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/
Speedevil has quit [Quit: No Ping reply in 180 seconds.]
Speedevil has joined #homecmos
<azonenberg> o/ diginet
<azonenberg> we were talking sol-gel TiN as a possible gate material not requiring vacuum
<diginet> o/ azonenberg
<diginet> yes
<diginet> mind you that this is all theoretical
<diginet> I haven't found much conclusively on doing sol-gel TiN
<azonenberg> Most of the work here is
<azonenberg> You could also potentially do it with relatively loose vacuum
<diginet> TiN seems to be one of the most promising candidates for metal gates
<azonenberg> like, reactive sputtering of a Ti target in Ar + N2 atmosphere
<diginet> so that's encouraging
<azonenberg> at tens to hundreds of mtorr
<diginet> I think if it's possible with a harbor freight vacuum pump, then it's reasonable
<azonenberg> Yeah
<diginet> anything which requires more isn't
<azonenberg> You can do sputtering with a relatively cheap 2-stage rotary vane pump
<azonenberg> idk about harbor freight, but like united nuclear or something
<diginet> sure sure
<azonenberg> i have one that advertises 40 mtorr
<diginet> the question is whether you can do it well
<azonenberg> The sputter coater for SEM sample coating at RPI runs down to 30ish
<azonenberg> you normally purge down to 30 then coat at 60
<diginet> my goal is to eventually get the cost of the entire system to $5-6k. That's what the cost of early hobbyist 3D printers was. IMO anything more is unreasonable
<diginet> *were
<diginet> err
<diginet> actually no was was right
<azonenberg> Let's make something that works first
<diginet> sure sure
<azonenberg> also... grr
<azonenberg> cant remember who the guy in canada that's trying to make a stepper is
<azonenberg> he's been pming me and i forgot to tell him to take it to the channel
<diginet> anyhow hard vacuum is generally not possible for an amateur, putting aside money the equipment is expensive and temperamental (at least the used stuff)
<azonenberg> Yes, hard vacuum isnt
<diginet> huh
<azonenberg> But soft vacuum? Much more so
<diginet> indeed
<azonenberg> Especially relatively safe processes like sputtering
<azonenberg> which dont leave the nasty wastes you get from RIE
<diginet> is sputtering difficult to control?
<azonenberg> Control what?
<azonenberg> Film thickness? Deposition rate? Grain size?
<azonenberg> Purity?
<azonenberg> Density?
<diginet> everything
<azonenberg> AIUI the parameters available are pressure, gases, and current
<diginet> mostly film thickness though
<azonenberg> We could probably build a quartz thickness monitor easily
<azonenberg> just open up a crystal oscillator and measure the output freq as the resonant element gets heavier from the deposited film
<diginet> that might be difficult to calibrate
<diginet> aren't piezos notorious for nonlinearities?
<azonenberg> No idea
<azonenberg> All i know is that this is apparently the industry standard method for measuring thickness
<diginet> sure
<azonenberg> If we can find something else i'm all for it
<azonenberg> Another option is two electrodes separated by a glass plate
<azonenberg> and measure resistance between them
<diginet> (not trying to shoot you down, I'm just trying to do think of all options)
<azonenberg> given known distance and area we can calculate sheet resistance
<diginet> */to/do
<diginet> yeah
<azonenberg> Which is IMO the more important variable to control
<azonenberg> vs thickness
<azonenberg> This of course only works for conductive films
<azonenberg> not dielectrics
<diginet> my favorite paper is on a using sol-gel CeO2 as a high-k
osfe has joined #homecmos
chris_99 has joined #homecmos
Textmode has quit [Quit: "It was one dev, naked in a room with a carton of cigarettes, a thermos full of coffee and bourbon, and all his summoned angels."]
kristianpaul has quit [Remote host closed the connection]
kristianpaul has joined #homecmos
<_Sync_> diginet I regularily manage 10^-10mbar at home
<_Sync_> azonenberg: using a crystal osc as a microbalance works perfectly well
<_Sync_> and yes, they do suffer from nonlinearities but you can just change them often enough
<azonenberg> _Sync_: that's what i was thinking
<azonenberg> just open up a stock crystal osc
<azonenberg> buy like a dozen from the same manufacturer, same lot number
<azonenberg> calibrate them against a single reference TCXO
<_Sync_> no
<_Sync_> too much hassle :D
<azonenberg> open it up, measure the base clock frequency
<azonenberg> Deposit at a constant rate, measuring the frequency, for X time
<_Sync_> you just measure the frequency at startup
<azonenberg> When you're done, pull it off and measure the film thickness
<azonenberg> using a SEM or something
<_Sync_> again, too much hassle
<azonenberg> Guess it depeneds on how muc haccuracy you need
<azonenberg> If you know the dimensions of the resonant element, you might be able to just calculate how much mass was added by the shift in the frequnecy
<_Sync_> yes, but for +-5% just going by density works
<azonenberg> Makes sense
<azonenberg> and like i said before, i dont care about film thickness so much as resistivity
<_Sync_> our only crystal balance we run is in the gold coater
<_Sync_> and we just go by eye on there
<_Sync_> and I manage to make films with ~20nm thickness variation
<azonenberg> nice
<_Sync_> and I'd just use Al for contacts
<_Sync_> or gold for that matter
* Speedevil ponders time of fight mass spectrometry.
* Speedevil is contemplating silly things regarding batteries.
<Speedevil> Not fully understanding the chemistry when passing current through molten solutions of flourides = exciting!
<_Sync_> protip: buy a used RGA
<_Sync_> if you need to resolve under 200AMU that's your best bet
chris_99 has quit [Ping timeout: 272 seconds]
chris_99 has joined #homecmos
chris_99 has quit [Remote host closed the connection]
chris_99 has joined #homecmos
chris_99 has quit [Remote host closed the connection]
chris_99 has joined #homecmos
kristianpaul has quit [Ping timeout: 240 seconds]
kristianpaul has joined #homecmos
kristianpaul has joined #homecmos
Textmode has joined #homecmos
Textmode has quit [Read error: Connection reset by peer]
Textmode has joined #homecmos
chris_99 has quit [Quit: Ex-Chat]