Topic for #homecmos is now Homebrew CMOS and MEMS foundry design | http://code.google.com/p/homecmos/ | Logs: http://en.qi-hardware.com/homecmos-logs/
ashaw has joined #homecmos
<ashaw> azonenberg, are you there?
<azonenberg> ashaw: back
<azonenberg> Was on campus
<ashaw> no probs.
<azonenberg> Whats up?
<ashaw> I am thinking about using LDI to image the surface of an IC.
<azonenberg> As in you want to do laser direct-write lithography for making an IC?
<azonenberg> Very feasible as long as you can align it, which is tricky
<azonenberg> I'd suggest doing it for mask fab and then contact alignment of the mask
<ashaw> I have been thinking, why not use a UV laser and have a Blu-ray like controll
<azonenberg> Well, the hard part is how to tell where you're pointing the laser
<azonenberg> before firing it and exposing whatever is under it
cheater_ has joined #homecmos
<ashaw> not really
<azonenberg> you suggest just doing a low-power pulse in a corner?
<azonenberg> In any case, i think it's very doable
<azonenberg> i seriously planned to do the exact same thing myself
<azonenberg> just been too busy to make it happen
<ashaw> no - I suggest using an electron beam to put tracking marks on the edge and then plating them with a metal
<azonenberg> If you have e-beam
<azonenberg> why not use ebeam for everything?
<ashaw> I suppose
<azonenberg> Since ebeam is pretty easy to use for low-power imaging and high-power litho
<azonenberg> Turning a SEM into an e-beam direct write system is a solved problem
<azonenberg> several companies sell conversion kits
<azonenberg> they can even still image
<ashaw> can you use a STM to image?
<azonenberg> Yes, but i dont know if you can make one do lithography
<azonenberg> and scanning probe techniques are much slower than scanning-beam systems like SEM
<ashaw> I know.
<azonenberg> I mean, there are so many tools i eventually want to have in my lab lol
<ashaw> as I
<azonenberg> its all a matter of deciding which ones to build/buy first
<ashaw> yeah
<azonenberg> Now that i have Ethernet working on PIC32s i can make networked lab equipment
<azonenberg> I'm planning to build an ethernet-controllable spin coater in the fairly near future
<azonenberg> gotta pick out a good motor and do some CAD work
<ashaw> how about using a single pulse on one corner to create the beam
<azonenberg> To do alighnment you'd need at least two marks
<ashaw> sorry the tracking generator
<ashaw> no.
<azonenberg> you need to know scale, theta, and position
<ashaw> I would not
<azonenberg> no way you can get that without at least two
<ashaw> I just need one
<ashaw> and then rotate it
<azonenberg> not sure i understand what you're planning
<azonenberg> in any case, the easiest technique IMO
<azonenberg> is to make the laser direct write be unaligned
<azonenberg> just write onto a mask blank
<azonenberg> develop and etch the mask (probably sputtered metal on a microscope slide)
<azonenberg> Then use the resulting chrome-on-glass mask in a contact aligner, which would be pretty easy to build
<ashaw> Interesting.
<ashaw> however, I still need scale
<azonenberg> I'd do a single initial calibration
<azonenberg> and then just assume it's accurate
<azonenberg> recalibrate periodically
<azonenberg> as in, write two lines X distance apart
<azonenberg> where X is nominally, say, 100um
<ashaw> So what do we know about the disk?
<azonenberg> then take a photo with a calibrated camera and measure
<ashaw> we know it's rotation velocity
<azonenberg> Bluray disks are polar coordinates
<azonenberg> and have tracking metadata already etched into them
<azonenberg> this will be cartesian
<azonenberg> on a blank substrate
<ashaw> no, I will write in polar
<azonenberg> Oh
<azonenberg> Mine will be cartesian
<azonenberg> i envision reusing the diode but otherwise doing totally my own structure
<ashaw> What type?
<ashaw> I too will only use the diode, I do not think the structure can be controlled well enough
<ashaw> My intent is to use a 2in wafer, spinning at 10k rpm
<ashaw> and image using a set of fibe beams with a high NA
<ashaw> lenses on the end.
