so theres a 5v rail involved?
weird
oh but its a 700mv signal?
no its not supposed to be, it rides on the rail voltage
Posted by renesis at 19:20 | permalink | 0 comments
lordpil's blog
yes, you are the first schematic
pretty good guess, hehe
its odd you dont see it on the ground side, shrug
yeah looks good
def worth trying
Posted by renesis at 19:15 | permalink | 0 comments
so i would guess after like 50 positive pulses youd have noticable drift down if the circuit was like in your diagrams
right
but prob an opamp >50mhz work fine
maybe, wouldnt be a cheap amp
you can make it differential
its an opamp
you get error if youre too close to the bandwidth limit and youre not doing much gain
well i can give you the ti app note to calc error vs bandwidth and gain but i dont think youll read it
http://www.ti.com/lit/an/slyt181/slyt181.pdf
its only 4 pages
Posted by renesis at 18:32 | permalink | 0 comments
well, yeah maybe because the resistors are big and the cap is big enough to keep the voltage mostly above 0v
anyway, signal average will eventualy drift to 0v, so half of it below 0v
doesnt seem to happen in your scope image
timecop: are the caps huge?
thats semi huge
so 50R
1/2pi50(.1e-6)
like a 30khz HPF
Posted by renesis at 18:26 | permalink | 0 comments
yeah thats not yours either
maybe theyre really big caps
grrr labview
timecop: http://i.imgur.com/ncR8x3z.png
see how in that one, the middle section is completely isolated by the C_aux caps
and that both signals are tied to ground through resistors
Posted by renesis at 18:20 | permalink | 0 comments
so if youre not using negative rails and a higher voltage, youre going to need to use their circuit or something similar anyway
heh
(yes)
well, its odd because your shit has dc offset on the scope but your hardware diagram says it shouldnt
seems like itll work more often than just clipping the shit out of the signal and trying to amplify that
does yours have to work for both?
Posted by renesis at 18:15 | permalink | 0 comments
13:07:16 < clannad> would it be recommended to start with basic electronics before doing microcontroller programming
i like this guy
timecop: copy their input circuit
well the shit is riding on rail voltage
Posted by renesis at 18:10 | permalink | 0 comments
$1000, $10000
student shit?
right so theyre trying to undercut solidworks
and you can route pcb in it?
oh
yeah thats what i meant
no i meant if they just put a DRC check in solidworks
and netlist shit
basically if they just made solidworks into EDA
think it just does wire routing
Posted by renesis at 17:58 | permalink | 0 comments
the interfaces, because usually you have to manage drawing traces and selection with filters and multiple grids and always so many layers and ways to visualize
so every program is totally diff
well
im not sure thats true in the machine world
and i dont know if its a good thing
okay
thats actually makes it true in the machine world, too, i think
because then only like, mastercam is good
heh
they should do DRC in solidworks
and just impliment layers as actual board layers
how much is inventor
Posted by renesis at 17:53 | permalink | 0 comments
doesnt matter how good theyr digikey integration and bom management is, if the cad engine sucks people wont use it
so im curious, just because most EDA UI pisses me off
i want something solidworks awesome
electronics workbench was like, almost
it was towards awesome
woddat
oh
hmm neat but yeah prob drive me nuts
yeah there, looks different
but yeah nothing in electronics cad approaches solidworks
have you ever done circuit boards?
theyre all a little on crack
Posted by renesis at 17:48 | permalink | 0 comments
yeah this was maybe 2009
$600 is a steal in that context
you dont like it, youre not out $5k
if they sell a ton, mentor has a new working business model
shrug
been there
the site and the situation
macegr: geeks will try it
cad geeks will know within an hour whether they like it or not if its decent
if its bad youll prob know in 3 minutes
Posted by renesis at 17:43 | permalink | 0 comments
Renegage System, its close enough to Renegade Hardware without being Renegade Hardware that people might actually take it as respectful
and if they dont fuck them
you brought it up
per year?
do you know how much pads costs?
i think its towards $10k a seat
possibly that was for a multiseat or network license
but i dont think so
Posted by renesis at 17:37 | permalink | 0 comments
if they do it right it obsoletes a lot of other shit for small-medium businesses
$600
like, if they make their BOM editor robust enough that companies can use it as a PLM system, theyll make bank
if its just enough to do real BOM, shit wont die
if you cant do anything but buy shit from digikey with it, wont last long
right why else would they get involved in EDA software dev
it prob links right into their search databass so the eda app is digikey
heh, databass
pfft, renesys is the obvious junglist dj name
eh?
i said junglist not kiddie
Posted by renesis at 17:32 | permalink | 0 comments
theyre removing the offset and just inputting normal diff
itll a/c couple the lines, so theyll eventually settle at the mid voltage between pulse trains
Posted by renesis at 17:01 | permalink | 0 comments
ha theyre bias circuits
timecop: thats basically what i was saying when i said do a trick with some resistors and caps
oh so that shit is ac
well why dont you just copy what theyre doing
well theyre in parallel, so like, the same thing?
