<DocScrutinizer>
they suggest a 4.7uF on VDD, you got 2uF
<DocScrutinizer>
actually 2.1
<DocScrutinizer>
what's with AT86 [12] and [13] DVDD?
<wpwrak>
yup :) they put all sorts of values there. e.g., page 32.
<wpwrak>
the at86rf230 has internal 1.8 V regulators for DVDD and AVDD
<wpwrak>
C11 and C12 are their bypass caps
<DocScrutinizer>
MHM, via DEVDD
<wpwrak>
yup
<DocScrutinizer>
and EVDD. I see
<DocScrutinizer>
(should've checked the datasheet of AT86...)
<DocScrutinizer>
wpwrak: check fig1.6 - you could hook up a storage scopte to nRST and see if it caches any 0 level. AIUI nRST also is an output. You could distinguish between VDD created resets and clock-failure related ones
<DocScrutinizer>
wpwrak: also what's the state you set MCD to? (missing clock detector enable)
<wpwrak>
(MCD) lemme see ...
<wpwrak>
i disbale everything but power-on/VDD monitor
<DocScrutinizer>
VDD is allowed to be between 3.0V and 3.6V
<wpwrak>
(that is, if i make it into the boot loader)
<DocScrutinizer>
(<wpwrak> i don't like the 5 V on the VDD rail at all.) :-o
<wpwrak>
i wonder if i have some weird solder bridge somewhere that i keep on overlooking
<wpwrak>
the layout is quite dense. i've gradually relaxed it a bit since. maybe i should just make a new board and see if things get better.
<wpwrak>
i wonder if the changed thermal stress pattern has done something
<DocScrutinizer>
I still wonder if you meant what you wrote
<wpwrak>
(vdd - regin/vbus short) that's of course what i tested first :)
<DocScrutinizer>
VDD is allowed to be between 3.0V and 3.6V
<wpwrak>
(what i meant) why ?
<DocScrutinizer>
(<wpwrak> i don't like the 5 V on the VDD rail at all.) :-o
<wpwrak>
i have two boards. board #0 has 3.577 V on VDD. on board #1, VDD is roughly at VBUS (~5 V)
<DocScrutinizer>
now that's a massive defect
<wpwrak>
board #0 doesn't even flash, so i've written that one off already
<wpwrak>
board #1 developed the 5 V issue recently. when i first experimented with it, it didn't have it - just the spontaneous resets.
<DocScrutinizer>
The voltage regulator is enabled on reset. When the device is self-powered from a 3V supply net, the reg-
<DocScrutinizer>
ulator may be disabled in order to save power. Important Note: If the voltage at the regulator input
<DocScrutinizer>
(REGIN) is greater than the Core Supply Voltage (VDD), the voltage regulator should not be dis-
<DocScrutinizer>
abled. Otherwise, permanent damage to the device may occur.
<wpwrak>
then, when preparing for adding the reset pull-up, i first wanted to check if the reset pattern repeats (varies between all the time to once every 10-30 minutes). there, the board didn't respond normally. so i checked and found VDD all wrong.
<wpwrak>
i don't deliberately disable the regulator :)
<wpwrak>
of course, with the chip acting all weird, there's no telling what may happen ...
<DocScrutinizer>
Output Voltage Output Current = 1 to 100 mA 3.0 3.3 3.6 V
<wpwrak>
yes
<DocScrutinizer>
5V VDD is a sufficient condition to declare the device dead
<DocScrutinizer>
I'd not bother about reason for it resetting
<wpwrak>
yeah. there's no telling what damage this does.
<wpwrak>
added 4k7 pull-up to board #0 (the one that had trouble even flashing). didn't help. actually seems a little worse - no response at all instead of badly garbled reponses. well, who knows what is wrong with this one anyway.