<ashaw> spaced on the axis of rotation 100 um apart
<ashaw> approx.
<ashaw> azonenberg
<ashaw> ?
fenn has joined #homecmos
<kanzure> patron saint of lithographers?
<ashaw> hey what?
<kanzure> just reading the backlog
<ashaw> who is?
<kanzure> /kick ashaw
<ashaw> why?
<azonenberg> Back
<azonenberg> So let me see if i understand your plan
<azonenberg> 10k rpm is pretty high, for starters
<ashaw> not really.
<azonenberg> So you want to have a single diode going into a set of optical fibers?
<ashaw> no, multiple diodes
<azonenberg> That's a LOT of power
<kanzure> in molecular biology labs you regularly get up to >50k rpm
<azonenberg> for a silicon wafer, i mean
<azonenberg> most spin coaters run at like 3-5k
<azonenberg> But for a 2" it should be fine
<azonenberg> i'd be cautious about an 8" or larger that fast
<ashaw> yeah
<azonenberg> anyway so ONE bluray diode is overpowered for litho
<ashaw> I mainly want a high linear velocity
<azonenberg> they're meant to actually heat and burn the metal on the disk
<ashaw> I will only run it a very low power
<azonenberg> You'll have to
<ashaw> it is more for control
<azonenberg> otherwise you'll overexpose hugely and heat it
<ashaw> I want to controll each laser seperately
<azonenberg> My thinking was to have a single diode
<azonenberg> x/y mirror for fine alignment
<azonenberg> (within one die)
<azonenberg> then a mechanical stage for die-to-die stepping
<ashaw> x/y mirror? how are you moving it?
<azonenberg> Open question, probably light show galvos
<azonenberg> then feed the resulting pattern through a reducing lens
<azonenberg> I'm far from sure of what i'm doing here
<fenn> light show galvos have terrible spatial resolution
<azonenberg> fenn: 100um would be enough
<azonenberg> if i was able to then reduce optically
<ashaw> you cannot.
<ashaw> you have a problem
<azonenberg> In any case, i had some other ideas
<azonenberg> one was to bounce the beam off a DLP micromirror array
<ashaw> You need a high numerical aperature
<ashaw> That was my first Idea
<fenn> i think you want to move the diode image around with flying optics, if you're going the continuous (vector drawing so to speak) way
<azonenberg> fenn: Again, open question
<fenn> DLP's are great, A+
<azonenberg> this is on the tools-to-build-eventually list
<azonenberg> not on the tools-under-current-development list
<fenn> also you can fake it with LCD's if you don't mind waiting a long time
<ashaw> What tools are you building currently
<azonenberg> I just crossed "fume hood" off my list
<azonenberg> "proper spin coater" is next up
<azonenberg> as in, one that isnt made of 2x4s and filling my entire hood
<azonenberg> this one will be much smaller, mostly metal construction
<azonenberg> and have proper feedback-loop speed control
<fenn> use a juiceman jr.
<azonenberg> i want it controllable over ethernet
<azonenberg> the plan is for all of my tools to be networked
<azonenberg> so i can download process recipes from my laptop onto it
<azonenberg> and then just click "run"
<fenn> i had the same dream... once
<azonenberg> I already have a PIC32 receiving UDP packets and decoding them
<kanzure> hey wait.
<kanzure> azonenberg: http://gnusha.org/skdb/
<azonenberg> and i know how to generate PWM for motor driving
<fenn> PWM is easy compared to decoding ethernet
<azonenberg> fenn: http://pastebin.com/fbWrjrdA
<azonenberg> that was a day or two ago
<fenn> most microcontrollers have PWM hardware built in
<azonenberg> The "read ethernet" is a solved problem
<azonenberg> i just have to put that and a PWM controller on one board
<azonenberg> add some kind of optical encoder for speed feedback
<azonenberg> and do mechanical design for the chuck etc
<kanzure> no really, check the link
<azonenberg> i did
<ashaw> so after the spin coater, perhaps a CMP machine?
