lordpil's blog
hey those hinges are pretty precision
and the way they mold them in is pretty cool
you should try and find some american company imports cheap ones
like i bet you can get them for $4 and the QA will be half as shitty
or maybe canadian
yeah because then you can return them and bitch to a real person
yeah fuck that
theres a machine does this shit by itself
you can prob find some $20 hipster shit
some wayfarer looking shit for prescriptions
haha
Posted by renesis at 19:03 | permalink | 0 comments
like you would never know if you never triggered it
yeah that last one
that last one will fuck you
i mean you never know, its like ESD
sometimes you decide you figured it out and later you figure out it was something else stupid
but yeah youre like, doing vibration testing on random electronics
blackmoon: subwoofer, but yeah if you play a kickdrum through it, its basically a kickdrum
when i was a kid my dad would buy tons for his pharmacy, i think just to have a steady supply
maybe last hime a few weeks
the frames?
they have been making that shit for like 30 years
yeah they have regressed they didnt do that in the 80s
i am dissapoint.
Posted by renesis at 18:58 | permalink | 0 comments
yeah theyre smaller capacity and who knows whats up with the protection components
capacity goes up, protection stuff prob has to get bit beefier
its walmart, shit prob has no headroom at room temp already
go into protection at like 2C, or never
apestate: do you know what MLCC are like inside?
the ceramic chip caps
its like layers of foil bonded to the little metal ends
anyway you can rip off the foil from the terminal from flexing
yeah
but then when its unflexed, the foil touches again and shit works
like shit will get hot or flex and fuck up and you will go nuts because maybe it almost always works
like you can fuckup a lot of that protection circuitry and prob shit still works fine until something bad actually happens
Posted by renesis at 18:53 | permalink | 0 comments
but right your idea is decent, you can prob find hardware
because youre biking a lot?
and you dont have spring fork
but right big shock loads that twist and bounce shit can fuck up pcb
kickdrums
kickdrums are so bad for pcb
and big transformers right in the middle of big pcb
they trampoline all the components to death
iunno
those things are supposed to splode if you look at them wrong
but ive been using 18650 in my ecigs like 3" from my face for years
few years unregulated
with the protection caps?
or just the safer chemistry
weird
Posted by renesis at 18:48 | permalink | 0 comments
or epoxy
you want something least a little flexible, and it cant be conductive
like, they will have copper doped shit for heat conducting at parts store, obviously dont get that
but i think all the other rtv for cars is okay
unless they make an especially pussy one
but you cant like, completely cover power components
they will overjeat
but some around the base of inductors is normal
bottom of the inductors just to keep them from moving
they can break their solder joints and the traces next to the pads
the toroids bounce
theyre decent transducers, heh
mostly you want to reduce board flex
but making sure connectors and inductors dont move arounds actually helps a bit with that
have you seen pc hard drive mounts?
with the shoulder screws and the gommet
and the like slide into other hardware
i mean theres variations
Posted by renesis at 18:43 | permalink | 0 comments
like, sometimes you can just blow hot air on shit
but its like of like hard drive or brake problems, usually doesnt get better
soldering iron?
so the shit is you cant test it by just pressing on parts
all sorts of high frequency shit and feedback mess
like, if it worked fine it would prob act diff pressing random shit
but you can like, kind of twist and bend it from the edges
oh its death
speakers and a lot of mass production accessory stuff will just have hot glue on everything
its ugly, and a lot of higher end shit wont have it, but it works
like, all my prototype shit, as soon as i build and test a section i pot it in hot glue
i dunno man, subwoofer is prob worse
ill just put a huge bead of hot glow over
or sometimes i will kind of make a mold with masking tape walls or something
and just put like 1/4" in
everything that doesnt get hot, just cover it, keep the tip in to keep it flowing until it surrounds parts
well
no on your thing maybe not okay because psu
could overheat stuff
also maybe melt into a mess
but you can get like, high temp RTV
Posted by renesis at 18:38 | permalink | 0 comments
u3 prob a uC for the leds, sitting on top of the input protection shit
you would think there would just a a chip, 'protects all the things'
every smps is a mess like this
its a totally normal product, the parts dont look super super cheap
also kind of normal
yeah they have lots of discrete circuits
so if you pop a SMD passive solder joint, or break internal bonding (happens on caps a lot), from board flex, you will go nuts troubleshooting
because itll work sometimes
like you will unscrew the board and unload it and the broken solder joint or track or end terminal will contact again
well
Posted by renesis at 18:33 | permalink | 0 comments
Q7 is maybe a pfet
Q9 to switch the pfet
is J? on the left input?
opposite the USB
top view, big component side
yeah so i think Q9 and Q7 are the shutdown pfet circuit, and all the shit underneath prob protection
Posted by renesis at 18:28 | permalink | 0 comments
like another couple of smps worth of shit
or ton of feedback stuff for protection
haha
so i think the big one is a 20V boost
and then everything follows
blackmoon: you think its all one stages?
you are powering your cpap on the run?
cpap ups?
where does 16.8 come from?
that some regged shit from the wall?
its external shit?
