#garfield Logs

Feb 25 2020

12:09 AM X704 is now known as tiwake
12:16 AM rue_bed: I think I have steppers large enough
12:32 AM rue_bed: - encoder interface for fpga
12:32 AM rue_bed: - editor
12:32 AM rue_bed: - motor driver fet modules
12:49 AM polprog: not familiar with north american style housing but is this a thing that the entrance leada directly to the living room?
12:49 AM polprog: leads*
12:50 AM polprog: re: https://twitter.com/skunkgod1/status/1231984176590540803?s=19
03:08 AM polprog: whats that CD40106 circuit doing there rue? isolated driver?
03:08 AM polprog: https://twitter.com/RueNahcMohr/status/1232140700424138752?s=19
05:07 AM Tom_L: hallway or hall
08:53 AM rue_mohr: polprog, recovers the differential state
08:53 AM rue_mohr: its a bistable (flip flop)
08:53 AM rue_mohr: polprog, hmm, livingroom or kitchen here
08:54 AM polprog: does it oscillate?
08:54 AM polprog: im trying to understand its workings..
08:54 AM rue_mohr: no
08:55 AM rue_mohr: when the transformer pulses, it latches depending on the pulse polarity
08:55 AM polprog: how, if the top ones output is highx the bottom ones input is low, so the bottom ones output is high, so the top ones input gets high
08:56 AM polprog: or am i missing something
08:56 AM rue_mohr: remember the one I did with a R S flipflop?
08:56 AM polprog: no, id say D and RS flipflops are my weak side
09:06 AM polprog: ok, off to side job
09:06 AM polprog: ive got a bunch of kids to teach linux to
09:08 AM rue_mohr: ok
09:08 AM rue_mohr: for the record
09:09 AM rue_mohr: the inputs of a logic chip bound the signals between the power rails
09:10 AM rue_mohr: if you take something like a pulse transformer, thats otherwise floating, and have it apply a 5V differential pulse to two pins of a logic chip, one side will end up logic low, and the other side logic high
09:11 AM rue_mohr: an RS flipflop maintains its state till one of only set or reset is triggered
09:11 AM rue_mohr: if the set is triggered, the output goes high
09:11 AM rue_mohr: if the reset is triggered, the output goes low
09:12 AM polprog: oh, that makes sense now
09:12 AM polprog: the transformer is floating, right
09:12 AM rue_mohr: so, if you apply the differential transformer to the flipflop inputs, it acts like a differential reciever
09:13 AM rue_mohr: the two inverters form an RS flipflop
09:13 AM rue_mohr: they hold each other in a constant state
09:13 AM rue_mohr: unless something over-rides the state of one of the inputs supplied by the resistors
09:14 AM rue_mohr: such as the transformer will
09:14 AM rue_mohr: interesting snag tho, the transformer, idle, holds both inputs at the same level
09:15 AM rue_mohr: this *COULD* be trouble, except that its made of schmidt triggers, which allow their outputs to be independent of the input *if* that input is between its trip-points
09:16 AM rue_mohr: anyhow, it works
09:20 AM Tom_L: morning
09:21 AM rue_mohr: hi
09:23 AM Tom_L: what's all the flipflop talk about?
09:24 AM rue_mohr: I was playing with restoring a pulse stream thru an ethernet transformer using a flipflop
09:24 AM Tom_L: oh
09:24 AM rue_mohr: it worked, interestingly enough to below 10Hz
09:24 AM Tom_L: emf don't bother the logic chip?
09:25 AM rue_mohr: its 5V pulses after the transformer
09:33 AM rue_mohr: I wonder
09:37 AM Tom_L: I wander
09:45 AM rue_mohr: might have just accidently found new door seals for my truck
09:45 AM Tom_L: are they leaking?