<qi-bot>
[commit] Werner Almesberger: atusb.brd: increased zone clearance from 8 mil to 10 mil, for solderability http://qi-hw.com/p/ben-wpan/ed591cf
<qi-bot>
[commit] Werner Almesberger: atusb.brd: added via from nRST/C2CK, to allow for addition of pull-up http://qi-hw.com/p/ben-wpan/7514fa8
<kristianpaul>
wpwrak: (m1 workshop) so i was confused, hmm yes yday was fpga intro workshop (warm-up)
<wpwrak>
kristianpaul: and today those who yesterday happily thought they knew everything are confronted with the harsh truth of how much they still don't know ;-)
<wpwrak>
ah, sebastien is here. so let me rephrase this: [...] harsh truth of how much still separates them from greatness ;-)
<kristianpaul>
;-)
<wpwrak>
kristianpaul: btw, pcb making update: if you mix 3x% HCl with 3x% H2O2, the acid rips the copper away within about half a minute. the reaction is powerful enough that the acid starts to boil. very impressive, although a bit scary :)
<kristianpaul>
boil? wow
<kristianpaul>
scary indeed, thanks for the advice
<wpwrak>
kristianpaul: i tried this with a 10x15 cm board. my usual boards are much smaller, perhaps 2x3 cm. the large board was single-sided, which is probably a good thing given the intensity of the reaction :) so, 12.5 times the surface, 12.5 times the energy ...
<kristianpaul>
Dit it may etch the tonner part if let it a bit more time "boiling"?
<wpwrak>
the board had two damaged traces. but i don't know if this was caused by the strong acid or if there was some other problem. (i did the thing in a bit of a hurry)
<wpwrak>
just to pass power (up to 3 A, voltage as desired). the board connects to a lab supply on one end and a ribbon cable with up to 10 pairs of 22 R on the other. the idea is to launch fireworks with it. alas, it didn't work for some reason, although it behaved well in tests. need to check. it may see action on the 31st again :)
<wpwrak>
andres-calderon: soon, we'll need to split the schematics history into several pages :-)
<wpwrak>
andres-calderon: by the way, regarding 0402, 0603, and such: the footprints in kicad-libs are designed for pick and place machines. if you envision manual soldering/rework to be a common occurrence, you may be interested in the 0402-M, 0603-M, etc., variants, which are 0.1 mm longer than the regular footprints.
<wpwrak>
(the regular 0402, 0603, etc., can be soldered by hand, but it's not very convenient. 0402-M, etc., are much easier.)
<wpwrak>
unfortunately, the length change between xxxx and xxxx-M also changes the pad centers. so you can't just replace one with the other in an existing layout.
<andres-calderon>
wpwrak:  mmm, interesting, may  be too late to change...
<wpwrak>
andres-calderon: also depends on whether it would actually be useful :)
<wpwrak>
andres-calderon: in my case, easy soldering is quite important. in your case, probably less
<DocScrutinizer>
wpwrak: for non-commercial sparkoff 22R aren't really good. I frequently used those small flash bulbs filled with oxygen and magnesium wool - usually blueish. they get so damn hot they ignite blackpowder easily
<wpwrak>
ah yes, they should be good too. only .. where would you find these in this century ? ;-)
<DocScrutinizer>
wpwrak: I'd think one of those flash bulbs taped to a fuse and maybe even wrapped with tin foil, should work nicely
<DocScrutinizer>
wpwrak: (find) good question
<wpwrak>
in my tests, the 22 R ignited things reliably. it was only in the non-test use that they failed. (naturally - what would be the point in alerting me to any problems before ?)
<wpwrak>
of course, maybe it was something else that broke. i didn't examine the thing yet.
<wpwrak>
yeah, i used them too. many years ago ... :)
<wpwrak>
nice. now in argentina ? :)
<DocScrutinizer>
you need to use argentinean ebaeh for that :-D
<wpwrak>
besides, they're quite big. the resistors are a LOT smaller. many of the firework items are smallish, too. it's not like we'd be firing a Saturn V here
<kristianpaul>
wpwrak: you can use nichrome
<wpwrak>
we have that, "mercadolibre" :)
<kristianpaul>
yup :)
<wpwrak>
kristianpaul: nichrome ... just plain wire ? i tried that too, steel wire. works, but takes forever to heat up. and it's bulky.
<wpwrak>
one of the goals is quite specifically not to tie some heavy weight to the rockets that may jerk them off course ;-)
<kristianpaul>
you can attach some nichrmoe/resistor  wire and some match (aka fosforo)
<wpwrak>
naw, a pair of resistors will burn nicely if treated properly :)
<wpwrak>
complexity is the enemy of a reliable design :)
<kristianpaul>
agree
<kristianpaul>
how fast you need the resistor to ge ignited?
<kristianpaul>
s/ge/get
<wpwrak>
i usually aim for 2-5 s
<kristianpaul>
are you  using just DC?
<wpwrak>
for things that are difficult to ignite, flares and such, i used a sparkler as a first stage. they burn with a lot of heat