<DocScrutinizer05>
(stylus) I made very detailed suggestion how to place sensor
<Oksana>
RoHS compliant, halogen-free
<wpwrak>
DocScrutinizer05: the two of us discussed it at length, but does nik know where you want that critter to go ?
<wpwrak>
BMX055 looks nice. so you'll suggest it to nik ?
<Oksana>
BMP280 operates at lower noise, supports new filter modes and an SPI interface within a footprint 63% smaller than the BMP180
<Oksana>
Hmm... Both these sensors measure Pressure And Temperature. As expected, BMP280 looks better than BMP180, in all regards.
<Oksana>
The clearance above the metal lid shall be 0.1mm at minimum. For the device housing appropriate venting needs to be provided in case the ambient pressure shall be measured. Liquids shall not come into direct contact with the device. Avoid the exposure of BMP280 to strong light sources. Soldering may not be done using vapor phase processes
<wpwrak>
Oksana: careful with SPI. we want I2C. fewer pins.
<Oksana>
Maybe, I could order a thousand of LP55231SQE/NOPB to Neo900 address?
<Oksana>
To Gerry:
<Oksana>
Dear Gerry, I have received 8 dome sheets. I have ordered 20 dome sheets. When will I receive 12 dome sheets? Sincerely, Oksana
<Oksana>
Sehr geehrte Gerry, Ich habe 8 Einheiten erhalten. Ich habe 20 Einheiten bestellt. Wann erhalte ich 12 Einheiten erhalten? Mit freundlichem Gruß, Oksana
<wpwrak>
# s/erhalte/werde/ :)
<wpwrak>
so gerry are already struggling with the unexpected sales volume. i guess we better make that dome sheet replacement a priority ..
che11 has quit [Ping timeout: 245 seconds]
che1 has quit [Ping timeout: 264 seconds]
freemangordon has quit [Ping timeout: 245 seconds]
Kabouik has quit [Ping timeout: 264 seconds]
<DocScrutinizer05>
no VPS? eeew
<DocScrutinizer05>
9, not 8
<DocScrutinizer05>
no VPS might be a KO criterion
<wpwrak>
what is VPS ?
<DocScrutinizer05>
Vapor Phase Soldering
<DocScrutinizer05>
>>may not be done using vapor phase processes<<
<DocScrutinizer05>
:-(
<wpwrak>
ah. well, taht would be something to ask to the fab. maybe they can also do a mixed process: first VPS, then reflow for the remaining parts.
obsed has joined #neo900
<DocScrutinizer05>
no
<DocScrutinizer05>
double-soldering is a complete nogo
<DocScrutinizer05>
most parts are explicitly not allowed to get exposed to soldering temperature more than once
<DocScrutinizer05>
but since we want baro/thermo as much close to the "outside" as possible anyway, we might be able to place it on upper UPPER in a corner maybe, and no other components nearby, so we could IR reflow or hotair that corner only
<DocScrutinizer05>
or we use a tiny flex PCB with a b2b conn, much like N97 cam
<DocScrutinizer05>
:-D or a slightly longer FPC so we could flip out the whole sensor from battery lid to the outside, for really accurate temperature readings :-)
vakkov has quit [Ping timeout: 264 seconds]
<wpwrak>
would "reflow everything" be an option ?
<DocScrutinizer05>
modem is hard to reflow
<wpwrak>
yes, i remember that from the galden [sp?] discussion. but would it be a complete no-no ? i.e., unavoidably resulting in poor yield ?
<DocScrutinizer05>
I'd rather glue the BP180 than IR/convection reflow the modem
<wpwrak>
okay ;-)
<DocScrutinizer05>
but that's sth to ask Nik
<DocScrutinizer05>
I'd even consider capton to seal BP280
<wpwrak>
the barometric critter does sound fairly fragile. "no liquids" ... what if we have condensation ?
<DocScrutinizer05>
but that's sth to ask Bosch
<DocScrutinizer05>
no light :-o
<DocScrutinizer05>
NB BP180 only says "no liquid" but doesn't mention VPS
<DocScrutinizer05>
which might just be a "we missed to mention it explicitly"
<DocScrutinizer05>
but honestly, no liquids? really now? what about washing the PCBA?
<DocScrutinizer05>
and yeah, what's about due point?
<wpwrak>
highly selective cleaning strengthens the immune system :)
<DocScrutinizer05>
~weather eddn
<infobot>
Nuernberg, Germany; (EDDN) 49-30N 011-03E 318M; last updated: 2014.10.31 0220 UTC; Dew Point: 48 F (9 C); Pressure (altimeter): 30.30 in. Hg (1026 hPa); Relative Humidity: 93%; Sky conditions: overcast; Temperature: 50 F (10 C); Visibility: 3 mile(s); Wind: Variable at 2 MPH (2 KT)
<DocScrutinizer05>
dew point
<wpwrak>
maybe they want to discourage submerged use. e.g., as liquid pressure sensor.
<DocScrutinizer05>
you're kidding, are you?
<wpwrak>
well, some people may get ideas ... :)
<DocScrutinizer05>
I think they meant exactly what they wrote
<DocScrutinizer05>
though I fail to get why
<wpwrak>
"we dunked that BMP180 into our oil pipeline and the measurements are all over the place. what crap did you sell us ?" :)
<DocScrutinizer05>
no
<wpwrak>
same as "remove hand before slamming car door shut" ;-)
<DocScrutinizer05>
never seen that
<nox->
sounds like "micowave is not for drying animals"
<DocScrutinizer05>
also the warning about liquids is under 2mounting and assembling", not under "recommended operation"
<Oksana>
9, not 8? Alright, and s/erhalte/werde/... Ask Bosch about ambient-vulnerability of BMP280 vs BMP180 As in: winter with snow? flood in India? Dew point?