09:45 AM rue_mohr: falling apart
09:45 AM rue_mohr: esp the driver side window seal
09:46 AM Tom_L: i suppose the rubber is getting worn out by now
09:46 AM rue_mohr: aged pretty good
09:47 AM Tom_L: https://www.amazon.com/Rubber-Weatherstrip-83-94-Chevy-Blazer/dp/B077YVVNZ5
09:48 AM rue_mohr: !oh
09:48 AM rue_mohr: Currently, there are no sellers that can deliver this item to your location.
09:48 AM rue_mohr: :/
09:49 AM Tom_L: https://www.rockauto.com/en/catalog/chevrolet,1991,s10+blazer,4.3l+v6,1060535,body,door+seal+/+weatherstrip,1430
09:49 AM Tom_L: i get stuff from them when i need something
09:49 AM rue_mohr: the kit looked good
09:49 AM Tom_L: cheaper than local usually
09:50 AM Tom_L: shows canadian currency so maybe they ship there
09:50 AM rue_mohr: I'd rather the amazon kit cause I can tell it was complete
09:50 AM rue_mohr: wanna do a reship? :)
09:50 AM rue_mohr: I coldn't see what the price was tho
09:51 AM Tom_L: shipping would be quite a bit i bet
09:51 AM Tom_L: it's getting ridiculous
09:51 AM rue_mohr: hmm
09:51 AM rue_mohr: I know a guy who goes in every so often to a dropbox
09:52 AM Tom_L: i know z gets stuff from them too
09:52 AM rue_mohr: k I gotta get going
09:52 AM Tom_L: my day pooped out on me
09:52 AM Tom_L: til afternoon anyway
09:53 AM rue_mohr: I'd really like to go back to bed
09:53 AM rue_mohr: think I'm ill
09:53 AM Tom_L: just when it's supposed to start snowing...
06:35 PM rue_shop2: "1 rock, two stoned birds"
06:55 PM Tom_L: ?
07:19 PM rue_shop2: thought it was funny
07:42 PM rue_mohr: there is my list?
07:43 PM rue_mohr: <rue_bed> - encoder interface for fpga
07:43 PM rue_mohr: <rue_bed> - editor
07:43 PM rue_mohr: <rue_bed> - motor driver fet modules.
08:26 PM Tom_L: what's up tonight?
08:42 PM X704 is now known as tiwake
08:55 PM X704 is now known as tiwake
09:00 PM X704 is now known as tiwake
09:21 PM rue_bed: trying to wake up again
09:21 PM rue_bed: one of the guys dropped in right after work, and we changed the door on the small shed
09:51 PM X704 is now known as tiwake
10:05 PM rue_mohr: I think I'm awake
10:06 PM Tom_L: nope, this is a figment of your reality
10:08 PM rue_mohr: encoder
10:08 PM rue_mohr: ok, I need to think
10:09 PM rue_mohr: oh, and the UPS in the shop needs new batteries
10:09 PM Tom_L: new old batteries?
10:09 PM rue_mohr: yea
10:09 PM rue_mohr: if I'm using expired batteries I suppose I cant complain about having to change them lots
10:10 PM Tom_L: you could but it wouldn't matter
10:10 PM rue_mohr: so, for the motor driers
10:10 PM rue_mohr: I have an idea for the mosfets
10:10 PM rue_mohr: to make a module thats the flyback diode and the fet in one little package
10:11 PM rue_mohr: trying to come up with something for heat sinks tho
10:11 PM rue_mohr: I have an idea to use live heatsinks
10:12 PM Tom_L: live?
10:12 PM rue_mohr: yea, energized
10:12 PM Tom_L: is it a good idea?
10:12 PM rue_mohr: what do you think of the idea of having a pcb between a fet and a heatsink?
10:13 PM Tom_L: how's it gonna transfer heat efficiently?
10:13 PM rue_mohr: yea
10:13 PM Tom_L: so no.
