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/
<whitequark> no manipulation required.
<whitequark> did I invent something well-known?
<SpeedEvil> Alternatively, you don't actually need wire.
<SpeedEvil> Take your anvil, and smack your blob a lot with a sledgehammer till you get a fine sheet
<SpeedEvil> Then just cut off strips with scissors
<davidc__> whitequark: is just purging a reaction vessel with argon at air pressure not going to work?
<whitequark> davidc__: well, the question of how to cast it remains
<whitequark> I could probably just blow argon into the crucible, yes
<whitequark> I really don't wanna purge some kind of casting mold with argon
<whitequark> That sounds involved and I'm lazy
<davidc__> whitequark: are the crucibles $$$?
<davidc__> I mean, if not, let it harden in crucible, remove from crucible with hammer?
<whitequark> they are not disposable.
<whitequark> well, maybe if you work in the US tech industry.
* whitequark shrugs
<whitequark> I have this small silica crucible (100ml or so?), $35 from China
<whitequark> It should probably just fall out of it tbh
<SpeedEvil> If you're just casting low-melt metal, you can groove some brick or something
<whitequark> Could work
<whitequark> I'd prefer to avoid anything porous and potentially contaminating
<whitequark> I got a graphite crucible for this as well, with nicer geometry than this silica one
<davidc__> whitequark: could you just pour it out onto heated glass?
<davidc__> or even just heat it in cheap glassware?
<davidc__> looks like melting temp for that is ~1600
<whitequark> davidc__: how I'm going to ensure it is always under argon?
<SpeedEvil> Calcium silicate seems pretty inert
<davidc__> whitequark: Oh, just build a box around the works and run argon into it
<whitequark> You could probably do that
<davidc__> I mean, my giant tank of Argon for .al welding was like $50 to fill
<davidc__> and that'll displace a hell of a lot more than you need!
<whitequark> here, Ar is $3 per 150L STP
<davidc__> 150L is a _lot_
<whitequark> It costs $30 to get a taxi to the station :p
<whitequark> well, no, more like $15
<whitequark> still, five times
<whitequark> davidc__: I just don't want to bother doing this in atmosphere. I already am making a vacuum induction furnace
<davidc__> fair enough
<whitequark> It's just as simple at that point, and superior in every respect
<davidc__> what are you using for pumping?
<whitequark> I bought the cheapest 1stage rotary vane pump I could find. it has no words on it except "2CFM"
<whitequark> Ostensibly for A/C but it totally works in continuous for extended intervals
<whitequark> as in, days.
<davidc__> Are you going to try and do a cold trap / etc?
<davidc__> (as in, so you don't deposit metal inside the pump)
<SpeedEvil> A sublimation pump is quite decent to get reasonable removal of oil backstreaming
<whitequark> I am going to get a cold trap in dry ice/acetone bath (prefer to carry it around over LN2 in a thermos flask; I'm too cheap for a dewar) but not sure if I will use it for this
<SpeedEvil> Ti film
<whitequark> I have a ~1m bellows connecting the chamber to the pump
<SpeedEvil> It has ~4L of heat-leak at ~+80C
<SpeedEvil> 4W
<SpeedEvil> If I replace the silly thing on top with a block of insulator
<SpeedEvil> Should work for LN2 just fine
<whitequark> Thermos flasks are totally OK for getting LN2
<SpeedEvil> Or, indeed, a vacuum chamber
<whitequark> Better not pour it all over you
<whitequark> Or close :p
<SpeedEvil> - the input hole is ~2.5"
<whitequark> eh. no fittings. my chamber is /full/ of various holes and fittings
<SpeedEvil> whitequark: remember - many vendors won't fill thermoses.
<SpeedEvil> whitequark: So paint it the right colour, and print out the right stickers.
<whitequark> pffff
<whitequark> I'm in RU
<SpeedEvil> Ah.