<kanzure> ok my work here is done
<azonenberg> ashaw: contact mask aligner is likely to come first
<Sync> closed loop speed control seems to be overkill, you don't have the load changing around alot
<azonenberg> Sync: but i need precise ramp rate contorl
<azonenberg> i need constant accel/decel
<azonenberg> and without knowing the mass of the load thats hard unless you have closed loop feedback
<ashaw> are we going to work on whole wafers here.\
<azonenberg> in any case a CMP system would be a pretty simple mod of the spin coater
<azonenberg> ashaw: I'd like to eventually but for now most of my work is done at the die level
<Sync> well, you know the mass pretty well
<azonenberg> since wafers are pricey
<azonenberg> ashaw: so contact mask aligner is high on the todo list
<kanzure> why not dlp
<azonenberg> in the longer term a stepper
<ashaw> how much are wafers?
<ashaw> The problem kanzure is the optics.
<azonenberg> ashaw: on the order of $25-50 depending on size, one vs two side polish, etc
<ashaw> how much is unpolished?
<azonenberg> Dirt cheap, i got an entire boat of 25x solar grade unpolished from ebay for $50
<ashaw> as wafers?
<azonenberg> Yes
<ashaw> what size
<ashaw> ?
<azonenberg> 4 inch N type
<ashaw> defect level?
<azonenberg> thats the spec
<azonenberg> I bought a cassette from that seller
<azonenberg> i've used two or three so far
<azonenberg> as mechanical dummies for testing spin coating
<azonenberg> as "cookie sheets" for heating single dies in the oven
<azonenberg> etc
<ashaw> Why not make a CMP machine as a top priority then
<azonenberg> Because i usually work with single dies, not full wafers, so cost of silicon is minimal
<azonenberg> that is enough for like a hundred experiments
<azonenberg> most of my dies are 2-3mm square
<azonenberg> If i were to do CMP, it'd be for a different reason
<azonenberg> namely, copper damascene
<ashaw> I think CMP is going to really make things easier
<azonenberg> ashaw: it's a valuable process to develop, i dont deny that
<azonenberg> I just dont think its top priority
<ashaw> perhaps.
<azonenberg> It depends, in part, on what you're trying to do
<azonenberg> for starters
<azonenberg> CMOS or MEMS?
<azonenberg> My current process is targeting MEMS
<ashaw> I think that with a Doubled BD-R laser and spin imaging we could get a 180nm process
<azonenberg> ashaw: I'm very interested lol
<ashaw> I am trying to target CMOS
<azonenberg> My current roadmap goes out to about 750nm
<azonenberg> 20um is the smallest i've done reliably and i've hit 5um experimentally but not with good yield
<ashaw> I want to do it with 10k in equipmemnt
<azonenberg> Well, I like it lol
<azonenberg> The first thing is particle contamination
<azonenberg> at 180nm its nontrivial
<ashaw> what is the process you are having the most problems with
<azonenberg> So far, it's simply that i cannot make a mask small enough
<ashaw> Yeah we will need a whole clean environment.
<azonenberg> i can reduce my 200um mask by 10x = 20um features reliable
<azonenberg> and by 40x = 5um but my 40x objective has a defect on it
<azonenberg> so i cant get good results
<kanzure> what is your current reducing optics like?
<azonenberg> Microscope objective
<azonenberg> the illuminator, though, is a halogen lamp
<azonenberg> as a result the FOV is small and it needs a really long exposure
<azonenberg> Both of those could be fixed if i built a LDI system for making masks
<azonenberg> then used high intensity UV tubes for a contact exposure
<azonenberg> Which is why once i get the spin coater built, LDI and contact aligner are next on the TODO
<azonenberg> fume hood was first as photoresist fumes can be a little nasty
<azonenberg> i have a spin coater but its so big it barely fits in the hood
<azonenberg> and obstructs airflow
<azonenberg> so i get vapors leaking out
<azonenberg> the new one will be a lot smaller
<ashaw> well for mask creation I cannot see a better technique than using a BD-R on a sputter coating of tungsten
<ashaw> and then ablating the tungsten
<azonenberg> why tungsten in particular?