<Oksana>
"micowave is not for drying animals"... Did Watson tell that to Holmes?
<DocScrutinizer05>
afk
<nox->
there was a court case in the us...
<Oksana>
About hot coffee spilled in McDonalds, and serious burns?
<nox->
that too
<Oksana>
microwave*, even
<wpwrak>
i vaguely remember having seen such warning stickers on car doors. in the us, of course :) can't find pictures, though
<nox->
Oksana, someone put a cat(?) in microwave and then sued the manufacturer
<wpwrak>
Oksana: it's what the cat told schroedinger. that's when he decided to use poison the next time.
<nox->
"only in america..."
<nox->
wpwrak, haha
<Oksana>
Hell, and did they expect to warm microwave from a freezing cold winter? Why would someone do that?
<Oksana>
to warm cat in microwave*, even
<nox->
to dry wet cat yeah
<Oksana>
Poor cat... Did it survive?
<nox->
guess not
<Oksana>
Soon animals will have to get EMP bombs to use against random technology: microwave, washing machine, vacuum cleaner, garbage collector...
<Oksana>
In self-defence against humans.
<nox->
there also was a court case about someone mistaking cruise control for "autopilot"...
<Oksana>
Sehr geehrte Gerry, Ich habe 8 Einheiten werden. Ich habe 20 Einheiten bestellt. Wann werde ich 12 Einheiten erhalten? Mit freundlichem Gruß, Oksana
<DocScrutinizer05>
google darwin award
<DocScrutinizer05>
9
<Oksana>
I remember a pilot who had to drive a car. He tried to fly over an obstacle...
<Oksana>
Sehr geehrte Gerry, Ich habe 9 Einheiten werden. Ich habe 20 Einheiten bestellt. Wann werde ich 11 Einheiten erhalten? Mit freundlichem Gruß, Oksana
<nox->
s/werden/erhalten/
<Oksana>
Sehr geehrte Gerry, Ich habe 9 Einheiten erhalten. Ich habe 20 Einheiten bestellt. Wann werde ich 11 Einheiten erhalten? Mit freundlichem Gruß, Oksana
<Oksana>
Weird. Synonyms?
<wpwrak>
s/11/die übrigen/ # clearer
<nox->
true
<Oksana>
Sehr geehrte Gerry, Ich habe 9 Einheiten erhalten. Ich habe 20 Einheiten bestellt. Wann werde ich die übrigen Einheiten erhalten? Mit freundlichem Gruß, Oksana
<wpwrak>
s/Gruß/Grüssen/ # singular is unusual. dunno whether it's ~ss~ or ~ß~ in the plural form. they didn't teach me that in switzerland :)
<nox->
ss in plural
<wpwrak>
but if in doubt, i'd just use english.
<nox->
heh yeah
<Oksana>
Dear Gerry, I have received 9 units. I ordered 20 units. When will I receive the remaining units? Sincerely, Oksana
<Oksana>
Sehr geehrte Gerry, Ich habe 9 Einheiten erhalten. Ich habe 20 Einheiten bestellt. Wann werde ich die übrigen Einheiten erhalten? Mit freundlichem Grüssen, Oksana
<Oksana>
About s/Gruß/Grüssen/ : I just copied from incoming letter from Gerry ;)
<wpwrak>
yup, sounds good
<wpwrak>
heh, maybe they would prefer english, too ;-))
<nox->
gerry does sonde kinda english name...
<Oksana>
I sent both. Tandem.
<Oksana>
Both English and German.
<Oksana>
Ask Bosch about ambient-vulnerability of BMP280 vs BMP180 ?
vakkov has joined #neo900
wpwrak has quit [Ping timeout: 245 seconds]
<Oksana>
Too quiet. Not just on this channel, but everywhere I try to look. It seems people have their own underwater/underground channels of discussing the projects.
wpwrak has joined #neo900
nox- has quit [Quit: Leaving]
fling has joined #neo900
<Oksana>
Good evening...
Nokiabot has joined #neo900
Nokiabot has quit [Client Quit]
Oksana has quit [Remote host closed the connection]
freemangordon has joined #neo900
fling has quit [Ping timeout: 255 seconds]
svree4 has joined #neo900
Oxyd76 has quit [Remote host closed the connection]
Oxyd76 has joined #neo900
fling has joined #neo900
krobelus_ has joined #neo900
vakkov has quit [Ping timeout: 256 seconds]
sixwheeledbeast has quit [Ping timeout: 256 seconds]
vakkov has joined #neo900
vakkov has quit [Ping timeout: 255 seconds]
krobelus_ has quit [Ping timeout: 264 seconds]
vakkov has joined #neo900
che1 has joined #neo900
che11 has joined #neo900
paulk has joined #neo900
paulk-collins has joined #neo900
paulk has quit [Remote host closed the connection]
paulk-collins has quit [Remote host closed the connection]
paulk-collins has joined #neo900
arcean has joined #neo900
che11 has quit [Ping timeout: 255 seconds]
che1 has quit [Ping timeout: 250 seconds]
norly has joined #neo900
paulk-collins has quit [Quit: Quitte]
kolp has joined #neo900
norly has quit [Quit: Leaving.]
Wikiwide has joined #neo900
<Wikiwide>
Hmm... 25*, even. Not 20. That explains origin of confusion around 9 and 8. When 9 are already delivered, and 16 remain, and there is limit of 10 (per order or per week?), 16/2=8, and it is expected that the 16 units are delivered as 2 parcels, 8 units each.
<DocScrutinizer05>
moin
krobelus_ has joined #neo900
<Wikiwide>
Just checked: no 'refund' from Gerry. So, waiting for a reply from the managing director :) Google+ gives funny mix-up of business and personal life of a human.