<whitequark> as long as I'm paying, they will fill a soda bottle
<whitequark> RU is great at not having bureaucracy stand in the way of good, slightly irresponsible research :p
<whitequark> A good vacuum chamber is ~2m of stainless
<whitequark> er
<whitequark> 2mm
<whitequark> less than 1.5mm is downright implodey
<SpeedEvil> yes, but that already has a vacuum in
<whitequark> I'm not sure if that thing will hold vacuum. I mean, it's not designed to hold vacuum /inside/
<whitequark> rather than inside the walls
<SpeedEvil> There is no seperate vacuum vessel
<SpeedEvil> It's just two skins of SS
<whitequark> yes
<SpeedEvil> So if you evacuate the inside, there is no stress inside it.
<SpeedEvil> on it
<whitequark> I've been thinking that the pressure inside could be structurally important
<whitequark> Not sure about it
<whitequark> also, how are you going to get a vacuumtight connection to the neck (much less add your instruments)?
<SpeedEvil> vacuum epoxy
<whitequark> ew.
<whitequark> get a TIG welder and a chunk of pipe
<SpeedEvil> yes, that's not happening.
<whitequark> why the hell not?
<SpeedEvil> - as in it'll suck in the very thin metal
<SpeedEvil> and you'd need to reevacuate it.
<SpeedEvil> Actually - silver solder
<whitequark> no, I mean, make the chamber out of the chunk of pipe.
<SpeedEvil> ah
<whitequark> with the appropriate holes for your instruments and such.
<SpeedEvil> yeah
<whitequark> I don't even know what you would do with that thermos
<whitequark> also, duh, you already have a vacuum pump, what's evacuating some crappy thermos
<SpeedEvil> And yes - that was an aside.
<SpeedEvil> I have plans for a nice large wooden vacuum chamber.
<whitequark> what
<SpeedEvil> :)
<whitequark> are you nuts
<SpeedEvil> Foil-lined.
<whitequark> can you please give me some of the substance you're on.
<SpeedEvil> Structurally, it's not that bad.
<SpeedEvil> Damn. Now I'm wondering about vacuum evacuated thermally treated wood as an insulant.
<whitequark> ...
<whitequark> (though they do it with some other materials that sound similarly implausible)
<SpeedEvil> Heat wood to 250C for a couple of hours under steam and it loses most of the volatiles. Probably not enough.
<whitequark> like... aerogel?
<whitequark> a vapor pressure of a few torr won't noticeably affect the thermal transfer characteristics
<SpeedEvil> I want to make a large chamber - a large slice of a meter cubed. I have the wood to do this with. So why not.
<SpeedEvil> whitequark: It does actually.
<whitequark> does it?
<SpeedEvil> whitequark: ~1Pa fucks up thermoses.
<whitequark> huh.
<SpeedEvil> Until you get so that the mean free path exceeds the gap, thermal conduction doesn't go down meaningfully
<whitequark> that makes sense.
<whitequark> SiC heater, $8.
<SpeedEvil> yeah - I was meaning more ~3kW :)
<whitequark> oh
<whitequark> the desc suggests putting it in a microwave
<whitequark> I love how the dude posts something like an alumina plate and then adds to every single description "Your imagination is the only limiting factor in the spectrum of possible applications for this component."
<SpeedEvil> There are _lots_ of venfors on alibaba
<whitequark> alibaba wants volume, no?
<SpeedEvil> rather
<SpeedEvil> whitequark: Well - it depends.
<SpeedEvil> It's china.
<whitequark> china is a fucking pain to work with
<SpeedEvil> I got a quote of only twice spot for a couple of kilos of lithium
<SpeedEvil> from a vendor on alibaba
<whitequark> ... why do you need that much lithium
<whitequark> also, how much is that?
<whitequark> and how pure?
<SpeedEvil> I wanted to make a battery
<SpeedEvil> Take lead/antimony, float the lithium on it with LiCl between
<SpeedEvil> (under argon at 400C)
<SpeedEvil> Voila, battery.
