#homecmos IRC log for Monday, 2011-10-24

azonenberg_workso i'm running through the comb drive process now17:45
azonenberg_workjust did top side hardmaskk evaporation, waiting for vacuum so i can do bottom17:45
azonenberg_workthings are looking good so far17:46
soul-dk what step that  be ?18:01
soul-dalso  maybe a note that   first you talk  front  and back then  change to  use top bottom18:02
azonenbergsoul-d: It pretty much depends on which way you are holding the wafer :p18:49
azonenbergI just did 3 and 5, 4 can be done before or after 5 so i havent done it yet18:50
azonenbergthat just involves dipping the wafer in acetone18:50
azonenbergWhich has to wait until i clean out my dirty classware ;p18:51
azonenbergglassware*18:51
soul-dwich tools are a must for this hobby btw   like you made somthing out of a microscope optics?  ( egg same  idea als those old negative pictures used for enlarging )19:06
azonenbergThe first thing to do is get a working contact lithography process set up at 200um feature sizes on prnited circuit boards19:07
azonenbergIf you use pre-coated PCB you can rule out variability in the photoresist as a source of problems19:07
azonenbergOnce you have that working, try spin-coating blank board and make sure that works19:08
azonenbergthen get a microscope and build a projection exposure rig, try doing exposures onto precoated PCB and then spin coated19:08
azonenbergAt this point you can start thinking about more complex processes like the ones i'm doing19:08
azonenbergOne big thing that has to be worked out is evaporation19:08
azonenbergHomebrewing an evaporator is possible19:09
azonenbergit's been done19:09
azonenbergbut nobody has published clear plans and drawings that i know of19:09
soul-dyeah was looking for vacuum pumps other day  but you gota watch out  what kinds of results you  get :P19:09
soul-dill browse to my  books19:11
soul-dwell not today  but got plenty of chemistry books need to read on theory of  filament evaporation it was ? first  sofar i could remmber was shematics for electric furnace19:20
azonenbergHmm19:45
azonenbergLooks like the top hardmask didnt deposit right19:45
azonenbergI was a little worried about that tbh, i saw a nice even deposition on the bottom side but the top for some reason didnt work right19:46
azonenbergNot sure why but the top is getting etched19:46
azonenbergGood thing i have another wafer, i'll have to run the process again tomorrow19:46
azonenbergI'm gonna leave this one in KOH for an hour to see what happens19:47
azonenbergI expect it wont be usable though19:47
azonenbergHmm, even the bottom side was weakened20:27
azonenbergsomething definitely went wrong20:28
soul-dthose 2"  expensive ?20:47
azonenberg$3020:47
azonenbergAnnoying but not the end of the world20:48
azonenbergBut i want to figure out what went wrong before trying it again ;p20:48
soul-dsure20:48
soul-donly need to fix my uv light box with a decent timer  and figure out if  this laser printer works well last trials on clear sheets din't go to well20:50
soul-dfor pcb20:50
azonenbergi see20:50
azonenbergLet me write up my lab notes for today, sec20:51
CIA-67homecmos r151 | trunk/ (2 files in 2 dirs) | Today's lab notes20:57
azonenbergSo basically what it looks like happened is that the Cr didnt deposit correctly20:59
azonenbergWhen i was depositing it, i saw a nice dark film on the bell jar near the top but very little at the bottom20:59
azonenbergI'm not sure what could cause a failure mode like that and need to do some reading / talk to some people20:59
azonenbergtop of the jar that is, not the wafer21:01
azonenbergIt looks like for some reason instead of the vapor spreading uniformly around the chamber, most of it went up21:02
soul-don wiki it looked so simple21:09
azonenbergThe deposition normally works fine21:09
azonenbergI probably did somethign stupid21:09
azonenbergall of my past evaporations have been successful21:09
azonenbergI think i got impatient and turned the temperature up too high21:09
azonenbergBecause, as I found out earlier today21:15
azonenbergYou don't evaporate chrome by melting it21:15
azonenbergit sublimes from the solid state21:15
azonenbergI kept turning it up higher because i expected it to melt21:15
soul-dtricky , reading your log forgot chat21:32
azonenberglol21:33
azonenbergAnyway so i am going to try and do it again tomorrow21:34
azonenbergon my last unused wafer21:34
azonenbergafter doing the contact mask tonight21:34
soul-dbought the final piece of equipment  since i went in  buying spree and stocked up on lots of smd stuff  bought a solder station to go with it  should help me get trough the winter  should start on some prototypes soon  also wanted to get one made sooner or later21:44
soul-dplenty of relativly cheap  pcb makers online21:45
azonenbergvery nice21:46
soul-dit hurts my wallet  though21:47
soul-dbut that's what hobby's are for i guess21:50
azonenbergLol, yep21:52
XgFMy friend dropped over £120 on cable connectors today... yikes...22:01
azonenbergo_O22:02
soul-dlol sounds like me  almost  but not that crazy22:02
XgF...Neutrik are expensive22:02
soul-ddid buy a lvds set for 13 eur :P22:02
soul-dwich i find expensive for connecting stuff22:03
XgFThe danger of working in the entertainments industry: You start acquiring your own cables. You realise how expensive the damn things are22:03
azonenbergLol22:04
XgFImpedance match cable gets expensive when you purchase it in non-negligible lengths in non-install forms22:05
XgF....Especially when it has to be designed so that it won't break the first time some monkey rolls over it with an 130kg flightcase22:09
Action: azonenberg wonders why most of that can't be run over ethernet / wifi as digital22:10
azonenbergthen going to analog right when it hits the speaker etc22:10
XgFThis is impedance matched DMX (2 core twisted pair + shield) cable for lighting. Fixtures don't take ethernet directly (yet), namely because having dozens of ethernet switches causes latency issues22:11
azonenbergDozens of switches?22:11
azonenbergWhat about a single 48 or 96 port22:11
XgFThats fine until you need to get 16 ethernet cables onto a truss...22:12
XgFOr a switch, which causes its own logistical issues, and also ethernet is more expensive that DMX22:12
XgFLots of venues will trunk multiple DMX universes over an ethernet cable from front of house to the stage and then break it out though22:14
--- Tue Oct 25 201100:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!