<DocScrutinizer05>
why you "considered" to get truckloads of LP55231?
<Wikiwide>
Because there is a large difference in price-per-unit for 500 and 1000 ?
<DocScrutinizer05>
I hope that google+ doesn't give _anything_ for me
Pali has joined #neo900
<Wikiwide>
Hmm... I don't remember sending you a email, and I am too lazy to jump into Google+.
<DocScrutinizer05>
well such price difference applies to absolutely every single component on Neo900 and even to heineken 6packs
<Wikiwide>
For Gerry, an image of snow-boarding was just too funny to not dive into his Google+. There is a photo of two cats...
<DocScrutinizer05>
weird
<DocScrutinizer05>
the guy is called thomas schreiber
<Wikiwide>
Well, if I knew where such difference is the largest between 500 and 1000, I would get 1000 of this unit, just because I do not like merchants making money off 'small' buyers.
krobelus_ has quit [Ping timeout: 255 seconds]
<DocScrutinizer05>
can't think of a private life of gerry-mobil
<Wikiwide>
:) :)
<DocScrutinizer05>
why THIS component?
<Wikiwide>
Just too obvious a difference
<DocScrutinizer05>
that applies to a lot of other components as well
<Wikiwide>
Maybe, others have more detailed scale (100 200 500 1000 10000) and this one has only two steps (500 1000)
<DocScrutinizer05>
anyway please don't order chip components for Neo900 UG
<Wikiwide>
Alright, alright
<Wikiwide>
Am I too noisy? :) I am glad that there aren't that much EOL components
<DocScrutinizer05>
it's still ok :-)
<DocScrutinizer05>
and any help appreciated. Thanks a lot!
<DocScrutinizer05>
actually without your domesheet action we hadn't started talking to snaptron yet
<Wikiwide>
How is it going? With snaptron?
<DocScrutinizer05>
very nice company. I need to prepare a pdf (et al) drawing of domesheet for them to give quotation
<Wikiwide>
No problem :) I just cannot stand still. That does not play well with slow software, unfortunately.
<Wikiwide>
How is drawing of domesheet going? Scan an existing domesheet into a PDF?
<DocScrutinizer05>
my highres scans are not compatible with their process
<DocScrutinizer05>
I ponder to create a pcb pdf printout
<Wikiwide>
I mean, scan as in paper-scanner? Not needle-3D-scanner? Or does it have to be something hand-crafted, nice and neat, like block diagram?
<DocScrutinizer05>
from eagle
che1 has joined #neo900
<Wikiwide>
From eagle... That would be nice for curious public :)
<Wikiwide>
If you prefer gimp, it's fine, too. :) I just find it too cluttered.
* DocScrutinizer05
finally starts slaughtering a domesheet, to rip off all the gory last details of internal structure
<bencoh>
oO
<bencoh>
a working domesheet ?
<Wikiwide>
:) There is a slaughtered keyboard here, though for a less noble cause. First apple juice was spilled upon it, then the electricity plus the juice ate the metal wires within it, then carbon glue was used upon it to 'restore' functionality of said wires, but since the attempt failed, the keyboard was left as puzzle-toy for inquiring minds...
<DocScrutinizer05>
W*T*F?! the white dots (lightspreader escape points) are on a separate foil glued to the "thick" lightspreader layer
<Wikiwide>
Easier to print?
<DocScrutinizer05>
prolly
<Wikiwide>
Printer cannot chew the thick layer? Only the paper-thin foil?
<bencoh>
Wikiwide: poor thing
<DocScrutinizer05>
the foil thickness is... incredibly thin
<Wikiwide>
No wonder they do not refurbish dome sheets...
<DocScrutinizer05>
actually I cannot say if it's a foil, a printed varnish, or just parts of the lightspreader surface specially polished so it looks different
<DocScrutinizer05>
I cannot detect any thickness at all
<Wikiwide>
Heh... Wondrous technology.
<DocScrutinizer05>
thickness of the *complete* domesheet plastic (sand domes and protective pink/blue foil) is 0.37mm
<DocScrutinizer05>
s/sand/sans/
<DocScrutinizer05>
the lightspread "thick" transparent foil seems 1/3 of that
<DocScrutinizer05>
the thin foil with dots on it - if it's really a foil - is <1/20 of the spreader foil thickness
<DocScrutinizer05>
gosh, where's my micrometer screw ruler? caliper fails for that stuff
che1 has quit [Ping timeout: 245 seconds]
Kabouik has joined #neo900
<drathir>
anyone hear about drivedroid thing?
<Wikiwide>
First time I hear about it..
<Wikiwide>
Turn your phone into LinuxLive for a PC? And there is a paid (no ads) version of it?
<DocScrutinizer05>
2014-10-31-120806_dotfoil-on-spreaderfoil.jpg has "0.5" ->cm
<drathir>
only association minds from my side...
<DocScrutinizer05>
you can clearly see the dotfoil on spreaderfoil there
<drathir>
thats dot in foil...
<DocScrutinizer05>
while funny enough the white dots on dotfoil look darker
<DocScrutinizer05>
they are not as glossy as the clear dotfoil
<DocScrutinizer05>
spreader foil also not as glossy so it also looks darker
<drathir>
and thats dots are very accurate...