<whitequark> local shop quotes me $150/kg at MOQ
<SpeedEvil> For lithium?
<whitequark> yeah
<SpeedEvil> I want that local shop
<whitequark> assuming you figure out how you ship HAZMAT out of RU, I will send it to you
<SpeedEvil> I stalled on that project when I realised that my naive implementation was potentially vulnerable if it got knocked to spontaneous discharge.
<whitequark> they also have very interesting things
<whitequark> they mail-order hydrazine (!!!)
<SpeedEvil> Which involves ~15 bar of lead/antimony/lithium/fluorine/chlorine at 2200C
<SpeedEvil> If it goes all the way
<whitequark> Ohhh yeah that high-temperature battery. IIRC there was also a design based on Fe?
<SpeedEvil> whitequark: Is actual Li hazmat?
<whitequark> it is hazmat.
<whitequark> batteries have metallic Li, which is why they are also hazmat.
<SpeedEvil> In metallic form?
<whitequark> it reacts with water and gives off a lot of hydrogen
<whitequark> no shit it is hazmat
<SpeedEvil> I gues.
<whitequark> oh I know this eastern european vendor on ebay, they ship calcium and sodium (in cans under argon) internationally
<whitequark> they are fucking nuts
<whitequark> lt's see if they have lithium for you
<SpeedEvil> Interesting.
<SpeedEvil> Unfortunately my 'I can do this' cell was basically large trays of the molten components with no seperators.
<SpeedEvil> While this in principle works, it gets way more complex if I need to do it properly - once I realised that shorting is an actual concern
<SpeedEvil> The metals can mix at a poorly limited very rapid rate
<SpeedEvil> dumping all their heat
<whitequark> Just stand a few meters back, upwind
<whitequark> And don't forget to record it and put on Youtube
<SpeedEvil> ~80kg, ~15 bar and 2000C of heavy metals and fluorine/chlorine/... is more in the chemical weapons class than 'just stand upstream.
<whitequark> Ok. Agreed. Put it on liveleak instead
<SpeedEvil> It is _really_ nice in principle as a home storage battery
<whitequark> $100/90g
<whitequark> But they have some very interesting metals.
<SpeedEvil> It is entirely molten state, and there is no aging mechanism
<SpeedEvil> And it has really good charge efficiency
SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
SpeedEvil has joined #homecmos
SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
SpeedEvil has joined #homecmos
<bofh__> lithium is weird.
<bofh__> it has a stable nitride. that's a superbase capable of deprotonating dihydrogen.
superkuh has quit [Read error: Connection reset by peer]
superkuh has joined #homecmos
nmz787_ has joined #homecmos
nmz787 has quit [*.net *.split]
GeorgeHahn has joined #homecmos
GeorgeHahn has quit [Read error: Connection reset by peer]
<_Sync_> whitequark: for custom feedthroughs just use macor
<_Sync_> and you can buy umicore brazetec CB4 in not really large quantities
kristian1aul has quit [Ping timeout: 272 seconds]
kristianpaul has joined #homecmos
chris_99 has joined #homecmos
_whitelogger has joined #homecmos
<berndj> TL;DR for the last 24 hours or so? vacuum induction furnace, shipping hazmat around the world, sintering?
_whitelogger has quit [Excess Flood]
_whitelogger has joined #homecmos
<whitequark> yes
<whitequark> _Sync_: cb4 is where i got the composition, yes
<whitequark> how does one buy it?
<_Sync_> write them a mail and get a kg of rods
<whitequark> hrm
<whitequark> still need to braze in vacuum though
<whitequark> or inert gas
<whitequark> macor is wetted by CB4 as well?
