<whitequark>
>This is why using tantalum capacitors in a power supply or for supply decoupling is not a good idea. Even when they are connected the right way round they can fail short circuit and burn like that.
<whitequark>
but they were recommended by TI in the step-down DS
<whitequark>
TI probably didn't mean Russian ones, I guess.
<wpwrak>
i'm beginning to understand why the west was so afraid of you guys ;-)
<xiangfu>
just double checked. we HAVE TO put the lockflash to the end of jtag command series.
<xiangfu>
wpwrak, thanks. :)
<xiangfu>
wrong channel.
<qi-bot>
[commit] Werner Almesberger: m1rc3/norruption/2/collda: oops, only look for collateral damage on standby.fpg (master) http://qi-hw.com/p/wernermisc/cb024e1
<qi-bot>
[commit] Werner Almesberger: m1rc3/norruption/2/dumplock: use correct offset; fixed block number counting (master) http://qi-hw.com/p/wernermisc/a6136bd
<whitequark>
darn. that buck converter just does not work
<whitequark>
I've checked the schematic (~identical to the one in DS, and verified by common sense as well), pinout (fine), soldering (fine), shorts (fine)
<whitequark>
and I only maybe heated it for 30 sec at 330 deg. C with hot air
<whitequark>
it's tps62060
<whitequark>
its output is just zero, and that's it.
<kristianpaul>
whitequark: btw what are you building?
<whitequark>
kristianpaul: a thingy for connecting an AT91SAM9G45 ARM board to a (one particluar) notebook LCD (which one my friend wants. as long as he understands that it will probably fail miserably _and_ cost him $100 in materials, it's fine imo)
<whitequark>
it accepts a TFT input and spits out LVDS
<whitequark>
also, there is a 5V->3.3V buck converter for backlight
<wpwrak>
regarding the applications, "This page lists all the programs available in the latest OpenWRT distribution image." is this true ? or are also installable applications on the list ?
<whitequark>
kristianpaul: ok, now it _is_ soldered correctly. still the same result.
<wpwrak>
(remove solder stop or silk screen) a knife should do for scratching. but be warned - the knife will also scratch away copper if used too much.
<wpwrak>
whitequark: you have a good chance that your contacts aren't soldered. this sort of package is difficult.
<wpwrak>
whitequark: also, can you see metal on the side ? or is it just the the bottom ? in the latter case, you would have to heat the solver via the copper, which would be very difficult with your large ground plane.
<wpwrak>
(well, or use hot air. but careful with that, too.)
<whitequark>
wpwrak: I've did it with hot air. I can see the metal at the sides; actually, that's how I determined that it is actually soldered
<whitequark>
I've attached a needle to a multimeter and checked the connections
<whitequark>
for shorts too
<whitequark>
actually, I've checked that thrice.
<wpwrak>
very good. then it could only be a bad chip or a design bug
<wpwrak>
did you check the chip's orientation ?
<whitequark>
yep
<whitequark>
it's correct too, of course
<whitequark>
I've picked the datasheet, rotated the board to match the print and manually checked all the traces from the chip
<whitequark>
it is, uh, very simple. a couple of caps on input and output, a coil, a resistor feedback divider and a pull-down on EN pin
<whitequark>
that's it
<whitequark>
not a lot of places to screw up there
<wpwrak>
EN should be pulled up
<whitequark>
actually, no-- the converter should be normally of
<whitequark>
also, I've checked it with both levels, just in case
<wpwrak>
ah, okay. but you're pulling it up for testing then.
<whitequark>
sure
<wpwrak>
do you see anything on SW ? (with a scope)
<whitequark>
hm, let me check
<whitequark>
(by the way, is DSO quad worth it? I currently have an analog Soviet scope, and it's kinda big and heavy. And it has a BW of 10MHz and the memory of how-much-can-you-fit-on-that-screen)
<wpwrak>
you want a DSO ;-)
<wpwrak>
i haven't seen any inexpensive quad-channel scopes with good memory, though
<whitequark>
the soviet one has only one channel
<whitequark>
I'm worried a bit about the resolution of DSO (8 bit), through
<whitequark>
does that matter?
<wpwrak>
rigol makes decent low-cost (in the USD 500-1500 range) scopes with deep memory. only 2 channels, though.