<DocScrutinizer05>
yes, distribution and size and density of dots determine how much light will escape spreader foil
<DocScrutinizer05>
when you send light into the spreader foil, the dots light up bright
<DocScrutinizer05>
so you want small and few dots near the LEDs and more bigger dots far away from LEDs
<DocScrutinizer05>
and of course the dots are tailored to match the keykaps and where those actually need backlight
<DocScrutinizer05>
keycaps*
<DocScrutinizer05>
~s/l/K/
<DocScrutinizer05>
~s/k/K/
<DocScrutinizer05>
sed=0 is kinda annoying in that it *completely* disables s/x/y/
che1 has joined #neo900
<obsed>
man i was hoping the n900 series from nokia was going to really take off
<obsed>
and that Nokia would leap onto the bandwagon and make high end smartphones based on Maemo lol
* DocScrutinizer05
thinks to completely split domesheet under test and place the spreader on black background to do a proper non-distorted complete shot of the dots
<obsed>
but microsoft implanted a trojan into nokia :/
<DocScrutinizer05>
we know
<DocScrutinizer05>
I had way less gray hair now when they actually did
<DocScrutinizer05>
but then building Neo900 is also fun, which i'd miss when Nokia had built and sold it
<bencoh>
well ..... would have still been simpler ;p
<DocScrutinizer05>
sure
<DocScrutinizer05>
and I could do a boring job to earn truckloads of money in industry
<bencoh>
huhu, true as well
louisdk has joined #neo900
<DocScrutinizer05>
now I'm doing an exciting and fun job (well at times, sometimes it's also a nightmare, thus the gray hair) and not earning what I could earn as streat cleaner
<DocScrutinizer05>
street*
<obsed>
yeah Neo900 is way cool =)
<DocScrutinizer05>
anyway, lightspreader: seems completely feasible to make such a transparent thingie from arbitrary plastic foil and print some dots on it. Place that on a standard domesheet and you're mostly done
<DocScrutinizer05>
it's not mandatory to glue the two parts in fab already
<Wizzup>
I saw a mention of DVI in the feasibility study, is it at this point still considered [somewhat] feasible?
<Wizzup>
(even with a casemod of some kind)
<DocScrutinizer05>
it's still somewhat under evaluation
<DocScrutinizer05>
but a tough item
<DocScrutinizer05>
I admit that I didn't do much to push it, during last maybe 3 or 4 months
<DocScrutinizer05>
btw we need another few (maybe 4) BeagleBoard-xM for R&D
<obsed>
hmm what about Host-mode USB, e.g. plugging in usb keyboards or other things
<obsed>
anyone interested in that?
<obsed>
hmm google says it might be in there already =)
<Wizzup>
DocScrutinizer05: alright, I wouldn't mind paying some extra for that if it would bring the costs up
<DocScrutinizer05>
huh? we have ultrafine OTG compatible hostmode, with a roaring 1000mA VBUS
<obsed>
n900 and maemo was the first phone and OS where i got really excited about them, and thought, wow, a phone i actually really love the design of hehe
Wikiwide has quit [Quit: ChatZilla 0.9.91 [SeaMonkey 2.30a1/20140613003001]]
<freemangordon_>
YAY! seems like the last function is down :)
<obsed>
hmm, bundled OS Debian
<DocScrutinizer05>
WOW!
<obsed>
man.. i have so been wanting to run linux on my phone for ages, specifically a distro like debian
<obsed>
i call android google tho. android = google. but replicant = linux
<obsed>
at least in ethos
<freemangordon_>
yep, at least EAP seems to be 100% ready, will test the record module when I am back home
<DocScrutinizer05>
drathir: yup, for e.g. playback of DVD video it will perform twice as good
<DocScrutinizer05>
at least
<drathir>
thats all great news...
<drathir>
and 1A in host mode looks amazing too...
<DocScrutinizer05>
well, it will kill your battery in ~60min
<DocScrutinizer05>
when you actually want to use that 1A all the time
<drathir>
yea but still posibility matter...
<DocScrutinizer05>
but most USB devices only need bursts that high, every now and then
<DocScrutinizer05>
e.g HDD only for spin up and for head positioning
<DocScrutinizer05>
greedy UMTS dongles only for TX - not that you'd need such dongle ;-)
<DocScrutinizer05>
thus a 1A is perfectly sane to provide, even when it dosn't make sense for constant full power
<DocScrutinizer05>
you're not supposed to operate your USB coffe mug warmer from Neo900 ;-P
<DocScrutinizer05>
(though you could, for around 45 minutes)
<Wizzup>
DocScrutinizer05: though, I guess, you'd need some ... interesting dvi connector
<Wizzup>
since even mini dvi is huge, no
<Wizzup>
micro-dvi*
<obsed>
i'd probably stick a big battery onto the Neo900 maybe
<DocScrutinizer05>
you're free to do that
<obsed>
nice
<obsed>
a phone with freedom:)
<DocScrutinizer05>
that's the idea
<obsed>
ive been looking for one of those hehe
<obsed>
i figure ok i wouldnt allow AdService.exe to run as a system service on my PC
<obsed>
so why the hell would i accept a verizon phone that has a hidden AdService program running, not killable by any of the default tools unless you root your phone lol
<drathir>
;p
<DocScrutinizer05>
haha
<DocScrutinizer05>
indeed
<obsed>
quite appaling design that is lol
<drathir>
obsed: install comodo...
<DocScrutinizer05>
my friend can't use google search on her note2 without the damn browser asking her to log in to google
<obsed>
yeah thats pretty ridiculous, the way these phones are designed, `sudo cp /* /mnt/googlescloud`
<obsed>
cp -R even
<obsed>
lol
sixwheeledbeast has joined #neo900
<obsed>
i remember one of my phones on first boot it wouldnt let me even use the phone at all without creating a google account lol
<obsed>
ive installed replicant atm
* Wizzup
still uses n900 daily, but is slowly becoming unhappy with some outdated sw (like browser)
<Wizzup>
I mostly just use ssh and music though :)
<Wizzup>
and mobile hotspot, of course
<bencoh>
it's not that outdated, it's mostly slow
<Wizzup>
It's firefox 3.5, no?
