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/
xiangfu has joined #homecmos
soul-d has quit [Remote host closed the connection]
xiangfu has quit [Ping timeout: 246 seconds]
xiangfu has joined #homecmos
icantdance has joined #homecmos
icantdance has quit [Quit: We be chillin - IceChat style]
xiangfu has quit [Ping timeout: 256 seconds]
xiangfu has joined #homecmos
Sync__ has quit [Ping timeout: 260 seconds]
Sync_ has joined #homecmos
_whitelogger has joined #homecmos
_whitelogger_ has joined #homecmos
_whitelogger_ has joined #homecmos
_whitelogger_ has quit [Remote host closed the connection]
drygol has quit [Ping timeout: 240 seconds]
drygol has joined #homecmos
Charlie has quit [Ping timeout: 260 seconds]
berndj has quit [Ping timeout: 255 seconds]
Charlie has joined #homecmos
berndj has joined #homecmos
B0101 has joined #homecmos
xiangfu has quit [Ping timeout: 240 seconds]
xiangfu has joined #homecmos
xiangfu has quit [Ping timeout: 248 seconds]
xiangfu has joined #homecmos
<B0101> +q45
B0101 has quit [Quit: Ex-Chat]
xiangfu has quit [Ping timeout: 245 seconds]
xiangfu has joined #homecmos
B0101 has joined #homecmos
xiangfu has quit [Ping timeout: 245 seconds]
xiangfu has joined #homecmos
qi-bot0 has joined #homecmos
B0101 has quit [*.net *.split]
qi-bot has quit [*.net *.split]
qi-bot0 is now known as qi-bot
B0101 has joined #homecmos
diginet has joined #homecmos
diginet has quit [Quit: Leaving]
glowplug has quit [Remote host closed the connection]
glowplug has joined #homecmos
glowplug` has joined #homecmos
glowplug has quit [Ping timeout: 255 seconds]
<glowplug`> Good afternoon all.
<Sync_> hello
<Sync_> glowplug`: thought about the direct litho for a bit
B0101 has quit [Ping timeout: 258 seconds]
<Sync_> I still think an interferometer is the cheapest and best way to do the position measurement
<Sync_> the resolution/cost factor is very good
<glowplug`> I'm working on the interferometer. Came to that conclusion as well.
<glowplug`> Will be constructing a simple michealson and a bunch of micrometers to position the sample. What was your conclusion with the direct litho?
<Sync_> there are 100nm resolution glass scales, which would be cool
<Sync_> yeah a michelson is easy to do
<glowplug`> It would be, but I can probably get 250nm accuracy with the cheap interferometer for pennies.
<Sync_> I should head to the scrapyard and see what machines they currently rip apart
<glowplug`> What are your opinions with the direct litho vs mask
<Sync_> well
<Sync_> direct litho is probably fine for experimental work
<glowplug`> At most universities from what I understand they use masks and spinning each layer with resist.
<Sync_> yup
<Sync_> it's cheaper if you have the equipment
<glowplug`> I think that will be fairly easily to accomplish.
<Sync_> and for most experiments you only use a single mask set
<Sync_> hmm actually, reading the paper again, I almost have everything I'd need
<Sync_> I'd only have to get some mirrors for the interferometers
<Sync_> and somehow build an xy stage
<glowplug`> If you give me a month I can send you an interferometer.
<glowplug`> I know thats quite awhile.
<glowplug`> Have a lot of personal crap going on right now thats keeping me from making progress.
<glowplug`> Do you have micrometers? Spinny doohickie? (technical term).
<Sync_> well I have a few for measuring stuff
<Sync_> I have a few laser xy stages
<glowplug`> We need a table with x y tilt and pan adjustments with micrometers.
<glowplug`> I can build one.. but I need time.
<Sync_> I'd just need to order stuff from edmund or thorlabs
<glowplug`> That will cost you a leg + arm.
<Sync_> well mirrors are cheap
<glowplug`> I can build the table and the interferometer for ~$50 total. Depending on how much strugle I have putting together the tiny gearboxes.
<glowplug`> I can use my CNC to cut 50/50 glass for 30 cents per 10mm square. Its a pain in the ass to cut glass but I can do it.
<glowplug`> Add cute little acrylic stands and a dollor store diode you have the worlds cheapest interferometer. Haha
<glowplug`> There are a few more things I haven't worked out
<Sync_> let me just look what thorlabs has for me
<glowplug`> Alignment markers. Is a cheap 100x microscope and a point and shoot camera sufficient?
<Sync_> oh, that's an interesting problem
<Sync_> but yeah probably
<glowplug`> The "actual" 100x microscopes that are ~$25 are probably unsuitable.
<glowplug`> We need to find the actual optics. Seperately.
<glowplug`> Then rig the point and shoot to the optics. This is pretty common stuff for telescope builder types. Do you know anyone that does that stuf?
<Sync_> well the actual optics sould be a microscope lens
<glowplug`> Right. I've never purchased anything like that.
<Sync_> one could just make a setup that you put on the lens to aling
<Sync_> and then take off
<glowplug`> But if I know its dimensions I should be able to get a ghetto mount for it that we could slap the point and shoot on.
<glowplug`> Then we have a cute little screen to align our markers.
<glowplug`> So the next. It's basically unaffordable to build a cartesian machine with 1um repeatable accuracy. It just wont happen.
<Sync_> hm, thorlabs has vis beamsplitter cubes at 135€
<glowplug`> I can build a beamsplitter for 30 cents.
<glowplug`> Thorlabs can jump off a cliff. Haha
<Sync_> or budget plates for 25
<Sync_> well I'm used to order from them
<glowplug`> Like I said if you can wait a month I can send you the stuff for free.
<glowplug`> Also there are beamsplitters in cdrom drives which is a low cost option that doesnt require waiting for me to setup my CNC for glass.
<glowplug`> You can probably find a pile of cdroms for free or <$5
<glowplug`> So the CNC thing. I'm not confident I can do it. So I've been trying to find an alternative.
<soul-d> indeed telescope work bit diferent but you might get some intresting tools
<glowplug`> I started thinking about laser projectors. Galvos aren't really suitable for exposing resist. They are super fast and not accurate enough.
<soul-d> google for ATM in full Amature telescope making
<soul-d> i assume theory behind the light stuff is same just have to do different calculations
<glowplug`> For the galvos?
<glowplug`> By the way did you just change names? Haha
<Sync_> yeah beamsplitters are not too bad commercially
<soul-d> no just another someone :P
<glowplug`> At any rate. Imagine a laser projector configuration. Two motors and two mirrors. The motors are microstepping servos with a 500/1 gearbox reduction.
<glowplug`> Oh I see. Haha
<glowplug`> So basically its the worlds worste laser projector that can draw a blue dot on a wall very very slowly and not make pictures at all.
<glowplug`> But it can expose photoresist with 1 micron repeat accuracy. =)
<glowplug`> I'm sorry. Microstepping steppers.
<azonenberg> glowplug`: i considered that
<azonenberg> I think cartesian mounting is a better option
<azonenberg> even 10um accuracy is good enough
<azonenberg> Bear in mind you only need 1cm or so of travel
<azonenberg> don't even think about that kind of accuracy across a large field
<Sync_> it's not that hard over a large area
<glowplug`> This thing might have a large field but it wouldn't be an extra cost if it did.
<glowplug`> Cartesian adds all sorts of costs that the projector configuration avoids.
<glowplug`> With no rotational to linear motion 90% of the cost is removed.
<glowplug`> No ball screws, ball nuts, no linear encoder ect.
<azonenberg> Maskless litho is great for making masks, but difficult to align to an existing pattern
<azonenberg> I still think your best bet is to use direct write for making masks, then a contact aligner
<glowplug`> That is the plan for sure. Masks make everything very easy.
<azonenberg> With a contact aligner you just have a little track above the mask assembly
<azonenberg> microscope slides in, tweak the micrometer knobs until the mask is lined up with the wafer
<glowplug`> At any rate. Let me attempt the projector configuration with 10um as an accuracy target. If it works I will make and send out masks. If I fail then I waste a bunch of time. Haha
<azonenberg> then slide it out and bring the UV lamp in
<glowplug`> Except we are going to have an awesome color LCD for alignment because thats what badasses do.
<Sync_> hmm, the new blue lasers sure made stuff easier
<Sync_> no need to drag the ALC909 out
<glowplug`> Yeah I'm just going to grab a blu-ray diode and call it.
<glowplug`> I even get a free beamsplitter from the drive.
<Sync_> I like my tankless water heater with operation light
<glowplug`> So does anyone have a plan for the spinny thingy?
