lordpil's blog


http://video.gq.com/watch/the-closer-with-keith-olbermann-how-would-trump-deport-11-million-people-think-about-it
really, really missed his anger

Posted by renesis at 19:45 | permalink | 0 comments


yup
i spent almost a week trying out diff gcc/eclipse based shit
got it working
tried keil, shit up in minutes
and im not sure, i dont think so but im not sure
keil is first party, prob why it actually works

Posted by renesis at 19:22 | permalink | 0 comments


nice
keil?

Posted by renesis at 19:15 | permalink | 0 comments


engineering pad paper can take a heroic amount of erasing

Posted by renesis at 17:47 | permalink | 0 comments


i dont like that EE dep at school is like, ya fuck units
its too hard to figure that you need to put a V/I (or ohms) on the independent sources, so theyre just like naw fuckit
ME department hella good about units, and i catch math errors a ton because units dont cancel or combine right
er, dependent source
yeah totally
engineering they let you work with actual numbers tho
and check your work along the way
well, except the ass teacher last semester who just got chewed out for not teaching me mesh and node
but yeah, fuck working shit out with variables until the very end
no way to see if your results are sane along the way
note that the physics teacher who was worse about this was admitted a shitty experimental physicist
MAYBE IF YOU ACTUALLY PUT NUMBERS INTO YOUR FORMULAS AND CHECKED YOUR WORK SHIT WOULD MAKE SENSE AT THE END

Posted by renesis at 17:24 | permalink | 0 comments


theyre 260mAH, itll be a little fire
mine has a cooler
i have a makita charger, it has a fan and little vents that mate with ducts in the battery
milwaukee is good
like, 3 speed gearbox?
thats neat
it has an ultra low gear?
ha
do you use it most?

Posted by renesis at 17:18 | permalink | 0 comments


but if i charge at 8C i can charge both batteries in 15 minutes!
WITH ONE CHARGER!
im hungry
i should fix my car already
or sell it

Posted by renesis at 17:06 | permalink | 0 comments


i like that the hk thing has adjustable current
where is it on that?
charge rate: .3A
its an 18650 charger, the cheap ones are ~.3A
yeah but the $13 thing has a knob
$30 for dual 2s/3s charging, up to 4.5A
they say 35-70C discharge, i can prob charge them faster than .3A

Posted by renesis at 15:31 | permalink | 0 comments


http://www.hobbyking.com/hobbyking/store/__26492__HobbyKing_8482_E4_Balance_Charger_US_Warehouse_.html
fuckit i get two of these
i hate the gator clip input
theres like, quad shit thatll do 4s or 5s or something
ya
i dont want to pay that much tho, i think they were like 70+
http://www.hobbyking.com/hobbyking/store/__18327__Turnigy_TQ4_4x6S_Lithium_Polymer_Battery_Pack_Charger_AU_Warehouse_.html
heh, i like their ID
right im just saying, i think thats the cheapest ive seen

Posted by renesis at 15:26 | permalink | 0 comments


yeah i was going to get the turnigy 2s thing with a/c input, but my thing has two packs for bipolar supplies so cheap dual charger would be neat
like, theyre 260mAH packs so not like charging them one at a time would take very long

Posted by renesis at 15:21 | permalink | 0 comments


timecop: you knowa cheap charger can do two 2s packs w/ jst connectors?
like, that isnt a fire hazard
preferably something with a recognizable brand, but ill click anything

Posted by renesis at 15:16 | permalink | 0 comments


if that doesnt kill you, you probably want to be dead

Posted by renesis at 13:21 | permalink | 0 comments


omg unlocked last TIS 100 square

Posted by renesis at 12:28 | permalink | 0 comments


fuck, im playing tis100 again

Posted by renesis at 09:52 | permalink | 0 comments


rdy for kickstarting

Posted by renesis at 06:38 | permalink | 0 comments


design doc section pretty good
seems to step by step the shit, instead of just throwing dozen equations at you
what i hate is they use x/y as a variable
but x/y shows up in the equations as not the variable
took me forever to figure out the equations because of that, ti, on semi, fairchild, i think nxp, same equation table
oh i thought you were champion for a sec
champion does not have $5k to spend on led project, is the point

Posted by renesis at 06:32 | permalink | 0 comments


its like a mosfet mc34063 with lower vin
prob does buck/boost/buckboost
still think get a surfboard or just solder wires to pins
and perfboard the shit
its like an $8 chip
is the downside, itd be pretty shit for a product unless that was the entire product and still kinda shit
the $4 chips only go to 15v
its because its higher voltage than the other low vin chips
also mosfet
yeah if youre broke, china does this shit for like, outlet psu blocks
you can afford an $8 chip

Posted by renesis at 06:27 | permalink | 0 comments


those switch IC always work, tho
and its not like youre trying to make a high margin product
blackmoon: did you see link i did?
its like, same shit less pins, but down to 0.8v
hmm doesnt have timing cap?
well, he can do the coil like that for any of the solutions
if theres a size issue with the train floor
and do you need neg voltage?
im not sure if kevtris means like, buck-boost or buck/boost
it does if they have something common

Posted by renesis at 06:22 | permalink | 0 comments


the chip i linked is 0.8v to 28v in, and can do boost/buck and inverting, and is a kind of common configuration
well all you would need is current lim resistors, because fixed boosted voltage
so you learn how to setup a pretty typical general purpose switching reg/controller, and the LED part is (V-Vf)/If simple
yeah i mean how you do the boost switching is kind of the variable
but the coil, switch, and cap part is going to be the same
you can adapt design equations you find in general purpose switching regulators, if you want to do your own
like a discreet boost reg?
i think because phase shifted transformer feedback is prob the chinacheap way