10:13 PM rue_mohr: there was a 3rd idea,
10:14 PM Tom_L: hope it's better than #2
10:14 PM rue_mohr: no
10:14 PM rue_mohr: and its not really togethor
10:16 PM rue_mohr: I'd like to use the tabs instead of the middle pin
10:17 PM rue_mohr: I'd also like to build the drivers on cpu heatsinks with fans
10:17 PM rue_mohr: that I'd put aside for it
10:18 PM Tom_L: they gonna run that hot?
10:18 PM rue_mohr: 8 to-220
10:18 PM rue_mohr: iirc about 11w
10:19 PM rue_mohr: the big stuff in the world uses energized heatsinks
10:20 PM rue_mohr: this is the crazy stuff that stalls me up
10:20 PM rue_mohr: if I was young and only knew one way, I'd do it that way, and it'd work
10:21 PM Tom_L: so pretend
10:28 PM rue_mohr: hmm
10:28 PM rue_mohr: what should trace size be for 20A?
10:28 PM Tom_L: big
10:28 PM rue_mohr: yea
10:28 PM Tom_L: https://www.7pcb.com/trace-width-calculator.php
10:29 PM rue_mohr: do you know what the clad thickness was on your old fr4?
10:29 PM Tom_L: 2 oz
10:30 PM rue_mohr: .5"
10:30 PM Tom_L: did i send you some of that?
10:30 PM Tom_L: gee i completely forgot about that
10:30 PM rue_mohr: na
10:30 PM rue_mohr: I wish
10:30 PM rue_mohr: tho I havn't etched in a long time
10:30 PM Tom_L: the pile is still sitting in the garage
10:31 PM Tom_L: and the big sheets are in the basement
10:31 PM rue_mohr: so, 1/2" trace, means no 0.1" parts
10:31 PM rue_mohr: really?
10:31 PM rue_mohr: thought you said you tossed it!?
10:31 PM rue_mohr: I should come over
10:31 PM Tom_L: i kept enough to last my lifetime
10:31 PM Tom_L: i did toss a bunch
10:31 PM rue_mohr: quick, give the border a story to get around me not having a passport at the moment
10:31 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/etching/Copper3.jpg
10:32 PM Tom_L: still have probably 4-6 of those
10:32 PM rue_mohr: I know I know
10:32 PM Tom_L: double sided
10:32 PM rue_mohr: hey
10:32 PM rue_mohr: there is a prject for ya
10:32 PM rue_mohr: mechanically etch a 555 flasher
10:32 PM rue_mohr: use smt
10:32 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/etching/Copper1.jpg
10:32 PM Tom_L: and that
10:32 PM rue_mohr: its all toolchain work
10:33 PM rue_mohr: I been meaning to try it
10:33 PM rue_mohr: I prefer etching, but
10:33 PM Tom_L: i've never had luck milling pcb
10:33 PM rue_mohr: I dont remmber you trying
10:34 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/etching/mills.jpg
10:34 PM Tom_L: that's why i got those
10:34 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/etching/mills1.jpg
10:35 PM Tom_L: sherline was a shit testbed for that
10:35 PM rue_mohr: exactly
10:36 PM Tom_L: i never quite understood mosfet packaging
10:36 PM rue_mohr: heh
10:37 PM Tom_L: you'd think the legs would be big and fat
10:37 PM rue_mohr: yea, no room for traces
10:39 PM Tom_L: https://www.littelfuse.com/products/power-semiconductors/discrete-mosfets/n-channel-standard/very-high-voltage.aspx
10:39 PM Tom_L: those look a little more respectable
10:44 PM rue_mohr: ya know, I was doin fine till I had to add the flyback diodes
10:44 PM Tom_L: i can see what you mean using the heatsink as the power path
10:45 PM rue_mohr: 3 heatsinks
10:45 PM rue_mohr: DRV+, output A, output B
10:46 PM Tom_L: https://www.cde.com/resources/technical-papers/design.pdf
10:48 PM rue_mohr: igbt is awefull at low voltage
10:48 PM rue_mohr: drops 2V
10:48 PM rue_mohr: ea
10:50 PM rue_mohr: maybe I should make it moderatly normal and maybe just blow some up
10:51 PM Tom_L: what was wrong with that rif design?