<Sync_> yes
<Sync_> after I finish my exams I'll try to come up with something
<glowplug`> Alright we are counting on you. No pressure. =P
<Sync_> other option is just to buy a spincoater
<Sync_> they are not too terribly expensive
<glowplug`> Oh and if I fail with the projector configuration then I will be building a SCARA robot to try and replace it. If that fails then I don't know. Trust me it wont fail. =)
<glowplug`> The way I determine if I DIY something is if the manufacturer of that thing doesn't have prices on their main page.
<glowplug`> =P
<glowplug`> I think spincoater meets that requirement. Haha
<glowplug`> It looks like the "bio" people do spincoaters.
<azonenberg> Building a spin coater isnt that har
<azonenberg> hard*
<azonenberg> i've homebrewed one already but it's too bulky
<azonenberg> i'm going to try and shrink it using a BLDC from an RC airplane
<azonenberg> will post plans once i finish
<glowplug`> What RPM does the coater run at and what accuracy +/- do you need?
<azonenberg> What RPM, no clue
<Sync_> haha that's interesting
<azonenberg> accuracy i've got, terrible
<azonenberg> accuracy i'd like, better
<azonenberg> My current coater is a flat plate on the end of an electric drill :p
<glowplug`> Interesting. Thats more than I can spend though. Haha
<Sync_> well it's clearly not worth that
<glowplug`> There is an open source BLDC controller now. That will give you the option to use an optical encoder and get extremely precise RPM control.
<Sync_> it's just a bldc
<glowplug`> If you need ~1% then an ESC won't work.
<Sync_> there are enough opensource bldc controllers since a few years
<azonenberg> Yeah
<glowplug`> They are all piece of shit dead ends though.
<glowplug`> The only one thats worth a damn, with decent code. Actually let me get the link onesec.
<glowplug`> You can get an MSP430 board for ~$5. Then you would have to add the sensored control but that wouldn't be too bad.
<glowplug`> It honestly might be good enough without sensored. Certainly better than a chinese ESC.
<Sync_> that's nothing special
<glowplug`> These guys have good MSP430 boards. https://www.olimex.com/Products/MSP430/Header/MSP430-HG2231/
<Sync_> designs like that were around for some time
<glowplug`> No it's not special other than the low cost of the MSP430.
<glowplug`> And the code isn't.... a giant mess.
<glowplug`> The interferometer is for aligning the sample correct? If we have a microscope/camera do we need it?
<Sync_> no the interferometer is for position feedback
<glowplug`> During which stage?
<Sync_> you do not need if if you use a projector
<glowplug`> Developing the mask?
<Sync_> but I'm not sure if that works
<Sync_> I'd try a scanner
<azonenberg> Who wants accounts?
<glowplug`> It is a scanner. Sort of. It slowly scans a beam in the x,y.
<azonenberg> For spam reasons i'm limiting editors to confirmed accounts
<glowplug`> You did it! Awesome!
<glowplug`> Account glowplug. =)
<glowplug`> for email you can use raincloudcomputers@lavabit.com
<glowplug`> The principle is that although it scans slowly compared to a laser projector. It will fully develop the resist along its path. And the path is any position on the x,y plane.
<glowplug`> Honestly if it fails it's only because I designed or built it poorly. In principle it will work.
<Sync_> azonenberg: that does not resolve for me
<glowplug`> I'm on the page. =)
<azonenberg> Sync_: i just made the subdomain like 15 minuts ago
<azonenberg> dns might not have propagated yet
<Sync_> yeah probably
<glowplug`> I have to go for a bit. See if you guys can hunt down and stalk a telescope maker type. I will try also.
<soul-d> even my super duper fiber con can't find it yet
<glowplug`> I'm so lucky. Haha
<glowplug`> Back in a bit.
glowplug` has quit [Quit: ERC Version 5.3 (IRC client for Emacs)]
<soul-d> if i had the money i would have made a 1meter scope already
<soul-d> or few small ones for that matter
<soul-d> still need guider for my 10"
<Sync_> meh
<soul-d> well plus im lazy to gather all the stuff i think it had to do with ability to damp up the aluminum layer
<soul-d> or mess about with silver stuff but you need to redo that like every few years
<Sync_> if you do not protect the al it will also go dull
<Sync_> you can just send the mirror away to a coater and get it coated
<Sync_> or if you want get a wideband dielectric coating
<soul-d> knowing netherlands /europe i better can buy the mirror if there is even company able to do it i know there was a
<soul-d> place but still size limited
<Sync_> a lot of companies do that in europe
<soul-d> havent looked at it since like 8 years it's that glowplugh mentioned it
<Sync_> you could also do it yourself
<Sync_> sputtering Al is quite easy
<azonenberg> Easy?
<azonenberg> i've heard Al in particular needs very high output powers
<azonenberg> due to the need to break through the oxide
<azonenberg> compared to other materials that have less tough oxides
<Sync_> yeah you need assloads of power but when you are evaporating it is not that special
<azonenberg> Evaporating vs sputtering is a big difference :p
<Sync_> true, but even sputtering is ok
<Sync_> I did not have an issue in my jar
<azonenberg> i want to build myself a DC sputtering rig
<azonenberg> How hard do you think that'd be?
<azonenberg> I have a mechanical roughing pump
<Sync_> not very hard
<azonenberg> but no other vacuum hardware atm
<Sync_> I did it in a jam jar
<azonenberg> Lol
<Sync_> that's enough if it can get low enough
<azonenberg> i think i'd get something a little nicer than that
<Sync_> no shit, but lazyness :D
<Sync_> now that I have the machines I'll build something more adequate
<azonenberg> combined with a homebrewed baseplate etc
<azonenberg> then i'd need a gas inlet, HV supply, etc
<Sync_> that's a little small
<azonenberg> That was just an example
<azonenberg> but bear in mind that i only need to fit a 2" wafer inside at most
<Sync_> yeah I don't really like bell jars
<azonenberg> The option i was thinking of actually was a cylindrical glass tube
<azonenberg> with flanges and two metal plates
<Sync_> well the problem is more or less that you need to have proper clearances for the heat
<azonenberg> that way i could have the HV coming in from the top
<Sync_> even with a cooling loop it is an issue
<azonenberg> Hmm
<azonenberg> Could you do it conductively?
<Sync_> also glass is annoying to get clean
<azonenberg> Not it's not
<azonenberg> effasol is what they use in the lab at school in the SEM stuff
<azonenberg> smear it on the inside of the jar
<Sync_> not getting it vacuum clean
<azonenberg> then just apply water and it dissolves, taking the film with it
<Sync_> that's easy enough
<Sync_> but getting your metals off it once they are there
<azonenberg> Yes
<azonenberg> that's what that stuff is for
<Sync_> I just don't like it
<azonenberg> it forms a water-soluble layer on the inside of the glass
<azonenberg> then it lifts right off
<Sync_> interesting
<Sync_> I'd probably just RF etch
<Sync_> 100% oxygen plasmas are nicely reactive
<azonenberg> Lol
<azonenberg> I want that in the longer term
<Sync_> I have a xds10 scroll here
<azonenberg> it would be so cool to be able to do SF6 plasma RIE
<Sync_> and two replacement scrolls
<Sync_> I'll try and see if I can get a recipient from work
<azonenberg> want
<Sync_> wee that thing is tiny
<azonenberg> Yeah, and $10K
<azonenberg> i feel like you could build it for a lot less
<Sync_> yeah it is not that hard
<azonenberg> I'd want to wait until i got a "real" lab set up first though
<Sync_> I wonder why such small plasma etchers are so spendy
<azonenberg> Because they can
<azonenberg> they sell to people with very deep pockets
<Sync_> you can get large evap setups for 15k used
xiangfu has quit [Ping timeout: 264 seconds]
<Sync_> I mean yeah they sure do but 10k
<azonenberg> well if i could get SF6 RIE
<Sync_> wat
<azonenberg> i could do nice cutouts in Si
<azonenberg> then sputter Ta + Cu
<azonenberg> boom copper damascene :D
<Sync_> meh no sensor head
<Sync_> interestingly nobody either sells the sensor with the controller or nobody sells the controller with the sensor
<Sync_> most also only sell the controller withot the exciter
<Sync_> I wonder if I could make my own EIES sensor
glowplug has joined #homecmos
<glowplug> Back.
<Sync_> if it had the sensor I'd buy it right away :/
<azonenberg> glowplug: you get my email?
<glowplug> I did but mu is giving me sass.
<azonenberg> ?
<Sync_> hmm
<glowplug> Mu handles my email. Not sure what it's problem is right now.