<blogic>
hmmm
<blogic>
i am just thonking about getting one of the rigol
<blogic>
RIGOL DS1052E
<blogic>
this one specifically
<whitequark>
ahh, heard a lot about it
<blogic>
looks like there are hw ans sw hacks to make a 100mhz
<wpwrak>
8 bit usually is plenty. you can sweeze out more with averaging if you have periodic signals
<whitequark>
dso quad is only $200. that's somewhat affordable for me, especially given its physical size. not sure if other characteristics are equally good, through
<whitequark>
is 4Ks of storage per channel enough?
<wpwrak>
for what's basically a "pc scope", they skimped quite badly on memory
<whitequark>
yep
<whitequark>
I guess external memory was too expensive/hard to add
<whitequark>
(iirc cortex-m3's do not generally have memory controllers, and stm32 ones definitely do not)
<kristianpaul>
a good place in wich i can find math snipets implemented in C? any ideas?
<wpwrak>
72 MSa/s is very low. the effective analog bandwidth is thus about 20 MHz with one channel, 10 MHz with two
<whitequark>
wpwrak: still better than my soviet scope. I doubt I can find more for $200
<wpwrak>
the DS1052E sounds quite decent to me. deep memory, sufficient sample rate, all the necessary controls.
<whitequark>
kristianpaul: a clone costs better. what do you think about quality?
<kristianpaul>
dunno... i would prefer a second hand rigol anyway :)
<zrafa>
are you using finally sdl-widgets?.. I would like to know how
<kristianpaul>
also this lack ethernet wich is not aceptable those days i think :)
<kristianpaul>
zrafa: me?
<wpwrak>
kristianpaul: (clone) the comments say it's a tek clone. in this case, check the amount of memory it has :)
<whitequark>
kristianpaul: for what purpose is there ethernet on a scope?
<zrafa>
kristianpaul: no idea.. I replied the sdl-widgets mail explaining why it does not work in BEn, and I see that you have ported it
<wpwrak>
besides that, teks are nice. handle better than rigols. but at equivalent specs, they're also considerably more expensive
<kristianpaul>
zrafa: i would like to use it, but need some time to learn the basis like making a fronted for sox like you did in jlime, remenber?
<zrafa>
kristianpaul: so I was thinking that you are starting to use sdl-widgets, and I would like to know how. it just works with mouse (the widgets)
<kristianpaul>
ahh
<whitequark>
wpwrak: (SW pin) nothing
<kristianpaul>
just mouse :/
<zrafa>
kristianpaul: but.. did you understand my mail?
<zrafa>
kristianpaul: maybe I did not explain it well
<kristianpaul>
i dont
<wpwrak>
kristianpaul: usb is also okay. sometimes even preferable. depends a bit on what you do. in the lab, ether or usb are pretty much the same. main difference is that you're more likely to find a web server on ether-connected devices.
<kristianpaul>
i took it as if you recomend it but warned about screen size problems?
<kristianpaul>
whitequark: remote managed or even scripting , well some scopes support that..
<zrafa>
kristianpaul: no.. the screen size problem was my answer for some music composer program which uses mod and s3m
<wpwrak>
whitequark: not good ... resistance of SW to ground ? (when the circuit is unpowered)
<zrafa>
kristianpaul: I did another answer for sdl-widgets werner found
<kristianpaul>
"The idea of those widgets are just to be used with mouse, no keyboard."
<kristianpaul>
ouch :)
<wpwrak>
kristianpaul: time to roll up your sleeves ;-)
<zrafa>
kristianpaul: I think that I was the only fan of music in demos and intros here :) (which used mod and s3m, which I played for a while with those programs to compose music with samples)
<whitequark>
wpwrak: 2.78K
<wpwrak>
good
<wpwrak>
what is MODE set to ?
<whitequark>
what does that mean?
<zrafa>
kristianpaul: at that time I ported several "tangos" songs to mod using those programs..
<whitequark>
wpwrak: (mode) high, i.e. fixed-freq
<whitequark>
it's connected to input power rail, directly
<wpwrak>
it means "not open", "no shorted". that's all i can read from it ;-)
<zrafa>
kristianpaul: reading the music scores and then porting
<kristianpaul>
wpwrak: time to look for another inresting widget project ;)
<whitequark>
wpwrak: (sw) ah. same here. I'm sure it should not be shorted, but why not open?
<kristianpaul>
zrafa: btw tht youtube vide you passed, can actually sound same in the nanonote?
<wpwrak>
because you have R1 and R2
<zrafa>
kristianpaul: the mods using tiny sound samples.. and they are combined, so yes, they should sound okey
<wpwrak>
did you load the output ?
<zrafa>
kristianpaul: they were used by intros and demos
<whitequark>
wpwrak: I've put an 1.8k resistor there
<kristianpaul>
ah combining is the trick i see
<kristianpaul>
well, sounds pretty nice !