10:52 PM rue_mohr: I dont think its set for 20A
10:52 PM Tom_L: probably not
10:52 PM Tom_L: i forget the mosfets on it
10:52 PM rue_mohr: oh its not abut the fets
10:52 PM rue_mohr: its heatsinking them
10:53 PM rue_mohr: and 20A traces
10:54 PM Tom_L: mmm i still have that eagle file
10:54 PM rue_mohr: whats it called?
10:55 PM Tom_L: hip4081_hbridge
10:55 PM rue_mohr: I didn't think rif designed that one
10:55 PM Tom_L: i dunno
10:55 PM Tom_L: no but he used it iirc
10:55 PM Tom_L: or at least told me about it
10:55 PM rue_mohr: it was part of openservo
10:55 PM Tom_L: maybe
10:57 PM rue_mohr: huh
10:57 PM rue_mohr: well, I ahve an idea that uses standing heatsinks
10:57 PM rue_mohr: I'd like everyone to be on their back
10:59 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/rue/4081/
10:59 PM Tom_L: still got eagle?
11:00 PM rue_mohr: hmm
11:00 PM rue_mohr: I should somewhere
11:00 PM rue_mohr: but I'm making *just* a fet board
11:00 PM rue_mohr: no control stuff on it
11:01 PM Tom_L: i did breadboard something using that chip but never actually put it to use
11:03 PM rue_mohr: heh
11:03 PM rue_mohr: I hahve mine all made up
11:04 PM rue_mohr: ok, I'm getting closer to something I like
11:04 PM rue_mohr: staggered leg thing
11:06 PM rue_mohr: yea ok
11:06 PM rue_mohr: pcb ends up upside down, but
11:09 PM rue_mohr: ok
11:09 PM rue_mohr: tumblers are falling into place
11:09 PM rue_mohr: I use two rows of devices, with staggered legs
11:09 PM rue_mohr: all wiring on pcb
11:09 PM rue_mohr: lots of wide floods work
11:10 PM rue_mohr: all deivces are insulted heatsink mounted to brackets
11:10 PM rue_mohr: that all come togethor to a cooled heatsink
11:10 PM rue_mohr: the pcb would have lugs made of standoffs screwed to the pcb, against the power traces
11:10 PM rue_mohr: (drv+, drv-, op A, op B)
11:11 PM rue_mohr: and 8 leads for driving fets
11:11 PM rue_mohr: (4 fets, 4 source connections as a ref)
11:13 PM rue_mohr: ok, this will be trickey to design
11:13 PM rue_mohr: DO I still have eagle on here?
11:13 PM rue_mohr: wait
11:13 PM rue_mohr: I use the old dos program
11:13 PM rue_mohr: ok
11:14 PM rue_mohr: ok I need 4 of these modules
11:14 PM Tom_L: i got 4.16 if you want it
11:14 PM rue_mohr: x, y, z, x
11:14 PM Tom_L: windows though
11:14 PM rue_mohr: no I cant etch from eagle
11:15 PM rue_mohr: I have an old dos program that can run my plotter
11:15 PM rue_mohr: which is burried under stuff :)
11:15 PM Tom_L: no surprise there
11:16 PM rue_mohr: actually, everything is burried under stuff right now
11:16 PM rue_mohr: its terribly frustrating
11:16 PM rue_mohr: but I have a design, this is good
11:17 PM rue_mohr: I'm getting closer to having to work out the servo stuff
11:17 PM Tom_L: gnite
11:45 PM rue_mohr: img!
11:46 PM rue_mohr: if I can find the right terminal
11:47 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1150312.jpg
11:48 PM rue_mohr: 4 fets, 4 diodes (to-220)
11:48 PM rue_mohr: all the screws can be reached
11:48 PM rue_mohr: I can use the heatsink with the fan
11:49 PM rue_mohr: the hip8041 and stuff go on another board