<Sync_> SF6 RIE has the annoying fluorine in the exhause issue iirc
<azonenberg> Sync_: Yes, i'd venti t outside :p
<azonenberg> vent it*
<glowplug> I've got my login though. =)
<azonenberg> glowplug: the server cannot send email if you're trying to do a password reset or something
<azonenberg> i havent configured that
<glowplug> I'll just use the default password for now.
<azonenberg> Once you log in you can change it
<azonenberg> you just cant ask the server to email you a new one
<glowplug> The problem I was having is that offlineimap pulled your email but mu didn't get it. Super annoying.
<azonenberg> oh lol
<Sync_> wat
<glowplug> Hardware guy learning emacs. Bumpy road.
<glowplug> Alright. Under Litho. "Sample alignment optics"
<glowplug> Supplies. I'll just do some basic things.
* Sync_ rages a little
<glowplug> Whats the rage? Haha
<azonenberg> Try to copy over whatever you can from the google code wiki
<azonenberg> i plan to remove most of that wiki so we don't have stuff in two places
<glowplug> Alright will do.
<glowplug> Everything under "supplies" are things considered "consumables" correct?
<azonenberg> yes
<glowplug> I will be dealing exclusively with things in the Litho category.
<glowplug> I have absolutely zero experience with everything else.
<Sync_> http://www.youtube.com/watch?v=TV5p36-c3R4 this hurts my vacuum self
<glowplug> This is what I mentioned earlier.
<glowplug> Do we need a magnetron?
<Sync_> no but magnetron sputtering is nice
<glowplug> That is stupidly cool...
<glowplug> Wait. Sputter copper metal onto glass.
<glowplug> Time to read the wiki....
<glowplug> What kind of materials can be deposited via sputtering in a DIY configuration?
<glowplug> This is a HIPIMS system?
<azonenberg> With DC sputtering, any metal pretty much
<azonenberg> with RF, insulators too
<Sync_> yay
<Sync_> there finally is cheap glass pipe on ebay
<glowplug> But we still need a spincoater for the photoresist?
<azonenberg> Yes
<azonenberg> and probably spin-on dopants
<azonenberg> unless you want to get into ion implantation
<glowplug> So many tools....
<glowplug> It sounds like a DIY HIPIMS and DIY spincoater is the way to go.
<glowplug> What is the glass pipe for?
<Sync_> temporary vacuum chamber until I get motivated enough to build something proper out of stainless
<glowplug> There is actually a video of my idea. http://www.youtube.com/watch?v=VjesdQa3mMU
<glowplug> He is using galvos though which is why it's so wobbly.
<Sync_> I wonder if that will lead to issues due to the angular beam direction
* Sync_ browses his stash of vacuum flanges
<glowplug> I thought about that. There exists optics that will take the beam from any angle and deflect downwards.
<glowplug> It also might not be an issue with a very small work area. Lets say 50x50mm.
<glowplug> Is 50mw sufficient to expose the resist?
<Sync_> laser with a maximum power of 3 W at 514 nm and 300 mW at 457 nm. This high power
<Sync_> output can be used for a second harmonic generation, while typical beam power required for
<Sync_> the exposure of color-center masks [1] is of the order 100 mW. An electro-optical modulator
<glowplug> Looks like I might as well use blu-ray then which is way over those power levels. But theres no cost difference.
<glowplug> I'm fine with that. =)
<Sync_> I'd still like to use my tankless water heater
<glowplug> For which process?
<glowplug> Under Metalization does "sputter coater" refer to the HIPIMS device?
<Sync_> haha no you do not get the joke
<Sync_> I have a huge argon laser
<Sync_> which basically just is a water heater...
<glowplug> OH Haha
<Sync_> well at least it tells you when it is on
<glowplug> Speaking of tube lasers. I need to figure out how to frequency stabalize shitty lasers so I don't need one. Think I'm going to attempt it with an FPGA.
<Sync_> what do you want to do?
<glowplug> For the interferometer.
<Sync_> use henes
<Sync_> they are cheap enough on ebay
<Sync_> an fpga is not needed, you just need to control the temperature and the current precisely
<glowplug> A frequency stabalized dollar store laser is 10% the cost though. =)
<glowplug> You need many discrete configurations though correct? Such as amplifiers, comparaters ect.
<glowplug> That could all be integrated into a ~$10 coolrunner type CLPD
<azonenberg> glowplug: if you can focus it sufficiently a bluray diode is going to be much cheaper
<azonenberg> Just use short pulses
<azonenberg> what you want to avoid is using so much power that you ablate the resist :p
<glowplug> Wait what am I thinking. We don't need the interferometer now correct?
<glowplug> Assuming I can keep the sample still during exposer.
<Sync_> you need a precision current source with a few passives and a pid temperature controller
<Sync_> which is just any micro with a high resolution temperature senor
<Sync_> if you do a projector you do not need any interferometers
<Sync_> they still are for a moving xy stage
<azonenberg> I think you will still have to do a fixed laser and moving x-y
<azonenberg> its much easier to move the sample by microns
<azonenberg> than to try to steer a beam by millionths of a degree
<Sync_> yeah
<glowplug> I can totally steer a beam by millionths of a degree.
<azonenberg> accurately? :p
<azonenberg> and repeatably?
<azonenberg> there's a reason nobody does this
<glowplug> One person did it. On Youtube. Hahaha
<azonenberg> With what kind of resolution?
<glowplug> Horrible. Because he used Galvos.
<azonenberg> And what do you plan to do
<glowplug> But steppers with a 200/1 gear reduction should give enough resolution.
<azonenberg> Lol
<azonenberg> Do the trig
<Sync_> enjoy your backlash
<azonenberg> If you want 1um shift
<Sync_> and yeah do the trig
<azonenberg> at say a 10cm distance
<azonenberg> any idea what fraction of a degree that is? :P
<glowplug> Backlash in the gearbox? I did do the trig but for 10mm above the sample.
<azonenberg> 10mm?
<azonenberg> you realize you need room for the optics
<glowplug> In 1/4 of a second of movement at 200 pulses per second you get 1/4 full rotation. 200 reduction gives you .001 degrees change in that period.
<azonenberg> Ok
<azonenberg> 0.001 degrees
<azonenberg> do you know how huge that is?
<glowplug> Let me find my notes onesecond.
<glowplug> I got .048mm travel with 1 second of motor rotation at 10mm from the sample.
<azonenberg> 48 microns
<azonenberg> thats huge
<glowplug> Nope I have 1000/1 reduction here not 200. Thats why.
<glowplug> Thats for a full second of motor rotation.
<Sync_> uhh
<Sync_> 1000 times reduction with gears
<azonenberg> LOL
<Sync_> do you know how large that is
<Sync_> what the hell are you smoking
<glowplug> How large 48 microns is?
<glowplug> That is with a full second of rotation or 200 pulses.
<azonenberg> 1000:1 reduction...
<azonenberg> how much backlash is that gonna be at the low end?
<glowplug> Right. That is with a 4 stage planetary. Laser cut out of acrylic.
<azonenberg> five or ten revolutions maybe?
<glowplug> I mean there is going to be backlash no matter what. Whether its in the gearbox or the ball screw.
<glowplug> At least this way the machine costs $100 total instead of $2,000 to get results.
<azonenberg> Completely useless results
<Sync_> you need to move 5.7µ°
<glowplug> The only reason it wouldn't have 10 micron repeat accuracy is if the gearbox's were absolutely horrible.
<glowplug> 5.7u degrees sounds scary. But we have to make adjustments of that accuracy everywhere. Even down to aligning the sample.
<glowplug> Having a stepper do it isn't that crazy.
<Sync_> o.0
<Sync_> I want to see that
<azonenberg> having a stepper do 175,000 steps per revolution?
<azonenberg> oh, and that assumes you can fit your optics in 1cm from the sample
<azonenberg> Which is rather unrealistic especially for a homebrew setup
<azonenberg> i'd expect more like 10cm
<Sync_> I mean shit, it is easier to build an xy table that does that
<Sync_> and then you can scan a *very* small area
<glowplug> No stepper has 175,00 steps per revolution. Thats what the reduction is for.
<Sync_> because your focus will deteriorate
<azonenberg> That was the other reason i wanted to go x-y
<azonenberg> you dont need to mess with the focus
<Sync_> yes but your quad planetary has an assload of play
<Sync_> you cannot avoid it
<Sync_> that's why there are hypocycloid drives and harmonic drives
<glowplug> I cannot avoid gear reductions anyways. On a SCARA, Cartesian, doesn't matter.
<glowplug> Hypocycloid drives are unsuitable due to wobble no?