<zrafa>
kristianpaul: they worked in older PC with dos
<wpwrak>
whitequark: sounds good, too. time to re-check the pins, i guess. with special attention to ground and Vin
<whitequark>
wpwrak: okay
<wpwrak>
oh, and did you measure SW-to-GND when the output was loaded with 1.8 kOhm ?
<whitequark>
wpwrak: nope, the load was disconnected
<whitequark>
so that was just R1+R2
<whitequark>
okay, I've rechecked pins, ensuring that I am touching the chip pads itself and not PCB ones
<whitequark>
they're fine
<wpwrak>
when you connect power, does Vin look good ? and what current consumption do you get ?
<whitequark>
yeah, Vin is 5.06V (USB)
<whitequark>
current, hm
<whitequark>
I have had a special cable somewhere...
<whitequark>
2.5mA
<whitequark>
hmmmm
<whitequark>
and where do they go?..
<wpwrak>
the rest of the circuit ? or are you measuring just the converter ?
<whitequark>
yep
<wpwrak>
A or B ? :)
<whitequark>
oh:)
<whitequark>
that 5V input is connected just to the converter
<whitequark>
nothing more
<whitequark>
and I'm mucking with it separately
<wpwrak>
hmm. then you do indeed have a good question :)
<whitequark>
yep.
<wpwrak>
are you sure your pins are really where you think you are ? e.g., is "top view" really "top view" ?
<wpwrak>
s/think you are/think they are/
<whitequark>
yes
<whitequark>
if they're correct in DS
<whitequark>
the chip looks like this
<whitequark>
. C6Y
<whitequark>
the . is on the ground, next pin is SW, next is GND, and so on
<whitequark>
the printing on the chip is vertical when the big connector on the board is on the bottom
<whitequark>
and the dot is on the right
<whitequark>
oh.
<whitequark>
it fucking works.
<whitequark>
how the hell?!
<wpwrak>
;-))
<whitequark>
no, really.
<whitequark>
HOW?!
<wpwrak>
perhaps you convinced it that there were no excuses left for not working ;-)
<whitequark>
maybe
<wpwrak>
you could try gently pushing against components and see if it stops working
<wpwrak>
defective solder joint, cracked traces, or broken bonding wires can all cause this sort of trouble
<whitequark>
well, the voltage changes when I touch feedback resistors, but that's normal
<whitequark>
ahhh, I have two ideas
<wpwrak>
bonding wires are particularly nasty. there, the defect is INSIDE the chip
<wpwrak>
(feedback change) yeah, they're very high-Z, so your finger makes a big difference
<wpwrak>
also flux residues and such could heavily upset this circuit
<whitequark>
first. when I've changed my multimeter from voltage measurement to current measurement mode, I've replugged the probe, but it has correctly measured input voltage and checked the continuity ("beeping diode mode", not sure how is it called correctly)
<wpwrak>
e.g., in atusb, i had flux mess with my reset timing. it contributed something like 100 kOhm.
<whitequark>
second. I've used a flux cleaning thingy--yes, that flux-off
<whitequark>
it is dimethoxymethane
<whitequark>
either this, or I have not dried it completely after bathing
<wpwrak>
bathing = ultrasound bath ?
<whitequark>
yep
<wpwrak>
what liquid did you use ? tap water ?
<whitequark>
it's filtered (not sure if you have same things in your country, but we do: it's an activated-carbon-filled (at least it looks like this way when disassembled) filter, you put tap water in, you get something better (and without any chlorine) out)
<whitequark>
and I've boiled it in a kettle
<whitequark>
poor man's distillation
<wpwrak>
phew. so much work ;-)
<whitequark>
yeah
<whitequark>
but now I'm a bit wiser
<wpwrak>
you could get demineralized water from car supplies
<whitequark>
yeah, I know. but I constantly forget about that. also, I'd need a whole lot of it
<whitequark>
*forget when I'm near the shop
<wpwrak>
(for cooling. without additives. hmm, if they have that in russia :)
<whitequark>
don't know about car supplies, but I've seen that kind of water in electronics shop
<wpwrak>
i have a DIY shop a few blocks away that has this sort of water in 5 l canisters. let's see how much it costs ...
<wpwrak>
then click on "Capital Federal y Bs. As." on the right
<whitequark>
I want to live in your country, whatever it is. DIY shops with websites, wow.