<_Sync_> I think so
<whitequark> hahah $200 for 100x100x3 sheet of macor
<whitequark> I think no
<whitequark> /maybe/ CB4 makes sense, though given how easy it would be to make, probably not even that *shrug*
<whitequark> also I am reading this
<whitequark> "MACOR® can be considered unique as its composition comprises 55% fluorophlogopite mica and 45% borosilicate glass"
<whitequark> which makes me doubt that the Ti content in CB4 will help wetting it
<whitequark> ah, no, it is.
<whitequark> I wonder how.
<whitequark> huh, you are actually also supposed to braze it specifically in vacuum
<whitequark> inert gas apparently doesn't work
<whitequark> even with a getter to sequester oxygen, too
<_Sync_> well if you need feedthroughs, buy them
<_Sync_> getting the ceramics and developing the workflow to deal with them makes it not worth it
<_Sync_> I investigated that 5 years ago
<whitequark> dunno we'll see about that
Noxz has quit [Ping timeout: 250 seconds]
<_Sync_> well, I tried to make them for cheaper
<_Sync_> but the whole process is too involved
<_Sync_> and then you need a helium leak detector to actually qualify them
<whitequark> now *that* can be easily done commercially
<_Sync_> sure, but is expensive
<whitequark> semi-related, I found some interesting device, developed in RU
<whitequark> which is, handheld helium leak detector
<whitequark> looks just like the freon ones, a bit larger. I wonder how they made it.
<whitequark> down to 1e-7 Pa*m3/s
Noxz has joined #homecmos
<whitequark> haha, they have a haloalkane leak detector which looks just like and has the exact same list of features as this one
<whitequark> so... they took a thermal conductivity based haloalkane one and recalibrated it for He?
<whitequark> does that... work?
Noxz has quit [Ping timeout: 240 seconds]
<whitequark> hum, no, nope. found the patent. the patent says it uses a membrane pump in combination with an ion getter pump
<whitequark> the closest US patent is https://www.google.com/patents/US7062954
<whitequark> the difference of RU patent from US patent is that they pump the sampled gas inside the case and around the capillary that is permeable to helium, thus making the gradient larger and more easily detectable
<whitequark> costs $6k new
<berndj> SpeedEvil, what's the salt in that Li / Pb-Sb cell?
<_Sync_> whitequark: I don't think that it works at the levels we need
<whitequark> mm, I see
<whitequark> I suppose the point is moot as my plan is to get an RGA anyway
<SpeedEvil> berndj: Li*
<SpeedEvil> berndj: Chloride, bromide, iodide and fluoride
<SpeedEvil> I think maybe not bromide
<SpeedEvil> You can leave some out, but the operating temperature then goes up
<whitequark> SpeedEvil: well, at least those halogens are bound, and you aren't getting a metal-fluorine fire
<whitequark> merely a metal-chlorine fire
<SpeedEvil> whitequark: well, in the worst case that I can't exclude - what happens is you get a propagating metal-metal mixing, which then heats to ~2800C or so.
<SpeedEvil> This then rapidly boils the electrolyte
<SpeedEvil> (the salts)
<SpeedEvil> It's perhaps technically not free Fluorine.
<SpeedEvil> And explodes the container
<_Sync_> fun!
mrdata_ has quit [Ping timeout: 264 seconds]
Noxz has joined #homecmos
Noxz has joined #homecmos
Logxen has quit [Changing host]
Logxen has joined #homecmos
mrdata has joined #homecmos
rk[1]_ has joined #homecmos
diginet has quit [Ping timeout: 250 seconds]
rk[1] has quit [Ping timeout: 250 seconds]
diginet_ has joined #homecmos
diginet_ has quit [Remote host closed the connection]
diginet has joined #homecmos
Textmode has joined #homecmos
chris_99 has quit [Remote host closed the connection]
mrdata has quit [Ping timeout: 240 seconds]
<_Sync_> hmm the new scope is actually pretty usable
SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
SpeedEvil has joined #homecmos
<SpeedEvil> :)