<azonenberg> yes, but it's better to move the sample a large distance
<azonenberg> than a mirror a tiny distance
<Sync_> the reduction you need is just unrealistically small
<Sync_> or rather large
<glowplug> If a Hypocycloid drive is suitable I will use that.
<glowplug> I assumed the offcenter mass would make it useless.
<Sync_> you need a good one
<azonenberg> i still think a Cartesian micrometer screw with a geared stepper is probably the best option
<glowplug> Ok then I will concede that the 4 stage planetary probably has too much backlash for mirror positioning of that accuracy.
<azonenberg> and interferometers for positional feedback
<glowplug> If I can get a hypocycloid drive laser cut and get my 1000/1 with only 2 stages then I think it is work trying.
<azonenberg> I want to build a Michelson interferometer some time soon just to play with
<azonenberg> And do the math for how many teeth you'd need on the gears
<glowplug> That would eliminate any backlash in the system.
<azonenberg> how big are these gears gonna be?
<glowplug> I did. With 4 stages that reduction is possible with gears that fit on a NEMA-23 sized stepper.
<azonenberg> How many teeth
<glowplug> But I agree that would likely have too much play.
<azonenberg> And yeah
<azonenberg> You will want positive feedback
<azonenberg> and you cant get that with beam steering
<glowplug> And the hypocycloid drive (if stable) would work.
<azonenberg> moving the sample, though, allows you to move with the beam blanked
<Sync_> are you really expecting lasercut pieces to be accurate enough for that
<glowplug> For a hypocycloid drive?
<Sync_> yes
<Sync_> I mean it'll spin
<Sync_> but will it really be backlash free?
<Sync_> the one I have is wire EDM'd
<glowplug> All hypocycloid drives are backlash free.
<glowplug> Just not wobble free.
<glowplug> And that is something I don't have the answer for. I don't know how much it will wobble.
<glowplug> The cost difference between my proposed system and the "traditional" is a factor of 50.
<Sync_> there are just too many engineering problems with your idea
<glowplug> I think it is worth persuing even for a short time if the design could enable people to contribute to the project who can't afford a $3000+ setup.
<glowplug> There are problems. Thats why I brought it up. And my backup plans.
<glowplug> But I think the hypocycloid gearbox could work if it was designed well enough to not wobble.
<azonenberg> what do you think of getting that
<azonenberg> if i want to fool with interferometry
<azonenberg> specs claim that 400nm movement accuracy is possible and so if i sample at a few KHz i should easily be able to pick up sub-wavelength shifts
<glowplug> 4 available.
<glowplug> Those are extremely nice.
<azonenberg> Well don't buy them before i get some :p
* azonenberg adds 2 to cart
<glowplug> Haha
<azonenberg> Ordered
<azonenberg> Seller has good feedback and they're only $35 each
<azonenberg> even if they arent good enough for fab i have a lot of other uses in mind
<Sync_> they are a little small
<azonenberg> Sync_: bear in mind we're talking about putting a single die on it
<azonenberg> I actually wanted to use them for microscopy
<Sync_> get me the other two azonenberg
<azonenberg> Sync_: You buy them :P
<Sync_> oh he ships world wide
<Sync_> that's nice
<azonenberg> I feel like it should be possible to hook a stepper up to them
<azonenberg> I want to make an x-y CNC stage for my microscope
<azonenberg> to do automatic scanning of large areas
<glowplug> You can send individual pulses to a stepper. The resolution of the machine I described earlier (assuming it could be built) would be 200/48 microns or 240nm.
<glowplug> So my backup plan. Lets see if I can get it shot down too.
<glowplug> The SCARA robot.
<Sync_> azonenberg: he has a nice xy stage too
<azonenberg> Sync_: i wanted linear ones for the interferometer experiments
<azonenberg> i'll look at his other stuff after payday :p
<glowplug> Actually it suffers from all of the same basic problems as the projector configuration. Plus some others.
<Sync_> XY axis RMS repeatability < 0.7 μm
<Sync_> nice.
<glowplug> Alright let me put it this way. I'm going to attempt the configuration I described in the off chance it does something useful we would have significantly cheaper litho. If I fail (~97%) then I'm only out a few bucks.
<azonenberg> for my cnc microscope i dont even need that good
<azonenberg> i just need to find an area of the sample
<azonenberg> oh and i need to get my camera working in linux
<glowplug> In the meantime I will also plan an x/y stage with interferometer feedback ect. ect.
<glowplug> What type of camera do you have?
<Sync_> aah I want that xy stage
<Sync_> that would solve all issues :D
<azonenberg> Sync_: go buy it?
<Sync_> yeah I'll do
<azonenberg> if you dont, in a couple of days i might :p
<azonenberg> Do you think a cheap pointer-level laser diode from digikey is stable enough to use for some quick fooling around with inteferometry?
<Sync_> not so sure
<glowplug> Yeah there was a video on YouTube I watched last night with a dollor store laser and he got a pretty good pattern.
<azonenberg> i mean i wouldnt use it for precise measurements
<azonenberg> but just to fiddle with the technique
<Sync_> I read a paper that said it depends on the diode
<Sync_> but I'd get a hene
<azonenberg> For a production system i might
<glowplug> You probably have to ditch the optics but the cheap diodes should work for experimentation.
<glowplug> If you have a really good quality current source it will work for sure.
<azonenberg> This laser has an integrated supply
<azonenberg> i have no idea how stable it is :p
<glowplug> I would pull its supply and the optics.
<azonenberg> its a pointer-style sealed unit
<glowplug> Yeah
<soul-d> they did that feynman expiriment thought experiment now btw
<glowplug> Just hack it up then use your own lense and current source.
<glowplug> And it will work awesome for a michealson setup.
<glowplug> The diodes are generally fine with everything else being junk.
<Sync_> azonenberg: a hene on ebay should be under 20$
<soul-d> oh
<azonenberg> One thing at a time lol, i dont have enough time to build the whole thing now
<soul-d> not the first
<soul-d> old news :P
<Sync_> just saying :D
<glowplug> The cheapest tubes I could find was a pair of two for $50.
<glowplug> But who knows what shape they are in.
<glowplug> The newish looking tubes are ~$75+
<Sync_> just wait and bid
<Sync_> I have a few and none was more than 25$
<azonenberg> Sync_: are diode lasers just not stable enough for interferometry?
<glowplug> They are!
<glowplug> But hene lasers are extremely stable by nature.
<glowplug> And a frequency stabalized hene laser is near perfect.
<glowplug> Diode lasers are good for experiment ect.
<glowplug> One of the best youtube videos from MIT. They make a simple michealson with a diode laser and calculate its wavelength. Very accurate.
<Sync_> azonenberg: the beam quality is a lot better
<azonenberg> Sync_: well i guess one nice thing about HeNe is a known output wavelength
<Sync_> that too
<azonenberg> So you can have numbers measured in nm rather than in wavefronts :p
<Sync_> yeah
<glowplug> The MIT experiment used a 650nm diode and 1/4 transitions were clearly visible. So that is fairly accurate.
<glowplug> Of course even if the laser was affordable they have top dollar optical equipment. =P
<glowplug> Yeah what you need to do for the diode laser interferometer is have really good optics and be able to adjust mirror very accurately.
<Sync_> one could just measure the wavelength
<glowplug> Absolutely. It would be a pain compared to hene but fun nonetheless.
<glowplug> Of course you need to already have a good laser to do that. O.o
<Sync_> it would be connecting the fiber to the wavelength meter
<glowplug> Do you guys really have every cool toy? Haha
<glowplug> Just so I understand fully. The current prevailing / preferred design of the direct litho.
<glowplug> It is a moving x,y stage driven by steppers?/servos? with a gear reduction. A mirror on all sides that are monitored by a series of interferometers for position encoding.
<glowplug> Flip down microscope with a point and shoot cam on the top. Micrometers for manaul adjustment.
<Sync_> well if I really wanted a wavelength meter I'd just go to the lab at work
<glowplug> Reading. This guy gets a good pattern with a radioshack diode. Skip to 5:30 http://www.youtube.com/watch?v=wdasvpErwsc
<glowplug> That was with manual alignment and no optical table. Pretty good in my opinion.
<glowplug> There is actually a PDF "laser scanner for direct writing lithography". Their design is quite a bit different than mine but with 1 micron accuracy and claiming that .5 is possible in principle.
<glowplug> Beam positioning is by modulation not mechanical actuation. Interesting.
<glowplug> Alright so a single HeNe laser gives feedback for both the X and Y. That is interesting.
<Sync_> yes
<glowplug> Did you see the cost estimations for the device? LOL
<Sync_> not too bad
<Sync_> this is 95
<Sync_> keep that in mind
<glowplug> We want manual digital control of the stage for positioning correct?