<Wizzup>
-based
<bencoh>
yeah something like that
<bencoh>
but for mobile browsing I couldnt care less, fancy css3/js animations wont run on it anyway (too greedy)
<bencoh>
wouldnt*
<obsed>
yeah it will be nice to get an upgraded perf with the Neo900
<bencoh>
I'm pretty sure web stuff will still be slow, and will quickly get "obsolete" anyway
<Wizzup>
yes, don't care too much about css/js stuff, mostly security related (if any)
<Wizzup>
(probably are some)
<bencoh>
(considering that those thing can bring down a core i7)
<DocScrutinizer05>
freemangordon_: what I'm looking at?
<DocScrutinizer05>
kerio: "nice thing about systemd: exposes poettering's mindset (if that's actually needed to get exposed even further, after PA and al the other stuff)"
<wpwrak>
hmm. placing the key mat directly on LEDs may not be entirely hopeless. but we'd need about 14 LEDs instead of just six
<DocScrutinizer05>
don't see this
<DocScrutinizer05>
you had a look at bottom side of keymat?
<wpwrak>
i put the mat over a 0603 LED i have and looked at how much light gets through. even if the mat touches the (relatively tall) LED, enough like makes it to clearly illuminate the larger items on a cloudy day
<DocScrutinizer05>
I'd say we'd need at least 38 LEDs
<wpwrak>
the'd have to use thinner LEDs, though
<wpwrak>
my experiment was with the LED between four keys
<wpwrak>
if i move it straight under a key, it gets a LOT brighter, of course
<DocScrutinizer05>
you cannot put a LED "between 4 keys", there's steel
<wpwrak>
s/the'd/we'd/
<wpwrak>
the rubber goes under the steel, so there is a bit of room
<DocScrutinizer05>
no it doesn't
<DocScrutinizer05>
well, at least nothing that helps
<wpwrak>
every 0.1 mm helps :)
<DocScrutinizer05>
with what?
<wpwrak>
getting the light out
<DocScrutinizer05>
and you know the distance between that structure in keymat and the PCB surface is exactle the 0.3x mm thickness of domesheet?
<wpwrak>
yes, the LED i used is thicker than that. so it has a spreading benefit there.
<DocScrutinizer05>
I just pointed a laser on the X crossing of steel between 4 keys. The light in keycaps been near undetectable
<wpwrak>
also, when i increase the distance, things get brighter
<DocScrutinizer05>
basically the light been stray light from laser
<wpwrak>
my LED has an angle of 130 deg and a base thickness of 0.28 mm. let's see if we could do better ...
<DocScrutinizer05>
I won't go to length of placing a carton with a pinhole under keymat
<DocScrutinizer05>
it's not working
<DocScrutinizer05>
won't fly
<DocScrutinizer05>
we have a pretty easy proven concept consisting of a sheet of plastic for 0.0001ct
<DocScrutinizer05>
even painting the dots with vanish marker on it will yield better results than even 38LEds
<DocScrutinizer05>
varnish*
<wpwrak>
ah, so we already have a replacement ? i don't think you ever mentioned that
<DocScrutinizer05>
[2014-10-31 Fri 14:19:44] <obsed> that's an interesting looking lightspreader
<DocScrutinizer05>
[2014-10-31 Fri 14:49:08] <DocScrutinizer05> rather "normal"
<DocScrutinizer05>
[2014-10-31 Fri 14:50:29] <DocScrutinizer05> just needs some 0.15 .. 0.2mm clear (hard) plastic foil, a cutter and a printer with white translucent toner to do your own
<wpwrak>
ah, you can see from this already how the dispersion in the plastic works. okay. your optics are way ahead of mine :)
<DocScrutinizer05>
yes, I can see that, from the density of dots
<DocScrutinizer05>
everybody else can as well
<DocScrutinizer05>
the larger and more dots, the dimmer the light "pressure" inside spreader at that point
<DocScrutinizer05>
of course there's also the factor of "how much light do we want at this place", otherwise there wouldn't be any gaps at all
<wpwrak>
yes, that interpretation of the pattern clearly makes sense. but what i couldn't predict so easily are just these spreading properties for a different material (i.e., your low-cost plastic)
<DocScrutinizer05>
meh
<DocScrutinizer05>
you think there's any difference to the low cost plastic used in nokia domesheet?
<DocScrutinizer05>
it's all just a light pipe
<DocScrutinizer05>
with two parallel 100% refractive parallel surfaces
<wpwrak>
i know that there's a lot of research in optical properties of materials. not sure how much they vary at that scale, though
<DocScrutinizer05>
you could use water ice, plastic, silicon, or glass
<wpwrak>
how thick is the thing ?
<DocScrutinizer05>
even liquid water if you can make a thin film with parallel surfaces
<wpwrak>
ice :)
<DocScrutinizer05>
[2014-10-31 Fri 17:03:52] <DocScrutinizer05> [2014-10-31 Fri 14:50:29] <DocScrutinizer05> just needs some 0.15 .. 0.2mm clear (hard) plastic foil, a cutter and a printer with white translucent toner to do your own
<DocScrutinizer05>
[2014-10-31 Fri 17:12:07] <DocScrutinizer05> you could use water ice,
<wpwrak>
i think that would stretch the analogies a bit too far :)
<DocScrutinizer05>
no, it doesn't
Guest87435 is now known as FIQ
FIQ has joined #neo900
FIQ has quit [Changing host]
<DocScrutinizer05>
you can test that any time, with a garden hose and a LED
<DocScrutinizer05>
looks beautiful
<DocScrutinizer05>
when the light escapes the water beam where it breaks
<DocScrutinizer05>
or when you hold your finger in the beam and see it bright illuminated
<wpwrak>
yes, but i would expect a crystal to have different properties from an amorphous substance
<DocScrutinizer05>
no
<DocScrutinizer05>
not on this level and testbed
<DocScrutinizer05>
plastic is no crystal
<wpwrak>
that's why i'm saying that water ice would take it a bit too far, because it is crystalline (at least in its most common form)
<DocScrutinizer05>
it's all just about light on medium break surface where media with different refraction factor touch
<wpwrak>
anyway, there's only one way to be sure: make an experiment :)
<DocScrutinizer05>
I won't make any experiments with 14 LEDs
<wpwrak>
one will be plenty to begin ...