Posted by renesis at 16:50 | permalink | 0 comments
1mhz signal and he wants like Av=4
is this coming from external devices over long cable runs?
250mv is pussy, bad EMI or crosstalk take your shit out
well, really bad
Posted by renesis at 16:20 | permalink | 0 comments
rrio amp, done
wtf
if theyre offset all stupid like that your options are limited
im assuming you only have lik 5x or 3.3v to work with?
5v
yeah im sure you can use all sorts of shit but if the negative is offset 3v, a lot of shits output wont be usable without some tweaking
no opamp is fine, he just needs a fast rrio
Posted by renesis at 16:15 | permalink | 0 comments
and falling edges itll pulse to GND then back to half VCC
so you can get period lengths easy by just having a comparator look for like 3/4 VCC
but yeah, you can video opamp, or do some bjt hacs
this shit got bidirectional?
oh this shit isnt ac?
you said vpp i assumed it was half negative
Posted by renesis at 16:10 | permalink | 0 comments
air isnt going to go around the big grey thing
right its going to greate turbulence in front of it and then itll all flow over the block things
black things look half as tall
i would line up all the black drives
on top
move the open frame psu to the bottome row on the left
center the grey thing in the bottom row
and move the small black things down to the right of big grey thing
timecop: you can maybe do a trick with caps and resistors if you just want to see pulses
you a/c couple it so its half VCC
and then on rising edges itll swing to VCC and back to half DCC in a pulse
Posted by renesis at 16:05 | permalink | 0 comments
1mhz square is prob equiv to like a 100mhz sine
timecop: http://www.ti.com/lsds/ti/amplifiers-linear/high-speed-amplifier-greater-than-equal-50MHz-products.page
i dont need it, youre the one worried about thermal
this is one of the few things FLIR are good for
they suck for any sort of absolute temp work
put the big grey thing in between some of the medium black things
whole point is not to test it
protection doesnt always work in a non destructive way
Posted by renesis at 15:57 | permalink | 0 comments
no like all test gear
if no one is buying used shit, new shit prob isnt doing that much better
he has a lot of tektronix, 80s leader stuff (japanese shit hes into), some older hp stuff, an audio precision and few other audio specific things
i think he sold a lot of the leader stuff
yes
get a video amp
ts912
try that shit
timecop: well how square do you need the shit?
Posted by renesis at 15:50 | permalink | 0 comments
other hacker says you cant get rif of test gear anymore
rid
sucks, because he has a lot of test gear
Posted by renesis at 15:41 | permalink | 0 comments
http://www.ebay.com/itm/HP-HP34970A-Data-Acquisition-Data-Logger-Unit-w-34908A-40-Channel-Mux-Module-/161472673244?pt=LH_DefaultDomain_0&hash=item25988575dc
theyre like $600 and most dont have any switch modules
this one just has the single ended module i dont think you can do thermocouples with those
agilent is now keysight?
oh damn
as of 5 days ago i guess not
spun them off
didnt the same thing happen with avago?
It also become part of product division of Agilent Semiconductor Group [5][6] before being spun off into a distinct legal entity by KKR and Silver Lake in 2005.
KKR = corporation stomper
Posted by renesis at 15:35 | permalink | 0 comments
or you can CA glue thermocouples with the HP multichannel datalogger (relay farm to an ADC with TC conditioning built in)
miss those things
Posted by renesis at 15:30 | permalink | 0 comments
zeeshan|2: can you steal a flir?
i would just set everything up pretty sane, enouogh that you can drive the mounted motors without actually damaging anything
then run a repetitive rapid cycle with the spindle on, let everything run a few minutes, start feeling shit with the back of your hand
open up cabs quick and touch subsystem surfaces
test for good grounding with current and vibration present at the same time you check for temp
Posted by renesis at 15:24 | permalink | 0 comments
heat sink models typically suck
its really easy to test with static loads
like, heatsink specs are good, but they usually dont come from models, they come from testing with loads
like, ive seen $10k dropped on thermal analysis of complex heatsink enclosures, and in the end you have to test as much as without because its too easy to think of variables which can throw the numbers off
but the simulation results are really pretty
Posted by renesis at 15:15 | permalink | 0 comments
zeeshan|2: i think youll be fine as long as youre not stupid
just give the drives as much space as you can in the given enclosure and make sure you have some sort of convection path for outside ambient to flow through if you dont have fans
right youre fine
most of those gap specs are prob something dumb like, our shit will work if you seal it in a box this big at 25C ambient
zeeshan|2: baffles and aiming the fans can make a huge diff
you prob dont need it unless you have some pocket that obviously wont get air
like if one of your drives takes up half the aream and the other takes up 1/8, you prob want to try and get most the air around the half area device
because its all going to try and go around the smaller devices
Posted by renesis at 15:10 | permalink | 0 comments
like, mad gap in a close box prob worse than mm gaps in a box with moving air in those gaps
oh
thats not a bad idea
k bbl
Posted by renesis at 15:05 | permalink | 0 comments
man i want coffee but coffe is maybe bad
i has to get to 9am-11am classes, than i want to do my optional shop lab from 11-12, than i has to go home and get ready for interview at 1
coffee now maybe fucks all of this up
zeeshan|2: gap has a ton to do with nearby ambient temps and airflow
Posted by renesis at 15:00 | permalink | 0 comments
blackmoon: i think older aluminum frames are brazed
you see brass looking metal around the joints
Posted by renesis at 14:37 | permalink | 0 comments
haha @ keep em all lined up and shit
im not sure if you mean before or after the blow up, tho i guess keeping them in line after the blow up is a safety feature
blackmoon: does he need the breakers or the box?