<glowplug> Analog joystick sort of thing.
<Sync_> not really, some kind of pc interface would be nice, linuxcnc would do just fine
<glowplug> Could just do a custom interface in linuxcnc.
<glowplug> With the camera output right in the software.
<Sync_> you don't even need a custom interface
<Sync_> the camera does nothing after you alinged everything
<glowplug> It would make automatic alignment possible. Thats just a toy though.
<glowplug> The PDF says servos. I wonder what the gear reduction is.
<glowplug> So you intend on using hypocycloid gearboxes and servos?
<Sync_> no
<glowplug> What kind of motor configuration?
<Sync_> some kind of dc servo with a high resolution encoder
<glowplug> Won't you need a gearbox?
<Sync_> probably, but 10fold would do
<Sync_> because the real positioning is done by the interferometer
<glowplug> I've been working a little on the encoder problem.
<glowplug> CD's have .5 micron resolution. You could build a very accurate encoder with cdrom guts and a ring with alternating 1-0 pattern written by the burner.
<glowplug> Could theoretically be millions of steps per rotation.
<glowplug> The pattern could be written by rigging the burner to be stationary and spinning the ring under it at a set rpm.
<Sync_> I'll just buy one
<glowplug> What resolution do you need?
<glowplug> How many steps I should say.
<Sync_> that depends on the gearing
<glowplug> Well for example a decent servo driven articulated robot has about 200,000 steps per revolution.
<glowplug> I'm sure we need more than that.
<Sync_> not really, I suspect the guys that make the stage I posted use something around 2800 step/rev wheels
<glowplug> Then there must be a massive reduction in the x,y stage?
<Sync_> ~20-30 I suppose
<glowplug> So thats 200-300 between the rotational and linear motion.
<glowplug> Thats 840,000 steps per linear travel.
<glowplug> Unfortunately I don't know very much about the design of those x,y stages. Does it make any sense to DIY the table itself?
<Sync_> no
<glowplug> I thought you would say that.
<glowplug> Well at least everybody else appears to be well funded. Haha
<azonenberg> well funded, lol
<azonenberg> i'm a grad student
<azonenberg> after rent and food i have a few $k a year to spend on fun stuff at most
<glowplug> So it's just Sync_ thats well funded? xD
<Sync_> probably
<glowplug> Hahaha
<glowplug> Well my encoder discs will be CD's. I'll put the designs on the wiki then hopefully other under-funded people can experiment. =)
<glowplug> I have fully conceded on the projection configuration. They exist (since 1989 apparently) but the design is not straitforward and there are too many variables that make it too error prone.
<Sync_> it's not really that I actually have a lot of money, but when I do I spend it
<Sync_> ah nice
<Sync_> avago makes 20000imp/r encoders
<Sync_> that makes for 80000 counts after quadrature decoding
<glowplug> How much?
<Sync_> good question
<glowplug> Uh oh. Prices arent on the website. Warning!
<azonenberg> glowplug: doing it on CDs is nontrivial
<azonenberg> they include error correcting codes and framing
<azonenberg> you cant just burn a timecode using regular burner firmware
<Sync_> unfortunately they are not avalible anymore
<Sync_> but I bet they are around 200$
<Sync_> maybe less
<Sync_> BUT they have nice new encoders
<Sync_> I'm just trying to figure out what they used in the stage
<glowplug> Yes I'm still working on that. It would be necessary to digitally modulate the laser diode directly and to read the photodiode output directly bypassing the drives uC.
<glowplug> In the PDF they say "DC Servo Motors". =(
<Sync_> yeah I guess they have a +-4-30 reduction
<Sync_> 5 is realistic with encoders avalible at that time
<Sync_> glowplug: actually they are not that bad
<glowplug> We need three of them though. =)
<Sync_> two
<glowplug> Oh right. Then the piezo for the Z which I have no experience with. =(
<glowplug> How much for disc + encoder?
<Sync_> I suppose one could live without autofocus for a while
<glowplug> Yeah I think autofocus will need to be added later.
<Sync_> around 100$ per set
<glowplug> Hey thats actually not that bad. Digikey must have very bad prices.
<Sync_> for a 17bit absolute encoder
<Sync_> that's pretty badass
<glowplug> I'll still take a crack at my CD rom encoders for free each. =)
<glowplug> We haven't talked about control much.
<glowplug> If we pipe everything into an FPGA, interferometer output, encoder output, and also use the FPGA for motor control we will have very good latency with very low cost.
<Sync_> I'd just rig linuxcnc to it
<glowplug> LinuxCNC can't be used for encoder feedback loops its too slow.
<glowplug> There is FPGA motion control IP that will do up to 8 servos with encoder feedback. Just integrate the interferometer output into the control logic.
<Sync_> tell that my cnc
<Sync_> yeah I have a mesa card
<glowplug> The PC software has 10us latency. Too slow for the actual control loops.
<glowplug> If you have a mesa card you are all set then. =)
<Sync_> that's the easiest solution
<glowplug> Absolutely.
<glowplug> Well that was easy. Haha
<Sync_> I mean further dicking around is just meh
<Sync_> at some point you have to throw money at it
<Sync_> we are still talking about positioning to 1µm
<Sync_> trying to make a custom drive controller consumes just too much time and effort
<glowplug> I agree. The mesa unit is actually affordable and adding the interferometer input and control logic won't be difficult at all.
<glowplug> Imagine if they had mesa cards in 1995. We just replaced $3,000 in equipment with $100. Haha
<glowplug> Do you have a 6125?
<Sync_> oh I'd have to look at my docs
<Sync_> I think it is a 5i22
<glowplug> For the 6125 (the pci express unit) if you buy 5 they are only $87 each.
<glowplug> I don't think there are any cost savings to be realized compared to that. O_O
<glowplug> That unit is quite a bit more expensive. But the 6125 will do everything we need for $87.
<glowplug> Is a blu-ray diode even accurate enough for 10um devices?
<azonenberg> The diode, sure
<azonenberg> you'd need better optics
<azonenberg> That's the paper i was talking about before
<azonenberg> under $1k end to end for ~1um resolutoin
<glowplug> Holy shit in a handbasket.
<glowplug> That is fantastic.
<glowplug> Their diode is violet with very low power. Thats great!
<azonenberg> They're using SU-8 negative resist but you can do the same thing with regular positive resist
<azonenberg> 405nm diodes work on that too
<glowplug> I have negative resist sheets thats what I'll be testing with so that works.
<glowplug> They are using steppers. No mention of gearing.
<glowplug> They are using the reduction in the x,y stage only???
<azonenberg> glowplug: quite possibly
<azonenberg> these stages will move maybe 100um per rotation
<azonenberg> so 1-degree precision is submicron
<glowplug> Ho
<glowplug> Ly
<glowplug> Shit
<glowplug> They dont even have interferometer feedback!
<glowplug> This is very encouraging
<azonenberg> That's something you'd put in if you wanted higher resolution
<azonenberg> the stages i picked up are supposed to be 250um per rotation of the knob
<azonenberg> 1-degree precision is easily doable with an ungeared stepper
<glowplug> Sub micron resolution.
<glowplug> Adjusting micrometers $150.
<glowplug> So there are steppers controlling the x,y stage.
<glowplug> The micrometers are where?
<azonenberg> The stage is probably micrometer screws on the axes
<azonenberg> that's typical
<glowplug> I see. So they added the micrometers TO the stage.
<glowplug> Attached steppers to the micrometers.
<azonenberg> the micrometers probably shipped with the stage
<glowplug> So all adjustment is done in sfotware.
<glowplug> *software
<azonenberg> a high-precision stage without micrometers is pretty useless
<azonenberg> And for that design it does seem like it's all open-loop contorl
<azonenberg> control*
<glowplug> Have you seen these?
<glowplug> It is open loop. But there is very little load on the steppers.
<glowplug> For submicron though I would think that encoded servos are more reliable.
<azonenberg> Submicron is a whole other animal
<glowplug> Haha
<azonenberg> At that point i might think about switching from laser to e-beam
<azonenberg> or alternatively using nearfield optics
<azonenberg> Neither is easy
<azonenberg> But ~1um with lasers should be very doable
<glowplug> A lot of these costs can be even further reduced.
<glowplug> Still need ~$100 for the blu-ray laser + lens.
<glowplug> But you can save $50 on the controller, $300 on the stage + micrometers, Digital microscope can be made for ~$30 thats another $120 savings.
<glowplug> Based on these numbers it can probably be done for ~$400 total.
<azonenberg> microscope for $30?
<azonenberg> lol
<azonenberg> i dont think so
<azonenberg> you cant even get a good objective for that much
<Sync_> yeah I'm not a huge fan of that design
<glowplug> The webcam design?