<DocScrutinizer05>
and I don't need to make experiments with a lightspreader design that's age old and used literally everywhere
<DocScrutinizer05>
I already did an experiment with something way easier, better, smarter than a set of LEDs: a laser pointer
<DocScrutinizer05>
result: your concept won't fly for Nokia keymat
<wpwrak>
yes, superman. only lasers for you ;-)))
<DocScrutinizer05>
afk, cya
mvaenskae has quit [Ping timeout: 255 seconds]
paulk-collins has joined #neo900
<wpwrak>
wow, 3M are creative. they made a transparencies film that's one side for ink the other for toner, so that you have to "stock only one". of course, it also means that you have a good chance of using the wrong side. the things we do to boost our sales ... :-))
mvaenskae has joined #neo900
vakkov has joined #neo900
<DocScrutinizer05>
spreader foil: 0.20mm
<wpwrak>
that's the specification of the N900 thing ? or the description of a product you found ?
Humpelstilzchen has quit [Ping timeout: 265 seconds]
<wpwrak>
i had a quick look at easily sourceable sheets for making transparencies. most seem to be in the 4-7 mil / 100-175 um range.
<DocScrutinizer05>
that's what my caliper says
<wpwrak>
the ink printer ones probably won't work so well since they already have a pre-roughened surface. so that would leave laser, copier, and manual. alas, 7 mil seem somewhat rare
<DocScrutinizer05>
thickness of the "schnellhefter" transparent frontside: 0.19mm
<wpwrak>
ah, good
<wpwrak>
oh, and are you sure these dots are actually printed on top of the plastic ? it would seem to make more sense if they were distortions that go into it
<wpwrak>
(of course, something printed on top may also produce distortions underneath. probably too small to tell with our means)
<DocScrutinizer05>
no, they are for sure on top, for all lightspreaders. ideally roughened surface, but printed stuff will prolly do as well
<DocScrutinizer05>
or rather, obviously does as well
Humpelstilzchen has joined #neo900
<wpwrak>
okay, that makes things simpler :)
<wpwrak>
now the white toner ;-)
<DocScrutinizer05>
the point is that refraction at medium frontier gets broken for the points
<wpwrak>
kewl. the first hits are for bleaching hair :)
<DocScrutinizer05>
you don't want discontinuities inside the pipe, you want gates to the outside
<DocScrutinizer05>
honestly, how hard can it possibly be to find a clear sheet of plastic with two plain parallel surfaces, of ~0.2mm thickness?
<DocScrutinizer05>
ideally some "hard2 plastic
<wpwrak>
one that will survive laser printing. but yes, that's the smaller problem
<DocScrutinizer05>
s/2/"/
<wpwrak>
finding a printer on the other hand ... (just google for white toner laser)
<wpwrak>
but maybe ink will do as well. may be easier.
<DocScrutinizer05>
use a stencil
<DocScrutinizer05>
or roughen surface directly with a laser engraving
<DocScrutinizer05>
or by etching
<DocScrutinizer05>
or even by milling
<wpwrak>
yes, i'm thinking about these processes. that's also why i asked about on top vs. inside the sheet
<DocScrutinizer05>
the dots don't need to be round either, and not even exact. As long as average density for the particular area is roundabout right
<wpwrak>
yes, could be some polishing process
<wpwrak>
for a mill, the tolerances are a bit on the tight side. not impossible, but it means suffering. if the tool can be softer, though (i.e., just to scratch the surface a little), then that would open more options
<DocScrutinizer05>
I'd prolly get them printed in a professional offset printing machine of a friend
<DocScrutinizer05>
can't see _any_ problem with that
<DocScrutinizer05>
he will also find the right ink for me
<DocScrutinizer05>
they print such stuff every day
<wpwrak>
the devil is usually in the detail :) e.g., things like unexpected process variations can keep you busy for a while. but yes, the only way to find out is to try it. see what works see what doesn't
<DocScrutinizer05>
"go to citydruck, tell them "need this printed in white onto 0.2mm clear polydunnowhat sheet" is proven to work
<DocScrutinizer05>
I don't care about any unlikely unexpected details they run into
<wpwrak>
well, ask him. then let's hope for the best. if you have a nice solution next door, even better :)
svree4 has quit [Ping timeout: 255 seconds]
paulk-collins has quit [Ping timeout: 265 seconds]
<DocScrutinizer05>
what worries me most is cutting this complex shape
<DocScrutinizer05>
but probably they have a way to stick tiny blades on huge rolls of their printing or engraving machine
<DocScrutinizer05>
embossing machine even
<DocScrutinizer05>
if all else fails, we got a laser cutter in fablab 1km away
Kabouik has quit [Ping timeout: 255 seconds]
<DocScrutinizer05>
what worries me most is the contact surface to white domesheet. It mustn't break the 100% refraction property of the medium surface
svree4 has joined #neo900
<DocScrutinizer05>
prolly needs a glue with significantly lower refraction coefficient than the plastic
<DocScrutinizer05>
aka optical density
<DocScrutinizer05>
aka lower lightspeed in that medium
<DocScrutinizer05>
err higher
mvaenskae has quit [Quit: Lost terminal]
<DocScrutinizer05>
or we use silver coating
Kabouik has joined #neo900
<DocScrutinizer05>
actually silver is a proven concept in N900. Though on the upside of lightspreader. The sticky tape that's white at visible outside is actually silver where it sticks on top of lightspreader
<DocScrutinizer05>
but seems a lot of lightspreaders work with white backside
<freemangordon>
hmm, still some distortion, anyway, way better than yesterday
paulk-collins has joined #neo900
<freemangordon>
looks like low freqs are lost from time to time
vakkov has joined #neo900
<freemangordon>
oh, there are some functions my test do not covre
FIQ has quit [Excess Flood]
mvaenskae has joined #neo900
b1101 has quit [Quit: b1101]
FIQ has joined #neo900
b1101 has joined #neo900
xes has joined #neo900
astr has quit [Ping timeout: 255 seconds]
astr has joined #neo900
astr has quit [Ping timeout: 244 seconds]
astr has joined #neo900
paulk-collins has quit [Ping timeout: 240 seconds]
nox- has joined #neo900
norly has joined #neo900
<wpwrak>
(leuchttafel) interesting. didn't know these critters.