Posted by renesis at 18:23 | permalink | 0 comments
feed forward with giant darlington regs would be neat
some to-3
just linear vregs with whatever vrefs
see yeah its big
they could jack the speed way up and use tiny tiny inductors instead
i bet those quad flat packs are some sort of USB supply
and then it looks like a big converter
oh haha at the bottom
Posted by renesis at 18:18 | permalink | 0 comments
right its prob at like 75% max current
its pretty much how fast the drive the fets, how fast the fets are, and what frequency vs cost/weight/audibility tradeoff they made
yeah systems with multiple stages prob like 75%
most class-d amps are like 75% because 85% supply 85% amps
unregulated feed forward with fet H bridge rectifier is pretty much a linear PSU
its just switching faster to use a smaller transformer
and barelt burning energy rectifying
its an SMPS but barely
Posted by renesis at 18:13 | permalink | 0 comments
like, youre burning more W charging faster, but the WH burnt in the battery might end up being close
i dont think its exactly the same tho
i think temp changes enough to swing efficiency one way or another, i think its bad
shrug
i dont know enough about specific charger chips to fuck with them too much
i know every variety i can think of exists, as far as measurement and management
5% is really good
if you can hear the switchers and it feels kind of heavy, its prob more efficient
then you want something heavier with audible switching
yeah
with cheap shit, switching low, its maybe pretty efficient
it depends on how theyre driving the fets
like, if its some ancient pwm IC with BJT switches its not going to be more than 80% probably
but like, modern shit is 90%+
Posted by renesis at 18:08 | permalink | 0 comments
\o/
like unbelieve good or bad?
92% is pretty good
oh im not sure what you mean
but its because of switching frequency vs size of magnetic components vs operating voltage
you can switch lower for more efficiency
or lower voltages
but the size of the magnetics is bigger, so heavier bigger and more $
at higher frequency, the mosfets spend more time slewing vs at stead state on or off
so you are less efficient, but you can use much smaller magnetics
plus with power banks there is battery losses
so i think with faster charging you lose more power
but i might be wrong
over time it might even out
Posted by renesis at 18:03 | permalink | 0 comments
they also got faired in rear wheels
cheaters
Posted by renesis at 11:17 | permalink | 0 comments
and then people die
and race fans get sad and angry
and france takes people to court
and it would just get banned
they run bigger diffusers now, so there is some ground effect, but they run flat floors and rather large rake (rear end lifted vs front) so its not a full floor ground effect
https://www.grandprixradio.nl/wp-content/uploads/2018/02/1920x1080-Alfa-Romeo-Sauber-C37-2018-09.jpg
see they got little tunnels in the back but they only go up a couple feet, and the floor few inches off the ground
http://static2.businessinsider.com/image/55549f12eab8ea1717d7b6e8/indycar-driver-helio-castroneves-went-airborne-in-a-scary-crash-during-indy-500-practice.jpg
not sure whats going on there, but indy has full length ground effects, which you can clearly see here, and might be an indication of what happens when they fail
Posted by renesis at 11:12 | permalink | 0 comments
yeah its goofy but its still not easy to pass
like, the DRS window is half a second i think
you have to put right on another cars ass to enable DRS
and if you had to get stupid in a turn to be in that position, you exit speed will be too shit for it to make enough of a diff
in practice its not as bad as it sounds
like, the problem is making the cars faster usually makes them harder to pass
gound effects, especially excreme ground effects with skirts, actually solve most of the issue
ground effects just use the air on the ground and dont care that much about dirty air, and dont leave the air that dirty compared to the front wings shaping vortexes to funnel air into the rear wing
but if you hit a curb, and in racing you have to, and it bounce the car up, you lose all downforce from ground effect, you break the low pressure effect from the expanding area chamber
Posted by renesis at 11:07 | permalink | 0 comments
they were burning oil
i think mercedes
they figured out how to use the oil to get some more power while keeping the fuel limit a little lower
like, this is a 0.5% sport
if you arent under 107% of the fastest time, you dont qualify
the fastest three are often within a few tens if that
they are like shaping vortexes to create fences to move air to create other vortexes
which is a bitch on a open wheel car, compared to endurance prototypes
like, you trade weight for an aero mess
no thats the #1 issue in the sport for like 10+ yearts