<wpwrak>
(in the "Clientes NO registrados" box)
<wpwrak>
;-)))
<wpwrak>
zrafa: do you hear ? the recruiting of new immigrants is progressing well :)
<whitequark>
$10 for 5l. that's nice
<wpwrak>
$ = pesos. oh, sorry ... it's less than USD 1.3 per liter. only about ... USD 0.50
<whitequark>
so, $5. even better
<wpwrak>
grmbl. x server hang :-(
<qi-bot>
[commit] Werner Almesberger: m1rc3/norruption/2/dumpotp: quick and dirty script to dump the OTP memory (master) http://qi-hw.com/p/wernermisc/97af8ea
<wpwrak>
whitequark: yeah, about 51 US-cents per liter of distilled water. affordable :)
<whitequark>
hah. I'm doing the final cleaning, as I've not bothered to clean it each time I've moved the chip 0.1mm to some random direction
<whitequark>
wonder if will work afterwards
<whitequark>
hmmmm. is that water in ultrasonic bath boiling?
<whitequark>
or is that just the fog?
<whitequark>
I couldn't tell easily. It's hot.
<wpwrak>
;-)))))
<wpwrak>
it's supposed to be room temperature
<whitequark>
not for this particular cleaner, I guess
<wpwrak>
that is, unless your have an ultrasound cleaner with heating. i guess there's no harm in exposing your circuit to 100 C, though
<whitequark>
it has, ahem
<whitequark>
it does not supposed to have a heater.
<whitequark>
*is not
<whitequark>
but it has a very chinese power supply
<whitequark>
(yes, that's an adjective)
<wpwrak>
one of those days, one of your "special" devices will kill you ...
<whitequark>
that's not very unlikely, yes
<whitequark>
(regarding that. I have -- yes, have -- a friend who once had a Soviet scope with a failed power plug. yes, the 220V one. when it did not worked, he pushed the exposed metal bars into the socket with his bare hands)
<whitequark>
somehow (surprisingly) he is still alive
<whitequark>
and does all sorts of weird stuff from refrigerators to solid-body lasers
<DocScrutinizer>
wpwrak: distilled water for cooling?? WTF? highly deprecated, will etch away your car's cooler
<wpwrak>
well, some people worry about how they'll live when they're too old to work. at least he seems to be free of such concerns :)
<whitequark>
exactly
<wpwrak>
DocScrutinizer: maybe it's for something else. but in any case, they sell it in the cars stuff section
<whitequark>
DocScrutinizer: etches? really? it's distilled water. isn't it supposed to not contain anything except H2O?
<DocScrutinizer>
yep, for lead-acid bateries
<wpwrak>
could be. next to it, they have water "with additives". no idea what these do, though.
<DocScrutinizer>
it's diluting any chalk that's already coating the cooler insdie, and it also dilutes the metal partially
<DocScrutinizer>
I've been puzzled as well that it's *that* bad, but I heard it several times: do NOT use distilled water for cooler refill
<whitequark>
wpwrak: :)) yeah, it's a well-known joke
<DocScrutinizer>
whitequark: cleaning with distilled or even clean tap water is OK, but you should finally rinse the board with 99% alcohol which is hygroscopic and will remove water from locations you can't reach with pressurized air etc
<whitequark>
DocScrutinizer: thanks, I'll do
<DocScrutinizer>
though virtually all of the residual substances fund after manufacturing/repair/rework are not water soluble, so water is useful only to clean board that suffered from spilled coke, coffee etc
<DocScrutinizer>
s/fund/found/
<DocScrutinizer>
as e.g sugar is hardly soluble in alcohol and non-polar solvents
<DocScrutinizer>
afaik
<whitequark>
yeah, but I'm not sure if ultrasonic bathing of the board with 99% alcohol is a good idea
<DocScrutinizer>
well, I'm not sure if ultrasonic bath is a good idea at all
<whitequark>
flux-off (dimethoxymethane) does something with the flux so it almost completely goes away, and bathing gets rid of what remains after flux-off
<whitequark>
as you can see from the image, this produces quite good results
<DocScrutinizer>
it's removing things by cavitation erosion (building up microscopic "vacuum" bubbles on the surface, that - when collapsing - really do hard impact on the surface)
<whitequark>
yeah, I know
<DocScrutinizer>
might even remove thin copper coatings etc
<zrafa>
wpwrak: recruiting : ;-)) .. and you just promoted some DIY shops.. :)
<DocScrutinizer>
?
<DocScrutinizer>
I just applied to a really blaergh job, so who's recruiting and do they pay good?
<wpwrak>
water also removes water-soluble flux. and yes, you can use pure alcohol i an ultrasound bath.