<wpwrak>
and now i also know where those LED strips come from :)
mvaenskae has quit [Ping timeout: 264 seconds]
louisdk has quit [Read error: Connection reset by peer]
jonwil has joined #neo900
<freemangordon>
jonwil: hi!
<freemangordon>
jonwil: need a little help :)
<jonwil>
hi
<jonwil>
saw your work, it looks good
<jonwil>
and what do you need help with?
<freemangordon>
could you please looak at the function CalcGainVector and tell me what I am doing wrong. If i feed mudrc_process with up to (and including) 0x10007F, the output matches to stock
<freemangordon>
one sample above make the outputs different
<freemangordon>
0x10007F is the number of the input samples
<freemangordon>
that functions is called only in EAP_MdrcDelaysAndGainsInt32_Process
<freemangordon>
I replaced the calls to EAP_MdrcDelaysAndGainsInt32_Gain_Scal and EAP_MdrcDelaysAndGainsInt32_Gain_Scal1 with those exported to stock, still the same
<jonwil>
let me take a look
<freemangordon>
so there is something wrong in either the other parts of EAP_MdrcDelaysAndGainsInt32_Process (which I doubt as those are very simple) or in CalcGainVector
<freemangordon>
which is simple too, but maybe the compiler fucks me with some signed/unsigned/16/33/64 bit nonsense
<freemangordon>
s/33/32
<freemangordon>
what I see in the disasm is that gcc emits umulll, while the original code is smull for "delta = EAP_LongMultQ15x32(instance->m_oneOverFactorQ15..."
<freemangordon>
BTW EAP_LongMultQ15x32is used only there, so that function might be the culprit
<jonwil>
how do you know its doing EAP_LongMultQ15x32?
<jonwil>
I dont see that function in either Fremantle or Harmattan
<freemangordon>
take a look in x86 code
<jonwil>
I dont see what could be wrong
<freemangordon>
me neither, but something is wrong in EAP_MdrcDelaysAndGainsInt32_Process as if I call the stock one instead ther is noddiference no matter the size of the input buffers
<freemangordon>
sorry for the typos :)
<freemangordon>
btw gcc plays tricks on me, see my last commit, the part "gain = gainVector[0];"
<freemangordon>
if I pass gainVector[0] (for example) directly, then it doesn't work
<freemangordon>
but if I assign it to int32 var gain, then it works
<freemangordon>
or I have NFC of what C is
<freemangordon>
:D
<wpwrak>
sounds like a possible signedness issue ?
<freemangordon>
something overflows, but I can't identify it
<wpwrak>
what are the types of gain and gainVector[0] ?
<freemangordon>
int32
<freemangordon>
gainVector is int32*
<freemangordon>
int32 is int32_t iirc
<wpwrak>
int32 gain, *gainvector, foo; gain = gainvector[0]; /* doesn't work */ foo = gainvector[0]; gain = foo; /* does work */ ?
<freemangordon>
well, it is a function f(), f(gainVector[0]) does not work, gain=gainVector[0]; f(gain); does work
<freemangordon>
a mystery, ain't?
<freemangordon>
:)
<wpwrak>
yes, looks as if there was a problem somewhere else
<freemangordon>
and f() is declared as "EAP_LongMult32Q15x32(int32 in1, int32 in2)"
<wpwrak>
what happens to the values ? what's gainVector[0] and what does it become when it doesn't work ?
<freemangordon>
wpwrak: hard to say, look at the code ;)
<freemangordon>
anyway, I am going afk for a while.
<freemangordon>
(GF :) )
<DocScrutinizer05>
err maybe I'm telling shit but "EAP_LongMult32Q15x32(int32 in1, int32 in2)" seems incompatible with f(gainVector[0]) and f(gain)
<DocScrutinizer05>
since, well, it has 2 params, right?
<wpwrak>
nasty code
<wpwrak>
i wonder if what you're seeing could be a pointer aliasing problem. e.g., if the first call overwrites the input of the 2nd, then having a copy in "gain" would help
<DocScrutinizer05>
nevermind
<jonwil>
I have no clue about all this math and stuff so I will let the experts figure it out :P
<DocScrutinizer05>
breakpoint and stack analysis would help a lot
<jonwil>
meanwhile I have to find a way to edit a pdf file and chance some black text to a color that looks black to the naked eye but will be printed by my printer in color rather than black since my printer is failing to print black for reasons I dont know :)
<DocScrutinizer05>
umm
<DocScrutinizer05>
inkjet?
