azonenberg changed the topic of #homecmos to: Homebrew CMOS and MEMS foundry design | Wiki: http://homecmos.drawersteak.com/wiki/Main_Page | Repository: http://code.google.com/p/homecmos/ | Logs: http://en.qi-hardware.com/homecmos-logs/
NullMoogleCable has quit [Ping timeout: 256 seconds]
NullMoogleCable has joined #homecmos
CaptHindsight has left #homecmos ["not here"]
CaptHindsight has joined #homecmos
<CaptHindsight> I just read the logs. I though that this channel was more active
<azonenberg> CaptHindsight: yeah its slow
<azonenberg> i'm finishing up a phd and havent had much time to do any work
<CaptHindsight> I'll fill in some blanks
<CaptHindsight> I'm working on some other projects now myself
<CaptHindsight> but this is my day job
<CaptHindsight> home cmos is asking a lot from DIYers
<CaptHindsight> lots of steps, alignment, vacuum chambers etc
<CaptHindsight> printing organic semiconductors might be a little easier for home
<azonenberg> Yeah its nontrivial
<azonenberg> at least for starters i want to work on MEMS
<azonenberg> My initial target is hackerspaces doing this as a team effort
<azonenberg> a sufficiently crazy and motivated one-man operation could do it, but nontrivially
<CaptHindsight> MEMS with just silicon?
<azonenberg> The initial goal would be a comb drive
<azonenberg> should be possible with silicon, glass, and metalization
<azonenberg> no transistors, off-die driver logic
mrdata has quit [Read error: Connection reset by peer]
mrdata has joined #homecmos
<Noxz> MEMS, or more specifically micro systems (non eletrical) is why I am here.. trying to make the smallest mechanical watch movement in the world ;)
<Noxz> currently my spare time is working on the CastAR (from jeri ellsworth) devkit.. porting a classic macintosh game that I reversed from M68k asm
<Noxz> I actually demo'd what I had for them on Friday
<CaptHindsight> so more MEMS interest than IC's
<Noxz> the current paper I am sifting through is: http://bdml.stanford.edu/twiki/pub/Main/PerchingPapers/bistable_beams.pdf
<mrdata> how small can magnetometers be?
<Noxz> magnetoresistive method seems to be tiny, from a quick search
<mrdata> okay, just how tiny are these?
<Noxz> well, it appears to be a meandering strip of NiFe
<Noxz> and you compare that against some really accurate resistors, most likely in some wheatstone bridge config like strain meters?
<mrdata> perhaps
<mrdata> but, i would want a large array of these
<Noxz> define: large
<mrdata> um, like CCD pixels
<Noxz> well, you still need to compare the resistance against a known resistance
<CaptHindsight> I imagine a single iron atom could be the smallest magnetometer
<CaptHindsight> well core anyway
<CaptHindsight> Ultrasensitive magnetometer using a single atom
<azonenberg> CaptHindsight: i would love to make CMOS eventually
<azonenberg> it just has tighter tolerances than MEMS
<azonenberg> in terms of chemical purity, trace metals, etc
<azonenberg> mems you mostly just care about particles
<azonenberg> my plan is to develop the process on mems initially
<Noxz> I forget, but you are NOT using su-8, right azonenberg ?
<azonenberg> Correct
<azonenberg> In the past i have used resist from injectorall.com
<Noxz> and is there a rule about DIY does not use expensive equipment? for example, a piezo platform for movement
<azonenberg> I am more than willing to use higher end gear
<azonenberg> I just want to do it only where necessary
<Noxz> roger
<Noxz> ughf, time to test this C# code
<Noxz> the plus side is the CastAR is super cool
<Noxz> nice when things work the way you imagined, huh?
chris_99 has joined #homecmos
NullMoogleCable has quit [Ping timeout: 246 seconds]
NullMoogleCable has joined #homecmos
<CaptHindsight> linear servos are often cheap on ebay. If you have a cnc shop you can build a positioner with 1um accuracy on the cheap.
<CaptHindsight> I mostly make my own materials and resists
<CaptHindsight> using that inverse microscope you can cure polymers down to ~1um lines
<CaptHindsight> so you can not only print masks but also functional polymer materials
<Noxz> nice, I just need to find time to get back into this.. have a million other things to do
<CaptHindsight> I'm going to start posting info at http://openlunchbox.com/smf/index.php
<Noxz> gotta catch the train to work..
<_Sync_> CaptHindsight: I built a stage that acheives that, the problem is the flattness over the area
<CaptHindsight> _Sync_: what was the work envelope?
<_Sync_> 50x50mm
<CaptHindsight> 3 axis?
