lordpil's blog


wow, far

Posted by renesis at 17:23 | permalink | 0 comments


he shows up in #electronics on freenode more
maybe been 2 or 3 years
i dont hang out there tho

Posted by renesis at 16:27 | permalink | 0 comments


ive seen it for UI stuff

Posted by renesis at 15:33 | permalink | 0 comments


$8.88
includes little protoboard
xtal is for the programmer on top, the target uC is running off RC clocks (has a second one for a real time clock)
so you could but that and some darlington arrays and drive couple dozen 10mA LED from a USB supply
build it on the little protoboard it comes with

Posted by renesis at 15:23 | permalink | 0 comments


shit just works
dev environment was small drama, but ST bought atollic so now truestudio is free
its gcc/eclipse, but it sees the st-link programmers fine and ive never had debug issues
stm32 has DMA for a lot of peripherals, frees up a ton of cpu
also discovery dev boards are crazy cheap
and they include an st-link programmer, so you can be programming and prototyping off pin headers for $8
https://www.digikey.com/short/p7w402

Posted by renesis at 15:18 | permalink | 0 comments


for $1.30, you can get an soic stm32f0, 25mA per io pin, 80mA max summed io
so like, you could drive 4 LED at 10-15 mA and be within spec
has internal RC oscillators
PWM is pretty easy, so you can do dimming for all 4 leds pretty easy
and compiler is free, HAL library w/ graphical setup for pins and peripherals is free
yeah you can program it however, can just stomp it out with a bunch of delays
avr is super cheap, too, internal RC and okay current sinking
yes/yes/yes

Posted by renesis at 15:13 | permalink | 0 comments


making a beacon sim, huh?
you basically run pwm to RGB LEDs, and make up patterns, on a uC
shrug, i think uC would be easiest, but you could just slow clock an eeprom or something

Posted by renesis at 13:37 | permalink | 0 comments


woofers

Posted by renesis at 11:42 | permalink | 0 comments


this has the resonance as 851hz, not 850, maybe has to do with why its off a bit
do an ac analysis
np

Posted by renesis at 11:31 | permalink | 0 comments


21:04:37 <@renesis> jero32: it settles at like 1.8vpk at maybe 150ms, 10pvk input
lesson learned, use as small a timestep as you can afford to wait for
hit go and have some coffee
judging from the ac response, seeing how accurate LTspice can model this crazy downspike in impedance

Posted by renesis at 11:26 | permalink | 0 comments


settles at around 200ms on mine
i was getting under half volt peak at 20us

Posted by renesis at 11:21 | permalink | 0 comments


was same for 10ns, 10us, running 1ns, 10ps was stupid slow
and ive added startup, since you did, maybe making it slower
okay at 10ms i get the curve youre seeing
the bounce goes away at around 30us

Posted by renesis at 11:16 | permalink | 0 comments


yeah
i think parasitic properties you put are just for the generator?
its prob timestep sensitive
when i left it blank i get a waveform like yours

Posted by renesis at 11:10 | permalink | 0 comments


jero32: it settles at like 1.8vpk at maybe 150ms, 10pvk input
link your spice image again

Posted by renesis at 11:04 | permalink | 0 comments


how far out did you simulate it?
at 100ms its like 2v peak
ti-89 on my phone is pimp

Posted by renesis at 10:59 | permalink | 0 comments


long as they work
hopefully can start packing moving truck by tomorrow

Posted by renesis at 10:36 | permalink | 0 comments


ive been doing outlines as a seperate gerber, prob same .005" width

Posted by renesis at 10:24 | permalink | 0 comments


yeah happens a lot
ccfl_man getting all political

Posted by renesis at 10:19 | permalink | 0 comments


ccfl_man: in the gerbers, the way the silkscreen is right against the pads and cutoff like that, some machine would just splat silkscreen onto the pads
sometimes they crop it out so it doesnt try and draw those
jero32: haha
at least you didnt put the mirror in space like that
at 100ms it hadnt settled
ccfl_man: neat

Posted by renesis at 10:12 | permalink | 0 comments


then do you get the right answer?
right click inductor
it says it defaults to 1mR
and youre calcing in series with resistance of 5mR
and youre only off by like 5%
are you sure?
shit on mine hasnt settled by 5ms
i think i timestepped to fine for it to go out that far, its chugging through 100ms

Posted by renesis at 09:36 | permalink | 0 comments


that doesnt make any sense jero
look at the equation, where does the negative happen?
i got 9.955 V
why is it negative
'Another choice is to define capacitive reactance as a positive number'

Posted by renesis at 09:28 | permalink | 0 comments


why is your schematic so funky
what is your output

Posted by renesis at 09:22 | permalink | 0 comments


thats overriding parameters for an NPN named NPN
prob same for caps and inductors
and yeah i wouldnt do that circuit like that irl
are you fucking up rad/s and hz?
Xc = 1/(2piFC), Xl = 2piFL

Posted by renesis at 09:17 | permalink | 0 comments


anyway for semiuconductors class, once i started overriding default values for beta, transconductance, early effect, the sim was crazy close to my calcs
few times it was dead on
so just override the default model
https://i.imgur.com/v59h70R.png

Posted by renesis at 09:12 | permalink | 0 comments


doesnt account for phase, but yeah
and ltspice uses pretty reasonable default values
but like, if you want it to match real life, you maybe have to override stuff like transistor beta, transconductance, inductor q, cap esr
not sure what you mean by ideal exactly
it models some non-ideal elements by default
but not all, like i think the small signal transistor model zeros out the early effect
anyway, you can override whatever it has in there for non-ideal elements
im not sure passives are non-ideal, like i think inductor and cap parasitics arent modeled by default

Posted by renesis at 09:07 | permalink | 0 comments


silkscreen on top of pads, did they just not draw those parts?

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


ccfl_man: which boards!?
hi!
nice

Posted by renesis at 06:05 | permalink | 0 comments


net|: 5v is higher current
higher current means heavier wire and bigger connectors
low voltage makes more sense for appliances in a DC and battery based building
but its spendy

Posted by renesis at 03:45 | 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