<wpwrak>
of course, using something like "int32" if we have well-established "int32_t" should, even if it works, be punishable be slow agonizing death
<jonwil>
yeah its an inkjet and the black ink is definatly there :)
<jonwil>
I can see black ink in the cartridge
<DocScrutinizer05>
do a printer test print or diagnostic sheet print
<jonwil>
I did that, it doesn't print black there either
<DocScrutinizer05>
prolly jet clogged
<jonwil>
the drvier says "ink level is full" and I ran the print head cleaning too with no help
<jonwil>
I may need a new printer or a new print head piece
<wpwrak>
freemangordon: i would add a bunch of printf and see what really happens. all the memmove and memcpy make it too painful to analyze
<DocScrutinizer05>
had that a half a dozen times, in a way that wasn't recoverable
<nox->
yeah if you dont remember to print once in a while a laser might be better for you than an inkjet
<DocScrutinizer05>
allegedly removing head and cleaning it half an hour in some mix of water and anthoer stuff shall help sometimes
<wpwrak>
it's the sort of fun that made me jump eagerly to laser printers as soon as they became affordable for home use
<freemangordon>
wpwrak: wanna share the code I use to test it?
<jonwil>
color laser printers are still far too expensive
<jonwil>
and I need color printing
<freemangordon>
wpwrak: you need n900 though
<freemangordon>
and (maybe) QtCreator
<DocScrutinizer05>
I guess you didn't print (black) for some extended period of time
<wpwrak>
freemangordon: naw, i'm not dirtying my hands with poettering ;-) besides, i like just being the smartass :)
<freemangordon>
ah, ok
* wpwrak
goes to wash his eyes now
<jonwil>
anyhow, I will probably see if I can find canon tech support somewhere that might have an answer for me but right now I just need to get this pdf file to print with the text visible
<DocScrutinizer05>
that code is unrelated to poettering
<wpwrak>
jonwil: i thought they were pretty affordable these days ? maybe not the greatest, though.
<freemangordon>
BTW that int32 comes form debug symbols, it is not something me and jonwil have invented
<freemangordon>
*from
<freemangordon>
DocScrutinizer05: :nod:
<wpwrak>
oh, that's RE'ed stuff ? i see
* DocScrutinizer05
suspects that somebody doesn't dare to try and reassemble a N900 kit ;-P
<wpwrak>
at the moment, in trying to work up my courage to disassemble my mill. because the Z drive is making noise. probably a tightening issue cause by too many tiny movements. already had to re-tension a wire to solve the same sort of issue in the X drive
<jonwil>
Compared to what an injket multifunction (print/scan/copy) costs around here, a color laser (just print no scan/copy) still costs a lot
<jonwil>
Not that I print a lot of stuff
<DocScrutinizer05>
that's the problem
<wpwrak>
jonwil: okay, that's true. but they went from the cost of a car to the cost of a cheap laptop, or similar
<DocScrutinizer05>
you need to print one page every week with a inkjet
<jonwil>
this printer (a Canon) has worked just fine without problems before even when I didn't print lots
<jonwil>
and it prints color just fine
<DocScrutinizer05>
yeah, it's not a 100% reliable effect
<jonwil>
it just wont print black
<DocScrutinizer05>
depends on mood, moonphase, air humidity a lot, timespan since last print, whatnot else
<DocScrutinizer05>
but it happened 2 dozen times to me, half a dozen unrecoverable
<jonwil>
well it HAS been quite humid around here this week
<DocScrutinizer05>
and yes, usually only one nozzle
<DocScrutinizer05>
doesn't matter, maybe it been salty
<DocScrutinizer05>
or high ozone
<DocScrutinizer05>
or a fly crapped into your printer
<DocScrutinizer05>
is canon disposable printhead (integrated into cartridge) or long lasting piezo printhead?
<jonwil>
Its a separate part to the cartridge
<jonwil>
and also to the printer
<jonwil>
I took the head out before and I got black ink on my finger when I touched the bit that the cartridge touches so ink is going from the cartridge to the print head
<DocScrutinizer05>
actually pressing head nozzles on a wet badge of tissues is a brilliant idea, it creates quite some water pressure on the nozzles
<DocScrutinizer05>
the faster and harder you press, prolly the better the result
<DocScrutinizer05>
also better don't use a bubbling waterjet, rather a plain jet from e.g. a hose or whatever
<DocScrutinizer05>
or your shower head
<DocScrutinizer05>
bubbles in water make the jet "soft"
<jonwil>
What I think I will do is to finish fixing this pdf, print it so it prints the text using the color cartridges then when I am out doing stuff I will look for some cleaning stuff (I dont trust the water in my house not to leave things picked up from the old pipes in my old apartment inside the print head)
<jonwil>
worst comes to worst, it looks like you CAN buy replacement print heads for my model
<jonwil>
Not sure I would trust one from fleabay or alibaba though
<jonwil>
would probably contact Canon and ask if they can point me at a legitimate supplier
<DocScrutinizer05>
dman you konversation! eating 10% of my 4core@3.4GHz
<DocScrutinizer05>
BWAHAHA kontact crashed
DocScrutinizer05 has quit [Quit: EEEEEEK]
DocScrutinizer05 has joined #neo900
<wpwrak>
indeed. ultrasonic cleaners are amazing. never worry about flux residues.
<wpwrak>
nox-: you can get demineralized water from car supplies. just make sure they didn't mix any "enhancements" into it. (but that's usually clearly indicated on the bottle)
<DocScrutinizer05>
and way more expensive ;-P
<wpwrak>
yeah, for 1 l the difference is small (+30%) but for 5 l distilled vs. anti-corrosive have a difference of some +130%
<wpwrak>
anyway, 5 l are a bit less than USD 3 here.
arcean has quit [Read error: Connection reset by peer]