<Sync_> it sure does work but it's openloop
<azonenberg> Sync_: I like their design but i feel like it can be improved quite a bit
<azonenberg> i'd want to clone it first
<glowplug> Oh the entire stage.
<Sync_> yeah the idea is great
<azonenberg> then add closed-loop feedback
<glowplug> I agree with closed-loop. Thats pretty much a no brainer.
<glowplug> Do you have a link to the sort of microscope optics that would be suitable?
B0101 has joined #homecmos
<Sync_> meh do I really have to fire up the vpn for access
<azonenberg> I'd use an Olympus objective from ebay
<azonenberg> machine a custom tube
<glowplug> You keep getting kicked? =(
<azonenberg> then use a nicer camera, maybe a HD webcam
<Sync_> oh
<glowplug> I suppose the worlds view of our progress will be through that microscope.
<Sync_> reminds me
<Sync_> microscope
<glowplug> So that is probably a good idea.
<azonenberg> Well
<azonenberg> you want to use a nicer one for inspection
<azonenberg> this would be just for alignment
<azonenberg> Actually, screw that
<azonenberg> you might not even need the uscope
<azonenberg> if you're doing direct write
<azonenberg> to make masks
<Sync_> yeah
<azonenberg> you can forget about alignment since once etched the mask wil then be used in a contact aligner
<Sync_> no need to align anything
<glowplug> The scope is for calibrating the machine.
<glowplug> It wont work correctly the first go.
<glowplug> And also uploading images / video to the wiki.
<azonenberg> You will want another scope for inspection of work
<azonenberg> the one on there won't be a nice one you can actually use for general imaging
<glowplug> How much will the "actual" scope cost?
<azonenberg> I got mine on ebay for $250 + $250 + $50
<azonenberg> body, head, and illuminator respectively
<glowplug> O.U.C.H.
<Sync_> that's not a lot
<azonenberg> that's a lot?
<azonenberg> lol
<Sync_> to put that in perspective
<azonenberg> New, it was probably five or ten grand
<azonenberg> $500ish is dirt cheap
<Sync_> shit I have spent over 4k on my mill
<Sync_> so what.
<glowplug> One thing you guys have to realize.
<glowplug> The cheaper this stuff is the more people can contribute.
<glowplug> There is a cascading effect of money and support to a project.
<azonenberg> It's never going to get to the point of being home-PCB-etching availability
<azonenberg> I'm shooting for "affordable by a hackerspace"
<azonenberg> Or by a sufficiently insane hobbyist
<glowplug> But it can get down to "enthousiest garage fab shop with a modest income".
<glowplug> Right. That seems a good target.
<glowplug> But equipment under $2000 total (for everything) is probably realistic for that goal. Not $20,000 total.
<Sync_> wat
<azonenberg> $2000? Lol
<azonenberg> $10K is more what i was hoping to, after lots of tuning
<azonenberg> the prototype stuff will probably cost me more like 50k
<Sync_> 30k is a good starting point imho
<Sync_> which is affordable for a hackersapce
<Sync_> ~space even
<glowplug> The spincoater can be made for ~$100. These guys have litho 1um litho with ~$500 in parts.
<azonenberg> well, once i get out of school
<azonenberg> high on my list of major purchases is a secondhand FESEM
<glowplug> $500 for a good microscope.
<Sync_> oh yes azonenberg
<glowplug> We are at $1,100. What other major components are absolutley necessary?
<Sync_> they can be had quite cheaply
<azonenberg> glowplug: Consumables
<Sync_> yeah comsumables
<Sync_> vacuum parts
<azonenberg> Dopants from emulsitone are $500 each for P and N type
<azonenberg> or is it $250?
<azonenberg> that much again for undoped spin-on-glass
<azonenberg> $300ish for a bottle of photoresist
<Sync_> photoresist
<azonenberg> HMDS can be had cheap in small volume
<azonenberg> trace metal grade TMAH
<Sync_> a wafer is 10-30$
<azonenberg> not sure what the going rate on that is
<Sync_> I guess tmah can be cheap
<Sync_> but in small units hmm
<Sync_> dunno
<Sync_> I'd say 5k minimum for a basic setup
<Sync_> access to a machine shop is neccessary tho
<azonenberg> Yeah
<glowplug> For consumables we can order in volume then do pre-packaged kits for the community.
<glowplug> You get the economy of scale plus the cost to each hobbyist is very low.
<glowplug> The vacuum is for which process? How strong of a vacuum?
<Sync_> seriously, this is not the kind of thing you can pull off with a tutorial
<Sync_> vacuum for any kind of metal deposition or plasma etching
<azonenberg> And for evaporation you need 1E-6 torr range
<glowplug> No I don't think tutorial. I think there are enough hobbyists out there with general knowledge of this stuff but thinking that its completely out of reach without school / work equipment.
<azonenberg> so diffusion pump range at best
<glowplug> They just need to be proven wrong.
<Sync_> plasma etching wants 10^-2mbar and evap 10^-4 or better preferrably
<azonenberg> Sputtering can be done as high as 30mtorr but that's after you have a good argon purge
<azonenberg> So you need to get the chamber down quite a bit lower first
<glowplug> What is the plasma etching for? This is the part of the process that I now nothing about.
<glowplug> *know
<azonenberg> I was initially hoping to get 10um features working with just wet etches
<azonenberg> that might be wishful thinking
<azonenberg> i've done 20 with wet etches experimentally and they worked fine, though i had some adhesion problems here and there
<Sync_> that's quite a good deal for the parts
<Sync_> three ion pumps, all the CF hardware
<azonenberg> Sync_: but not as a SEM
<Sync_> yeah
<Sync_> just browsing general vacuum stuff
<Sync_> too lazy to study :D
<glowplug> The guy on Youtube with the diy magnetron. What was his vacuum source?
<Sync_> refridgeration vac pump
<glowplug> Is that sufficient for everything we need?
<azonenberg> Was that the guy who had all kinds of problems getting deep enough vacuum?
<Sync_> that's probably ok
<azonenberg> oxygen contamination everywhere?
<Sync_> yeah everything oxidized
<Sync_> no glowplug by far not
<azonenberg> I trained on that same setup lol
<Sync_> oh
<Sync_> snap!
<Sync_> we have one too
<Sync_> :D
<azonenberg> an 840?
<Sync_> yep
<azonenberg> Lol nice
<glowplug> I see that. Copper oxides on the slide.
<Sync_> it's our oldest one
<azonenberg> Sync_: same here
<azonenberg> The JSM-6335 is the one i've used more recently
<azonenberg> it's a FESEM
<azonenberg> $20/hr more to use but thats not a huge deal
<azonenberg> then if you're made of money you can use the Zeiss in the cleanroom :p
<Sync_> we can use all the equipment for free
<glowplug> So what processes require vacuum that are *Absolutely* necessary for a 10um device?
<azonenberg> glowplug: Metal deposition
<Sync_> metalization
<azonenberg> You can use spin-on dielectrics and wet etching for large features
<azonenberg> so no need for RIE, PECVD, etc
<glowplug> Alright so wet etching refers to using the spincoater for metal deposition. That I get now.
<azonenberg> No
<Sync_> pecvd is something that is not happening with my pumps :D
<Sync_> too much ugly stuff involved
<azonenberg> it refers to etching stuff using a liquid acid
<azonenberg> rather than a plasma
<glowplug> What are spin on dialectrics?
<azonenberg> Which means it can be done at atmospheric pressure
<azonenberg> glowplug: stuff like http://www.emulsitone.com/sif.html
<azonenberg> I suggest you learn more about semiconductor fab in general
<glowplug> I'm reading nonstop already. I just discovered this stuff a week ago. Haha
<azonenberg> Then work on how to scale it down to hobbyist level
<Sync_> the second hitachi I posted seems to be ok
<Sync_> it's no fesem for sure but it'll do
<azonenberg> Sync_: an 840 would be nice to have at home as long as i could rent time on a nicer scope when i needed it
<glowplug> But I mostly focused my research so far on the litho because I understand those principles well.
<azonenberg> Sync_: But stuff like http://i.imgur.com/esVwLlK.jpg
<azonenberg> i dont think you'll get that kind of resolution with an 840
<glowplug> It was already determined that the wet etching is not very reliable with devices smaller than 10um correct?
<Sync_> probably not
<azonenberg> the round dots are 60nm nominal colloidal silica left over from from CMP
<azonenberg> the device is a 45nm FPGA
<Sync_> hehe nice
<azonenberg> glowplug: I forget when industry stopped using wet etches
<azonenberg> For submicron i'd definitely want to use plasma
<azonenberg> possibly sooner
<Sync_> but if you are into that you can just buy one, reverse engineer shits and pay it off
<glowplug> And the magnetron is enough for every metalization step for a device correct?
