see the yt?
07:45 PM rue_mohr: ....no
07:45 PM Tom_L: ^^
07:45 PM rue_mohr: wow way back there
07:45 PM rue_mohr: thats the pi?
07:45 PM rue_mohr: on the lcd?
07:45 PM Tom_L: yes
07:46 PM rue_mohr: huh
07:46 PM Tom_L: lcd?
07:46 PM Tom_L: no lcd there
07:46 PM rue_mohr: the monitor?
07:46 PM Tom_L: yeah
07:46 PM rue_mohr: you need a bigger motor eh?
07:46 PM rue_mohr: spindle
07:47 PM Tom_L: motor leads to more power which i don't have for it
07:47 PM rue_mohr: heh
07:47 PM Tom_L: it woudl be nice
07:47 PM rue_mohr: did you see the huge DC motor I got from the scrap pile last year?
07:47 PM Tom_L: no
07:47 PM rue_mohr: I'll photo it later
07:47 PM rue_mohr: its for an OLD cart
07:47 PM Tom_L: hmm
07:48 PM rue_mohr: its one of those 75lb 8"x16" things
07:48 PM rue_mohr: 24V, likley 200A
07:48 PM Tom_L: no good for a spindle
07:48 PM rue_mohr: I grabbed it for a spindle, but realized its more valuable as a power train
07:48 PM Tom_L: too heavy
07:49 PM rue_mohr: that said, I made the power cart from a scooter, so ..
07:49 PM rue_mohr: agreed
07:57 PM rue_mohr: polprog, https://github.com/ruenahcmohr/595VFD
07:59 PM rue_mohr: my eyes keep closing, wonder what I can get done
10:13 PM rue_shop1: mmm stripping pcb's