Posted by renesis at 06:16 | permalink | 0 comments


but you could just jack the voltage way up and use resistors
get really accurate current
thats why i think you should use a boost reg
gpf: i mean, this isnt really diff
it just has a voltage ref and comparator feedback
and somewhat accurate timing
those little multi config switchers are pretty inaccurate and unstable with switching freq, cap sets it and its depending on vin and some other shit
but itll get you within 20% and it wont ever have startup issues
champi0n: can you make pcb?
or solder wires onto SMD pins?
yeah i think you should try and make a boost reg
its pretty simple
yes just with resistors

Posted by renesis at 06:11 | permalink | 0 comments


http://www.digikey.com/product-detail/en/maxim-integrated/MAX629ESA-/MAX629ESA--ND/948123
maybe this
kind of looks like a mc34063
w/ mosfet switch, tho
champi0n: how short?
those 'big' square SMD ferrite core coils are maybe 5mm tall, for what youll need
yeah lots of ways to do the current reg after the boost

Posted by renesis at 06:06 | permalink | 0 comments


this is for little leds?
whats the train run on?
fuck a 555, theres prob some 1.8v boost reg that will work
yeah get some 1.8v shit
but its a 555
id rather use a 1.8v uc to make squares!
i <3 inverter hacks tho
like, one shots and supply sequencer type stuff
or just in parallel because fuck a pair of transistors

Posted by renesis at 06:00 | permalink | 0 comments


there are probably curves in the datasheet that answer your question

Posted by renesis at 05:36 | permalink | 0 comments


jero32: its a funny joke because its true
ive personally read about a few examples, so my guess is its just a common mak:durr strategy
i need to time two things, so i need two arduinos

Posted by renesis at 04:58 | permalink | 0 comments


in a ms tick loop environment, you should be able to code delays extremely simply with accuracy and the ability to process in parallel
the thought of a uc just nop looping to create timing, in 2016, is pretty fucking disgusting
i blame arduino
that said, i wonder if macegr still has my demo code up with pasted nop delays
i didnt even bother with loops, code didnt use any flash otherwise so just hammered ^v

Posted by renesis at 04:25 | permalink | 0 comments


systick
which i think is a cortex thing, not an st thing
i think by default its setup for 1ms
i actually like that a lot as i almost always used a counter tick to run my mainloops
i think those delay function just compare to systic counter
delay functions are always pretty sketchy
ive been like 'fuck c' and just done shit in asm because delay functions were either to coarse in res or couldnt go low enough
or were just wrong
like they dont acount for loop overheads on small delay values
like, HAL delay()s make me think of people buying multiple arduinos for simple projects which required parallel timing

Posted by renesis at 04:20 | permalink | 0 comments


no its basically the replacement
i think some simpler applications, you can get it to generate finished code
setting up counter functionality and DMA based stuff
cubemx will generate source code with peripherals setup, so things that dont really depend on mainloop interactions will sometimes just work
right, but something like a comm bridge you can maybe just click click generate done
yeah usb and some fs stuff and realtime stuff are middleware with diff licensing shit

Posted by renesis at 03:52 | permalink | 0 comments


rab: theres like, one decent st doc for the stm periph library i found
and i dont remember finding something similar for cubemx

Posted by renesis at 03:33 | permalink | 0 comments


jero32: ya
theres more than one, heh
ya
pops out a source tree, it almost works with that workbench ide thing

Posted by renesis at 02:49 | permalink | 0 comments


lope: im not sure BMS is clearly defined
like, if someone sold something that just tracked cell voltage and opened the batter pack on low voltage, you could prob sell that as a cheap BMS to customers
some sort of low voltage protect override to allow charging, no-bypassable overvoltsge shutoff
and itd be safe, shrug
nik1: yes, exactly, the same as the MAKE:DUR community
but i mean, if you sell a range of BMS for better makrup on more expensive units, you need a low functionality BMS in the range

Posted by renesis at 02:28 | permalink | 0 comments


no problem, good luck
i think the nature of most of those things is that they will work okay in parallel with charging voltage
i think balancers are just switching in parallel loads to drop current in the higher voltage cells
so operation isnt affected to much by charging
its possible they go super cheap somehow and it doesnt disconnect the charging voltage, and cant shunt enough current around the batteries to keep them from overvolting
if both of those things are the case I guess charging with a high current charger could be dangerous
but i think these things are basically made to avoid that so prob fine
timecop has experience with BMS, so might have specific product suggestions

Posted by renesis at 02:15 | permalink | 0 comments


short of opening the thing up and going about standard troubleshooting i got nothing
if you can reproduce it by excersizing particular functions of the synth thats clues that there is something wrong inside
ya
basically, open it up, inspect everything
bulging caps, burnt resistors, discolored PCB
if you comfortable doing it, touch things to see if changes the sound, maybe a solder joint or internal connection has gone bad
be careful about HV mains, obviously
doi the faders change the sound unexpectedly?
a synth is basically a creative preamp
which makes them less than straightforward to troubleshoot, compared to an amp
do you have a scope?
yeah thats going to make things difficult
try and check power supplies, make sure theyre reasonable

Posted by renesis at 02:10 | permalink | 0 comments


nik1: its a synth not an amp so the vol fader could be final stage
so it could be something wrong with one of the generation circuits
but yeah, i suspect its some sort of hum issue
lost my cooking tweezers
i need to get some tongs already
lame my tostadas are all curly now

Posted by renesis at 01:59 | permalink | 0 comments


Top | Add to Technorati Favorites


© 2007 lordpil.   XHTML 1.0! CSS! Site design by GNAA  Blog Engine by pbx | MULTI2 | ian hanschen | lolwat