i dunno if he means throw the box or throw the breakers into the vfd cab
Posted by renesis at 14:05 | permalink | 0 comments
probably anything succeptible is covered with steel in some way, and they likely spray ghz equiv nastiness out every vent in their case
i wonder if they do EMC on shit like that
Posted by renesis at 12:36 | permalink | 0 comments
this isnt connector datasheet this is amodule datasheet?
okay makes sense
Posted by renesis at 12:12 | permalink | 0 comments
right, so its for grip or maybe there is an option with shield sleeve that wedge into that little crack
because they figured out the last pins suck for some reason
because then the pins side that would be the last pins and youre using those so that would be very bad
anyway, its an oddity so either a weird feature thats unimplemented, or its a manufacturing fix
they just have to make sure they dont come off
they dont have to make sure they work
not breaking is easy, consistent contact after n reinsertions is hard
how are they numbered in the datasheet
how is the other numbered
im not counting your pins
how are they not mentioned
the drawing isnt accurate?
so they dont label them?
Posted by renesis at 12:07 | permalink | 0 comments
they sound like big giant variable DC motors with a fuckton of rumbling
shield, thats the destination side, assuming shield is common with source
is a guess but that shit has big metal ends like for extending a shield
also note the key would on the PCB side would prevent the last pins from even fitting
no im saying if they extended that block to put more pins it wouldnt fit
the block cant fit anymore pins at the same pitch without being extended which would interfere with the pin
Posted by renesis at 12:02 | permalink | 0 comments
ha
macegr: gotta do it with to92 3904
can build it into a set of a dozen aluminum suitcases
or maybe itd be 1200 suitcases
Posted by renesis at 10:53 | permalink | 0 comments
blackmoon: heh
are they steel?
if steel, prob means dont be torching the handlebars for rainbows
Posted by renesis at 08:17 | permalink | 0 comments
so 1 - cuberoot(-1) + 1.5root(-1) is zero
yeah basically
well
i guess 1 = -1^(1/3) + -1^(2/3) looks reasonable somehow
Posted by renesis at 06:38 | permalink | 0 comments
thats the same link as before
ok
so second output should be 0
well maybe it has more than one result so if you did f[x] /. %[[2]] it will work
or something lire that
what does %% mean
okay so // isnt a comment
Posted by renesis at 06:33 | permalink | 0 comments
if you pull out another electronic device during class, you fail
you have a calculator fucking watch, you fail
right im saying it will not help you in a math class
dude instructors arent blind
so yeah, math aint easy when all you got is your head
right thats not gonna help
im going to repeat every math theorem ever?
also, i dont cheat
im a rare idiot like that
tho i seem to be a lot less clueless when i finish programs than average
Posted by renesis at 05:22 | permalink | 0 comments
sculptor: it sucks man =(
like, all the calc shit is easy, its all the algebra and trig thats killing me
yeah most the time
honestly algebra and trig are easy, most the time
which is typical B student shit
its the 10% of it that isnt obvious, that starts showing up about 2/3 the way into every asignment
its like, i have a variable in 8 diff random places in an equation and i have to get it all into one variable on the other side
in 3 minutes
weekend assignment im a math super genius
because i can spend an hour checking shit and going over shit and etc
yeah you cant do that in an exam
thats my point
if we could wolfram in class, id have an A
were not even allowed calculators in exams
we get docked points if we have to go up and ask for gemoetric formulas for area and volume
but at least he is letting us ask
irl math is not the same as math courses
wtf app
ur phone is in a box full of 40 phones in front of the class
along with your tablet and calculator
Posted by renesis at 05:17 | permalink | 0 comments
well, at least not an hour or whatever after the convo started going circles
hahaha
i would be so impressed
so youre doing 1^2 - 1 + 1
i dont understand your strange syntaxes
kk
i should be doing that
have a test in a week
man hes maybe fucked unless hes already pretty on it
Posted by renesis at 05:12 | permalink | 0 comments
pissed me off, generally end up arguing about shit were not arguing about
like, if i dont want to understand his specific thing, im an idiot, even if his specific thing isnt really what im going on about
Posted by renesis at 05:06 | permalink | 0 comments
rab: http://i.kinja-img.com/gawker-media/image/upload/s--DTrcZJ-W--/c_fit,fl_progressive,q_80,w_320/wktoidqjitxwvjfnzx60.jpg
putin looks uncomfortable, heh
timecop: now its just because im a shitop
Posted by renesis at 04:21 | 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