dirty air
aero doesnt work following, so passing is a bitch
to the point where theyve made an active suspension system that only enables on straights is you are closely following another car for position
i think they note it but its just a result of how they tune the cars
Posted by renesis at 11:02 | permalink | 0 comments
no idea, the cars are basically robots, could be doing anything you can think of
looks like around 50 psi
https://www.reddit.com/r/formula1/comments/2zds92/boost_pressure_levels/
its variable tho
and the flow limit isnt a plate
its a sensor
the teams control the engines with flow directly, direct injection dealies
ferrari figured out how to like, preload their system
stretchy tube or something
so they got around the limit for periods while staying under it on average, as seen on the monitored sensor
yeah basically they made a fuel capacitor
there was probably a second pump
thats basically all of f1
its hard to explain that to people
half of it happens in between races
and when someone figures out how another team somewhat cheated
then FIA has to decide if they cheater, and prob change the rule
its engineering to a moving spec
Posted by renesis at 10:57 | permalink | 0 comments
in that case it matters more because power output will be less overall
like, going back to 70s small car power levels is realistic
so like, 60-120 hp
yugos were supposedly great off road
so light and spingy they just bounce over everything
turbos are indisputably a huge efficiency win
https://www.motorauthority.com/news/1112999_mercedes-amg-f1-engine-achieves-50-percent-thermal-efficiency
right thats what the highest level of performance cars are doing
mclaren, ferrari and porsche all have hybrid turbos, specifically to deal with lag and torque non linearity
heh
they put a fuel flow limit
and V6 hybrid turbo is mandated
they would have done i4, like the 80s, higher efficiency
but ferrari threatened to quit
no i dont think so
macgyver0: omg i saw it omfg
at monterey historics
Posted by renesis at 10:52 | permalink | 0 comments
lights, a/c pulleys, all electronics, tons of bigass car sized actuators, even without stereos
no idea
but turbos, motor/generators, and batteries gonna become one system
car shit is just made to be cheap and last long in crazy conditions
anyway, i think the big push is going to be lightweight safe cars
and hybrid turbos
because the solutions are proven, cant not work
Posted by renesis at 10:47 | permalink | 0 comments
what you get out?
macgyver0: are they clutched?
CVT on an alt seems a little complex for car guys
but a/c clutch on an alt sounds exactly like car guys
well thats the thing, if you can turn the alt off instead of dumping its power somewhere when the battery is topped off, you maybe gain tenth of a mpg
is there a shunt or something to reg max batt voltage?
okay so why couldnt an ecu just control the reg
heh
wot cutout sounds neat
yeah but then you cant do cool shit
like turn the alt off at wot
no way
right
so how much a car load?
all the bullshit is prob couple kw easy
Posted by renesis at 10:42 | permalink | 0 comments
car ECU are so fucking slow, prob all the filtering they have to put in
modern cars are basically 3 levels deep embedded systems
i imagine the hackers and the mech engineers at car companies go at it a lot
BITCH YOU CAN JUST GROUND MY SENSOR ON THE MIRROR MOTOR BRACKET
*cant
macgyver0: right because the fucking things are basically dataloggers now, no?
ya
everything is feedback tables
blackmoon: variable pulley or something wouldnt be nuts
or a clutch
averages, yo
theyre like pushing the edge of the envelope for efficiency for a decade now
like, now they basically have to do everything
every econobox prob gonna be hybrid turbo 10 years from now
just disconnected i think
so prob not what you mean
right that thing not gonna spin free wired in
Posted by renesis at 10:37 | permalink | 0 comments
i dont trust protection circuits to last more than a few events
car is bunch of random parasitics, battery might be far
car companies would figure out how to make a 10ft distance into 50ft of wiring in a harness
plus shittiest grounds in embedded, ever
unibody its prob depending on a couple mexican welds
right im saying its prob just hardware issues on most cars
trucks prob still have silly current paths
i wonder how much modern cars are using chassis for return on shit
like the led lamps and sensors
Posted by renesis at 10:32 | permalink | 0 comments
glass ones have the noise
but so much fun burning string of 1n914 with dead batt
i think this is same issue i was talking about
there is actual inductance and the apparent inductanct from switching the coils around
right i think its the switching that produces most of the potential
right
so you have this mad variable switch action
i guess in this case there is a rectifier, no?