<azonenberg> and why ablation vs lithography (which needs a lower power dose and gives cleaner edges)
<ashaw> nothing, actually any material
<ashaw> no reason at all actually
<ashaw> How long are you prepared to wait for a mask
<azonenberg> A while, even overnight would be fine if the equipment wqasnt expensive to run
<azonenberg> But vibration would be a concern at those feature sizes
<azonenberg> you'd need to find some way of damping it
<ashaw> where are the vibrations coming from?
<azonenberg> Um, everwhere
<azonenberg> people walking by
<azonenberg> cars outside
<azonenberg> HVAC system
<ashaw> on what time scale?
<azonenberg> several Hz
<ashaw> above 100k?
<azonenberg> no, single Hz
<azonenberg> like, maybe 0.1 - 10
<azonenberg> even at 400x with a microscope you can see the image shaking if its not on a realy stable mount
<ashaw> then just compensate the lithography
<azonenberg> and at 1000x when people walk by it's nearly impossible to see anything
<ashaw> Why not suspend the entire aparatus on an air bearing
<ashaw> and use a wireless adaptor
<azonenberg> Again, there are a lot of possibilities
<ashaw> for the signal
<azonenberg> in fact, many commercial anti-vibration tables are based on air bearings
<azonenberg> the SEM at my school's cleanroom uses one for example
<azonenberg> But it adds to the cost and development complexity
<azonenberg> Not saying you cant do it, i'm all for it
<azonenberg> But i also want to point out potential problems in advance
<ashaw> I like that.
<ashaw> allways important
<ashaw> we also need an ultra high purity water generator
<azonenberg> Again i'm doing MEMS, not CMOS, so i've been fine with regular distilled water from a pharmacy
<azonenberg> didnt need to do DI on site
<ashaw> yeah
wolfspraul has joined #homecmos
fpgaminer has joined #homecmos
soul-d has joined #homecmos
cheater_ has joined #homecmos
wolfspraul has joined #homecmos
wolfspraul has joined #homecmos
<cheater_> why are air bearings superior to magnetic bearings?
<cheater_> you'd think a couple of electromagnets would be easy enough to cobble together
<Sync> you do not need advanced control circuits
<ashaw> and there is no possible stable magnetic bearing in a still system
<ashaw> so the bearing is by nature a dynamic system
<Sync> the amd diffusion plant in dresden did some major effort to get the vibration out of their building
<ashaw> it is really important
<ashaw> as at that level micro-vibrations are huge
<Sync> I don't work with that fine structures at the moment
<ashaw> not YET
<ashaw> we'll get there
<ashaw> so talking about the equipment problem.
<Sync> no, I won't, I don't need to :P
<ashaw> Why?
<Sync> I manufacture fets in my sidejob at the university
<Sync> they won't get smaller because they will not work as good then
<ashaw> aaah
<ashaw> hehehe
<ashaw> you will not use smaller microstructures on your fet
<Sync> okay "make" is a bit overstating it
<ashaw> to reduce capacitance and such
<Sync> I yell at the ion implanter and hope it does what I want
<Sync> this is not an issue
<ashaw> kihihihi
<Sync> we develop ion sensitive fets
<Sync> for measuring pH values
<ashaw> oh cool
<ashaw> which uni might I ask?
<Sync> hannover, germany
<ashaw> oh cool
<ashaw> are you a stdent or a resercher?//
<Sync> I'm a student
<ashaw> Bachelor, Masters or PhD?
<Sync> first semester bachelor ;)
<ashaw> hehe, Just finished my 1st year
wolfspraul has joined #homecmos
wolfspraul has joined #homecmos
wolfspraul has joined #homecmos
wolfspraul has joined #homecmos
bitr has joined #homecmos
azonenberg has joined #homecmos
lekernel_ has joined #homecmos