<azonenberg> Sync_: i'm the guy behind siliconpr0n.org :p
<Sync_> yes I know
<azonenberg> glowplug: The full suite of film coating capabilities you'll want
<Sync_> but if you wanted to do that commercially
<azonenberg> is spin coating, DC sputtering, RF sputtering, evaporation, and PECVD
<azonenberg> all but the first require vacuum
<azonenberg> You can probably live without PECVD if you do reactive sputtering for dielectrics, or use spin-on glass
<glowplug> The PECVD we don't need immediately correct? What about evaporation? That is a different device from the magnetron?
<azonenberg> Evaporation and sputtering might be able to use the same chamber depending on how you design it
<azonenberg> the processes are different but they have some similarities
<azonenberg> DC and RF sputtering can definitely share the chamber
<glowplug> So a well designed chamber will take care of every vacuum process.
<azonenberg> Not necessarily
<glowplug> Then you have litho and the spin coater.
<Sync_> shits
<Sync_> I want
<glowplug> Thats only three primary devices.
<azonenberg> Sync_: *drool*
<azonenberg> glowplug: well litho is not one tool
<glowplug> I mean three main monolithic tools.
<azonenberg> you need both the direct-write system for mask making and then the contact aligner
<glowplug> Ahh right. For exposure to the mask. So that is 4 primary tools then.
<glowplug> *through the mask
<azonenberg> Then general lab glassware etc for wet processing
<azonenberg> microscope for in-process inspection
<azonenberg> If you're doing copper metalization or multiple metal layers CMP would be helpful
<glowplug> Right. ~$500 scope for checking work.
<glowplug> So for an absolute bare-bones lab.
<azonenberg> But for a simple 1-metal process using evaporated aluminum you could get away without CMP
<glowplug> $200 spincoater, $500 litho, $500 scope, $200 exposure setup, $XXX magnetron?
<azonenberg> For bare bones you'd want the spin coater, laser direct write for masks, contact aligner, and an evaporator
<azonenberg> I would do evaporation as the first vacuum tool
<Sync_> yeah
<Sync_> it is the easiest
<azonenberg> it needs deeper vacuum but you dont need to add gas to it
<azonenberg> its basically just a light bulb
<Sync_> still around 1k realistically
<azonenberg> that happens to have somethign stuck to the filament
<glowplug> The evap is ~$1k?
<azonenberg> You could use spin-on dielectrics
<azonenberg> so the evap would only be for metal
<Sync_> and even 500 for a microscope is quite a bargain
<azonenberg> Sync_: yeah, i got lucky
<azonenberg> asking price was 500 each on body and head
<Sync_> I know
<Sync_> :D
<azonenberg> i bargained them down
<azonenberg> Oh, and then a furnace for implants
<Sync_> or oxides
<azonenberg> I'd do spin-on glass
<azonenberg> since you'd need that for an ILD anyway
<glowplug> Does the furnace need vacuum?
<Sync_> (then an AFM for measureing the thicknesses) ;)
<azonenberg> might as well use it for field oxide
<azonenberg> glowplug: No vacuum, but it might be nice to purge it with inert gas
<azonenberg> argon or something
<azonenberg> if you are trying to heat and not oxidize
<glowplug> I think I've seen nitrogen used.
<Sync_> yeah but the inert gas needs to be fairly clean
<azonenberg> Yeah
<Sync_> which has been the issue for us
<glowplug> So the evap is ~$1k?
<azonenberg> and all of the chemicals have to be completely metal-free
<Sync_> we tried and we got oxide
<Sync_> yeah
<Sync_> abou that
<Sync_> the scope too
<azonenberg> Sync_: Thats why i want to focus on MEMS first
<Sync_> if you calculate with bargains that are possible
<Sync_> and not 1 in 100 bargains
<Sync_> yeah
<glowplug> And you think the $1k scope is absolutely necessary to check work at ~10um?
<Sync_> I know why you want to do that azonenberg ;)
<Sync_> yes
<glowplug> Alright so we need to narrow down the oven cost.
<glowplug> ~$200?
<Sync_> as I said, around 5k is a _realistic_ starting price
<azonenberg> glowplug: LOL
<azonenberg> 200?
<azonenberg> this is not an oven
<azonenberg> its a furnace
<azonenberg> you have to get up to around 1200C
<Sync_> I got my ceramic burning oven for 50
<azonenberg> Sync_: metal-free? :p
<Sync_> but it's not for that
<Sync_> haha sure not
<Sync_> :D
<azonenberg> A new one from MTI is probably gonna run you about a grand
<azonenberg> Do you think AL2O3 insulation is safe for CMOS?
<Sync_> I could make a SI insert
<Sync_> yep
<glowplug> Is it possible to get to 1200C with a DIY unit?
<glowplug> Induction cant be used can it?
<azonenberg> This is what i was looking at
<Sync_> oh hmm
<Sync_> let me ask our techs tomorrow
<azonenberg> Sync_: you think that one would be cmos-compatible if you were careful about only putting freshly RCA-cleaned silicon in it?
<glowplug> You can get to 1200C with induction for pennies.
<Sync_> we had all sorts of shits going on in our ovens
<azonenberg> Sync_: lol
<azonenberg> well high-temp processes are where the worst ion contamination happens
<glowplug> I dont know if it would just blow up the sample or something.
<Sync_> we have a company in the institute that does ovens
<Sync_> you cannot use induction
<glowplug> Your positive?
<Sync_> yes
<glowplug> Well that sucks. =(
<Sync_> you could use induction to heat the sides of the oven
<glowplug> Good point. Then heating from the oven itself to the sample would be by convection.
<Sync_> convection at over 900°C?
<Sync_> not so much
<azonenberg> more radiation i'd think
<Sync_> yeah
<glowplug> Does that work?
<glowplug> From induction to radiation heating?
<glowplug> What kind of material would the oven be?
<glowplug> Steel?
<Sync_> something ferrous
<glowplug> Iron oven?
<Sync_> actually azonenberg I think we use si wafercarriers
<azonenberg> you'd have to be really careful about metal particles falling out and landing on the samples
<Sync_> and quartz holders for those
<azonenberg> Sync_: i was thinking quartz boarts
<Sync_> yep
<azonenberg> the question is whether being near hot Al2O3 is dangerous to transistors
<azonenberg> i wouldn't actually have them touching
<Sync_> well it is used as gate dielectric
<Sync_> so
<azonenberg> But that's post-diffusion righht?
<azonenberg> the big time you have metal issues is just before an implant or oxidation bake
<Sync_> yeah
<azonenberg> after that, no more furnace
<azonenberg> just whatever heating you get during an etch or deposition step
<glowplug> I think that a 10kw powerstage is enough.
<azonenberg> You realize that you can only pull around 2kW from an average (American) mains circuit, right?
<glowplug> This is for garages and hackerspaces.
<Sync_> azonenberg: it seems to be safe
<glowplug> They can figure out how to draw 10kw. EV chargers do.
<glowplug> It's not like the $1,000 oven draws much less anyways.
<glowplug> Yeah 10kw is enough. That much power can heat to 1100C in 9 seconds.
<Sync_> wat
<Sync_> yeah uh well a nut
<glowplug> We need to heat the sample overnight correct?
<azonenberg> glowplug: For diffusions normally an hour or so
<Sync_> hmm
<Sync_> I guess you could make a sled for the wafer out of quartz
<glowplug> If we need to ramp up to 1200C and down again within an hour. I think it can be done with a 10kw powerstage, some copper, and an iron box.
<Sync_> so that metals do not affect you that much
<Sync_> now that is another thing you have to develop
<Sync_> a 10kW smps is not easy
<Sync_> let alone a 10kW induction heater
<Sync_> the tank caps are quite spendy
<Sync_> the currents involved are large
<glowplug> It may be the case that 10kw is actually very overpowered.
<Sync_> yes
<glowplug> Do we know the power draw from the retail furnaces at the wall?
<Sync_> the one azonenberg had uses 700W
<glowplug> OH
<glowplug> Then the induction version should have comparable power usage.
<glowplug> Could just use an ATX SMPS
<Sync_> ...
<Sync_> yeah right
<Sync_> did you actually build a induction heater before?
<glowplug> I understand their principles.
<Sync_> understanding their principles does not show you the pain to make one
<glowplug> It looks like it can be done without an SMPS at all.
<Sync_> it *is* a smps
<glowplug> In this design I don't see any IGBT's / FETs.