<_Sync_> no, the 3rd axis doesn't have to move, although I could rig something up
<_Sync_> the thing is, that my stages have an average flattness of 400nm
<_Sync_> which in the real world translates to a high spot of around 600nm in one place
<_Sync_> but that is enough to defocus the laser
<_Sync_> using S1813 resist
<_Sync_> which is annoying because you need HMDS surface prep or it doesn't stick at all
<CaptHindsight> I make my own resists
<CaptHindsight> why the Z runout?
<CaptHindsight> bearing runout or does does platform need more polishing?
<_Sync_> it's not runout, it's just the parallelism of the surfaces
<_Sync_> which is still pretty good
<_Sync_> not much you can do about
<_Sync_> well, making your own resists has the problem of metal ion contamination
<_Sync_> and inconsistency
<chris_99> can you actually buy the inks for that now?
<CaptHindsight> I didn't look at their exact details but they should be available, if not let me know
<CaptHindsight> Silver Conductor (C2131014D3, 100m /sq) make your own
<chris_99> i'm on https://embodied.mpi-inf.mpg.de/research/printscreen/ they don't seem to be selling them at the moment
<CaptHindsight> ohh I should beat them to a kickstarter
<chris_99> heh, have you made some?
<CaptHindsight> what they used: minescent display: Silver Conductor (C2131014D3, 100m /sq), a barium titanate based dielectric (D2070209P6), phosphor (e.g. C2101125P4) and a PEDOT based translucent conductor (D2070209P6, 500-700 /sq)
<CaptHindsight> chris_99: I make materials and the deposition equipment
<_Sync_> http://sync-hv.de/~tmp/forward_logscale_dead.png what a little bit of metal ion contamination does :3
<CaptHindsight> some I open source
<chris_99> cool CaptHindsight :)
<CaptHindsight> use Linuxcnc for motion and process control
<CaptHindsight> the problem they will have is getting reliable jetting from the Canon TIJ cartridges
<CaptHindsight> first they have to flush the ink cartridges since Canon doesn't sell virgin cartridges
<CaptHindsight> and inkjet inks even for industrial piezo heads is tricky, let alone getting solids through TIJ
<chris_99> oh so it'd be harder to do it using Epson printers?
<CaptHindsight> yes and no
<chris_99> i've got one i'm planning on ripping appart for the piezo pumps
<CaptHindsight> the Epson heads are piezo but the ink chambers are not friendly to solids, there is no way to circulate fluid...
<chris_99> ah
<CaptHindsight> so you need very stable fluids
<CaptHindsight> you can fill, print and then flush
<CaptHindsight> if you keep the fluid moving it will help
<CaptHindsight> don't stop printing
<chris_99> heh
<CaptHindsight> with TIJ if they clog you just toss them and use a new one
<CaptHindsight> but you're limited to ~3cps max for viscosity
<CaptHindsight> maybe an ink kit with synthetic paper and polyester film
<CaptHindsight> might be best with a printer as well with all the mods
<CaptHindsight> you could have 4 fluids and print 4 layers with a single pass desktop inkjet
<chris_99> that'd be awesome
<CaptHindsight> since the printer can't back up and register itself again for more passes
<CaptHindsight> their EL has a driver of up to 220Vpp
<CaptHindsight> and material safety is another issue
<CaptHindsight> get people to wear gloves and goggles
<chris_99> are all those chemicals you need easily available, or do you need to buy in bulk
<CaptHindsight> well they always want you to buy in bulk (~200kg drums)
<CaptHindsight> sometimes you can get 20kg pails
<CaptHindsight> or bags if powder
<CaptHindsight> https://www.sigmaaldrich.com/ carries quite a bit but at 10-100x the usual price
<CaptHindsight> 500ML Isopropyl alcohol $49.70
<CaptHindsight> meets USP testing specifications
<CaptHindsight> lets see about the chems in their secret sauce above
<chris_99> $49 seems high for isopropyl doesn't it
<chris_99> neat!
<chris_99> is that basically the same set of chemicals
<CaptHindsight> yes, and a few other kits
<chris_99> how much does it cost from there? i can't see any price
<CaptHindsight> I bet that kits is >$1k
<CaptHindsight> you have to ask for an invoice
<chris_99> ah
<CaptHindsight> if you want to sceen print then it's cheaper and easier
<CaptHindsight> you can make your own silver conductive ink
<chris_99> interesting :)
<CaptHindsight> I don't want to go too far outside the realm of this channel
<CaptHindsight> if the focus is on MEMS and CMOS
<CaptHindsight> maybe move the printed electronics to #openlunchbox
NullMoogleCable has quit [Ping timeout: 250 seconds]
SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
SpeedEvil has joined #homecmos
chris_99 has quit [Quit: Ex-Chat]
NullMoogleCable has joined #homecmos