<wpwrak>
the main issue with ultrasound is resonance. the better devices (for cleaning electronics) therefore vary their frequency.
<whitequark>
wpwrak: I definitely can, but it is _not_ easy to get large quantities of pure alcohol here, guess why.
<whitequark>
I have a 5l canister, and it is, well, valuable :D
<DocScrutinizer>
use one of the other alcohols then, methanol, isopropanol
<DocScrutinizer>
doesn't really mater as long as it's pure (not that 70% stuff)
<whitequark>
I'm afraid of methanol (as of any other quite poisonous easily evaporated substance), through isopropanol may be a way to go
<qwebirc56550>
Does anyone know about Nanonote Manufacturing, who should i contact ?
<wpwrak>
whitequark: yeah, i was thinking of isopropanol. not your 5 l canister of premium vodka ;-)
<DocScrutinizer>
methanol is relatively harmless, it's the formaldehyde that your liver produces from methanol that's the problem (that's why you get huge amounts of ethanol in hospital when they diagnose and tread a methanol intoxication. Same liver enzymes that methabolize ethanol to acetaldehyde also are producing formaldehyde from methanol, so diluting methanol with ethanol helps to reduce the formaldehyde level in blood)
<wpwrak>
qwebirc56550: i guess that depends a bit on what exactly you're after. but probably wolfgang
<DocScrutinizer>
s/tread/treat/
<qwebirc56550>
Yes a saw his name, but i''m new here, how can I contact him ?
<wpwrak>
DocScrutinizer: so to get your doctors to give you a prescription for lots of booze, you need to tell them you had consumed some methanol ? interesting
<whitequark>
DocScrutinizer: yeah, I know that. this does not changes the fact that it is dangerous to work with
<DocScrutinizer>
that'S what worked several times already, yes
<wpwrak>
qwebirc56550: either lurk here or send mail: Wolfgang Spraul <wolfgang@sharism.cc>
<whitequark>
wpwrak: that's not vodka, just some pure alcohol :D it has happened to get to me untouched somehow. I don't quite recall how I got it
<wpwrak>
now we know DocScrutinizer's secret :)
<qwebirc56550>
thanks a lot !
<wpwrak>
whitequark: temporary memory loss is often associated with in vivo quality control of ethanol :)
<whitequark>
wpwrak: :) somehow the things has arranged in an interesting way: I haven't ever drinked anything alcohol-containing
<whitequark>
i.e. in my entire life
<wpwrak>
or at least you don't remember such events ? :)
<whitequark>
:)
<whitequark>
nor does anyone else
<wpwrak>
(lack of memory) at least that's how you should express it to avoid tarnishing russia's hard-earned reputation
<wpwrak>
(anyone else) hmm, social drinking would explain this :)
<whitequark>
I don't really think it is a good reputation, especially given that it is very close to truth
<wpwrak>
also the voltage regulators look a little tipsy
<DocScrutinizer>
LOL
<DocScrutinizer>
AT98C2051 \o/ my private pet
<DocScrutinizer>
awesome lil critter
<whitequark>
huh? have you developed it? :)
<DocScrutinizer>
89 of course, not 98
<whitequark>
not that I'd be surprised if you did
<DocScrutinizer>
not exactly, just had an own design with it (pusteflipper, easy to google) - the PCB for that one looked nicer than the one linked above though: http://maemo.cloud-7.de/dscf33*
<whitequark>
DocScrutinizer: that URL does not work
<whitequark>
googled it (the device). what the hell it does?!
<DocScrutinizer>
the controller operates some 5 LED/phototransistor light barriers, and flash some LEDs and move some magnets
<whitequark>
2000
<whitequark>
it's quite ancient
<DocScrutinizer>
yup
<DocScrutinizer>
back when I had a real "company name" jrwsoft
<DocScrutinizer>
running a real company is a PITA though, so I stopped that ;-D
<DocScrutinizer>
btw the whole device is a medical aid for kids with problems to speak correctly - you blow a small plaic ball with a straw (or without) and manage it into some target holes
<whitequark>
weird
<DocScrutinizer>
kinda micro table air minigolf
<DocScrutinizer>
plastic*
<DocScrutinizer>
the controller gives a kind of scoring and feedback
<DocScrutinizer>
btw "flipper" is the german name for pinball
<whitequark>
and "puste" is...?
<DocScrutinizer>
blow
<DocScrutinizer>
so it's blow-pinball
<DocScrutinizer>
which pretty exactly describes the whole thing