<Sync_> a smps does not need igbts or fets
<Sync_> it can also use bipolars or thyristors
<glowplug> I dont think there are any switches.
<Sync_> oh sure
<Sync_> otherwise it would not work
<glowplug> I'm looking!
<glowplug> So far I just see diodes.
<glowplug> And a massive bank of caps.
<Sync_> it starts at page 10
<glowplug> Got it!
<glowplug> I'm not sure that these FET's are switching 10kw however.
<Sync_> they could
<Sync_> they are not but they could
<glowplug> This is an interesting design.
<glowplug> The FET's are switching AC?
<Sync_> they receive dc
<Sync_> and make ac out of it
<glowplug> Ahh ok I thought it was a solid state AC voltage regulator.
<glowplug> So we are positive that a furnace with gas purging and 1200C max temp is available for $1000?
<Sync_> about that
<glowplug> The DIY option is probably in the $300 range + ~100 hours labor.
<glowplug> Maybe less. Hard to say.
<glowplug> Plus misc and consumables thats hovering at $4500-$5000
<glowplug> For an absolute basic setup
<glowplug> Those numbers are with the retail furnace.
<Sync_> yeah what I said basically
<glowplug> I wonder if it is realistic to complete a device in a distributed way. I do the litho, mail out to doping, mails out to oven, ect. ect.
<glowplug> It would be slow but maybe possible.
<Sync_> some processes cannot have a lot of time between them
<Sync_> because of oxidation and contamination
<glowplug> Ahhh yeah.
<glowplug> Plus USPS likes to drop kick packages.
<glowplug> Litho can be outsourced but everything else is probably best kept in a single lab.
<Sync_> well mask making can be outsourced
<Sync_> otherwise it takes ages
<glowplug> By outsourced I mean one hobbyist makes a set of masks then mails those out to another hobbyist with the lab.
<Sync_> yeah but you could also get commercial masks
<glowplug> Absolutely.
<Sync_> it's the same thing basically
<glowplug> Agreed. Just the price difference. And quality. Haha
<glowplug> What is your opinion about the PDF litho setup with the steppers? Asside from obviously switching to closed loop servos. What about the lack of interferometer feedback?
<glowplug> Maybe thats something we can add later?
<Sync_> yeah
<glowplug> Thinking the same thing.
<glowplug> I have a way way way better picture of whats going on now.
<glowplug> Is the furnace considered FEOL?
<glowplug> Is it its own category?
<azonenberg> Furnace is probably FEOL since its not used once you put down metal
<glowplug> Alright I'll add to the wiki. =)
<glowplug> What else is FEOL?
<azonenberg> Everything that happens before you put down the first metal layer
<glowplug> That would include Litho then.
<azonenberg> I'd file litho separately
<azonenberg> since its done everywhere
<glowplug> Good point.
<glowplug> So for a basic setup there isn't much FEOL then.
<azonenberg> Not too much, since we arent doing ion implantation etc
<azonenberg> The acronyms stnad for "front end of line" and "back end of line" respectively btw
<azonenberg> in reference to the order wafers normally go through the tools during fab
<glowplug> Yup. =)
<glowplug> Just burned through the wikis.
<glowplug> These are the parts I knew *zero* about. Haha
<glowplug> So the contact mask aligner and sample alignment optics are part of the same apparatus correct?
<azonenberg> Yes
<azonenberg> Initially we'll do manual alignment
<glowplug> At this stage we are aligning the sample to be exposed in UV through the resist.
<glowplug> *mask
<azonenberg> Mask is to be clamped in place firmly
<azonenberg> so we just need an x-y-theta stage
<azonenberg> (manual)
<azonenberg> with a relatively imprecise Z axis
<azonenberg> Move the wafer up to just below the mask
<azonenberg> line it up
<azonenberg> bring it up into contact
<glowplug> Right it needs pan and tilt also. :/
<azonenberg> slide the microscope off and load the exposure lamp
<azonenberg> Nop[e
<azonenberg> nope*
<azonenberg> Just rotation
<glowplug> Ohh I see. Derp.
<azonenberg> As long as you machine things such that the stage and the mask are sufficiently parallel
<azonenberg> you just need to line up two in-plane images
<glowplug> This apparatus also needs a small vacuum correct?
<Sync_> pfft no fancy thing as 3d capacitors like we have on dram now? ;D
<glowplug> OR are we just bolting the sucker down?
<azonenberg> Light vacuum for the chuck to hold the wafer down is helpful
<azonenberg> but that's nothing
<azonenberg> any pump can do that
<glowplug> Right.
<azonenberg> Same with the spin coater
<azonenberg> You could potentially even just use double-stick tape
<azonenberg> but vacuum is less likely to damage stuff
<glowplug> Agreed. And the tape is probably more expensive.
<glowplug> So the microscope for this apparatus.
<glowplug> Expensive unit or cheap unit?
<Sync_> I guess getting the wafer off is quite tricky
<azonenberg> relatively nice
<azonenberg> But you only need one objective
<azonenberg> not a full set
<glowplug> ~$200 optics?
<Sync_> I guess you could use a single optic microscope
<azonenberg> One objective, custom-made tube
<azonenberg> then a camera on the top
<azonenberg> You need two of them
<Sync_> they are not too expensive even new
<azonenberg> to see two sets of alignment marks at once
<azonenberg> otherwise you wont get the theta right
<glowplug> Right. One comes into focus you take a snapshot then bring the other into focus.
<azonenberg> One on each side of the wafer
<azonenberg> you align it approximately by naked eye
<azonenberg> to get it into the FOV of the microscope
<azonenberg> then tweak the stage to line it up exactly
<glowplug> Two cameras you mean? And one scope?
<azonenberg> you'd use a split-field view on the computer
<azonenberg> No
<azonenberg> Two objectives, two cameras
<azonenberg> an inch or so apart
<glowplug> I've seen a single scope/camera configuration.
<azonenberg> That is... not nearly as good
<glowplug> You focus onto one marker then snapshot it.
<glowplug> Then switch focus to the other.
<azonenberg> This is the setup i'm talking about
<azonenberg> Two vertical camera/microscopes on the right
<azonenberg> UV exposure system at the left
<azonenberg> chuck at the middle
<azonenberg> you slide one or the other into the central slot to use it
<azonenberg> the mask stays in one spot
<azonenberg> and the wafer doesnt move except for tiny alignment corrections
<azonenberg> needless to say that process is UV sensitive
<azonenberg> so you want say red or yellow LED illumination
<glowplug> How many watts for the entire grid?
<glowplug> That design is actually quite good.
<Sync_> not much
<azonenberg> for the microscope illumination? One or two LEDs each is enough
<azonenberg> for the main exposure assembly i'd use party-style blacklight fluorescent lamps
<azonenberg> maybe six or eight of them side by side
<azonenberg> less if only doing 2" wafers
<azonenberg> Professional shops use kW-range mercury vapor lamps for like 5-sec exposures
<Sync_> hm cut open mercury arc lamp
<Sync_> with a shutter
<azonenberg> but if you use a 20W lamp and it takes a minute orso
<azonenberg> big deal
<Sync_> that'd be easy to do
<azonenberg> you're not mass producing here
<azonenberg> Sync_: remember this isnt DUV resist
<glowplug> Wouldnt you just do what the growers do and find the wavelength mix needed and get LED's for those wavelengths only?
<azonenberg> its 405nm or so
<Sync_> yeah but the lamp will still do
<glowplug> Ahh I see
<glowplug> I have 3w violet LED's
<azonenberg> Something like that
<azonenberg> Fluorescents will probably be a lot cheaper
<glowplug> They probably produce more usable power in that wavelength than the blacklights.
<glowplug> I got them for only $12 each.
<azonenberg> Hmm
<glowplug> They are probably cheaper now.
<azonenberg> Just make sure you diffuse them well
<Sync_> yeah
<azonenberg> or you get some spots unevenly exposed
<glowplug> Good point.
<Sync_> I wonder when uv lets show up that actually last a while
<glowplug> So there are no software tricks to get that aparatus down to 1 scope?
<Sync_> not with annoying mechanical effort
<glowplug> The scope outputs are superimposed? Stereovision? Or seperate screens?
<Sync_> just two screens
<glowplug> I see. Page 3 explains what I was missing.
<glowplug> I've seen this proces before. But I forgot that you are aligning 4 total markers. Two on the aparatus and two on the sample.
<azonenberg> Yep
<glowplug> That would be a massive pain with only 1 scope. =(
<azonenberg> Yeah
<azonenberg> Possible, but you really dont want to
<azonenberg> spend the extra $200 or whatever on the second scope
<azonenberg> Two webcams, two tubes, two objectives