right that
so instead of normal inductive rise and pushback, you get these bigass spikes from reversing current flow
Posted by renesis at 10:27 | permalink | 0 comments
tinsel will do enough current for like 4000W across 4R
tinsel leads are legit the hardest part of high power woofer design
sometimes they are woven into the spider
occassionally there are flat ribbon lead setups
but generally, its about how you position them, and how to do strain relief at the ends or with a sheath or coating, that makes the diff
like, heatshrinking sections or entire lengths is an option
RTV or glue at the ends
worst case, they will jump rope and if youre not careful, short against each other
like you basically have to grab them and try to touch them together to make sure they can, just checking interference with loose tinself isnt enough
room temp vulcanizing
Posted by renesis at 09:58 | permalink | 0 comments
pretty sure that specs a temp rise and ambient or something
Posted by renesis at 09:50 | permalink | 0 comments
yeah if starburst could actually go stale, they would basically be now and laters
i got the sour ones, i got exactly 6 apple, 6 cherry and 6 watermelon
alternatively, a now and later is a soft opaque jolly rancher
macgyver0: haha, so now and laters are apparently just starburst now
like, theyre not breaking and sticking to my teeth, maybe im supposed to let them sit for a week
Posted by renesis at 09:29 | permalink | 0 comments
you can just Xc = 1/2piFC and Xl = 2piFL all the components and solve for given frequency
im sure you could automate that in c or python to make accurate ac response for all sorts of shit
im saying without just cheating
because we both know virtually all passive circuits that make practical sense will simulate in a few seconds
onlt thing that spice seems to flip out on is certain positive feedback situation, and im pretty sure because they divide by zero and break ltspices fragile universe
TI has a doc about opamp feedback oscillator where they talk about it
WTF
HOW LONG HAVE NOW AND LATERS BEEN SORTED INTO BAGGIES OF INDIVIDUAL FLAVORS!?!?
dude i dont know whats up with that
but its not as bad as labview
flavored hard plastic chewies
Posted by renesis at 09:24 | permalink | 0 comments
rab: ac motors are weird
the flipping inductive change shit doesnt not work for me intuitively
like, i think the switching effect makes the inductance looks larger than it actually is?
like, there is more than one thing limiting the current
like, i think its inductive but not in typical RLC sense
jero32: no
not at all
RF is weird because its 3 dimensions and a lot of the math turns inside out
but even when just view as very big light, it makes intuitive sense
ac circuits obey ohms law
Posted by renesis at 09:18 | permalink | 0 comments
omg parts at ups store bbl
Posted by renesis at 06:57 | permalink | 0 comments
apestate: pliers
or a hammer
sometimes its not actually a bead, its halves and the shit falls off once you get the clip or case or whatever off
Posted by renesis at 06:29 | permalink | 0 comments
i wish i had chisels
i dont think i have
Posted by renesis at 04:40 | permalink | 0 comments
maybe there is a usb chip you dont have to buy 10,000 of and you can do a psu yourself
bc #electronics
they like $30?
yeah its their cheapie pro shit
they got $10 cortex-m stuff too
man so this LCD i got for this perf prototype is 2mm pitch
which im excited about in a future layout sense but im like, damn in a proto perf context
like i am going to have to slice through a row holes or something
Posted by renesis at 04:35 | permalink | 0 comments
last phone was a usb-c charger, pretty sure i charged other things with the usb-c to micro-usb cable
this phone is a usb-a charger but the phone is usb-c and it charges crazy fast
also i think usb is going to be fine with a few amps and most laptops power consumption going to go down
theyre supposed to
if it doesnt it means they could have made more money and someone gonna get canned
Posted by renesis at 04:30 | permalink | 0 comments
would be a little inverter, too
im pretty sure usb-c is caped at low current and voltage until negotiated to crazy big voltages and current for a tiny low pitch connector
yeah its nuts
like i havent researched it much because it seems insane and maybe not so implimented yet
into a c jack?
or sourcing?
itll push at low current, 5v without anything special
Posted by renesis at 04:25 | 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
