sb0 changed the topic of #m-labs to: ARTIQ, Migen, MiSoC, Mixxeo & other M-Labs projects :: fka #milkymist :: Logs http://irclog.whitequark.org/m-labs
ylamarre has quit [Quit: ylamarre]
cr1901_modern has quit [Quit: Leaving.]
cr1901_modern has joined #m-labs
<GitHub115> [migen] jordens pushed 1 new commit to master: http://git.io/vqg53
<GitHub115> migen/master 8d6aa82 Robert Jordens: mibuild/openocd.py: add support...
<rjo> ysionneau: notification on gitter.im
travis-ci has joined #m-labs
<travis-ci> m-labs/migen#54 (master - 8d6aa82 : Robert Jordens): The build passed.
travis-ci has left #m-labs [#m-labs]
FabM has quit [Remote host closed the connection]
* whitequark grumbles really really hard
<whitequark> turns out i absolutely soaked my turbopump in forepump oil
<sb0_> how did that happen?
<whitequark> no check valve on forepump
<whitequark> I have this new large chamber. pumped it down. turned turbo pump off
<whitequark> then opened the vent
<whitequark> however, the vent on the turbo pump is tiny and the chamber is large
<whitequark> so the huge volume of the chamber has sucked oil from the forepump
<whitequark> it all went in the bottom bearing oilpan and pretty much destroyed it
<whitequark> not to mention that the whole pump is splattered in oil
<sb0_> why didn't you keep the forepump running while venting?
<whitequark> because I'm dumb
<whitequark> anyway, I was able to get the oil out of it completely using large amounts of HCFC-141b for flushing
<whitequark> the problem is now the sponge that used to contain the oil for the bottom bearing, which is now pretty much destroyed
<whitequark> 141b is pure magic. You splash it on a surface that's covered with a half-polymerized mix of two oils that isn't even washed away with regular soap--it still leaves a thick film--just splash it
<whitequark> and in a few seconds you're looking at a mirror shiny surface without any trace of oil, where the only thing you see is tool marks
<sb0_> do they still sell CFC refrigerants in Russia?
<whitequark> sure
<whitequark> "it's so good, it's illegal" would make a good marketing slogan *ahem*
<whitequark> really a pity there are few functional replacements for CFCs.
kyak has quit [Ping timeout: 276 seconds]
kyak has joined #m-labs
kyak has joined #m-labs
<chiggs> I've always wondered what on earth @whitequark is up to and come to the conclusion you're a real life VXJunkie?
<whitequark> I have no idea what/who that is
<cr1901_modern> I'm very confused and scared reading the subreddit and wiki
<chiggs> There's a similarity between 09:00:19 whitequark | then opened the vent │
<chiggs> sorry
<chiggs> There's a similarity between whitequark's twitter posts and https://www.reddit.com/r/VXJunkies
<chiggs> I've been trying to guess what you're building but I've given up
<ysionneau> if you were to know what he is building, you would call the police :p
<cr1901_modern> I would guess whitequark is building a lot of things, culminating in a mini fab for CMOS chips inside his own home.
<cr1901_modern> In any case, it looks like the VX subreddit is an elaborate joke, designed to poke fun at ppl who use too much technical jargon.
<whitequark> chiggs: lol
<whitequark> so far just a vacuum chamber
<chiggs> and a fridge?
<chiggs> just your standard household appliances pretty much
<whitequark> well, yes
<cr1901_modern> whitequark: Mini fab?
<whitequark> the refrigerator will drive a cooling circuit
<whitequark> the turbo pump requires cooling, at least
<whitequark> some of the things I want to do inside the chamber, e.g. an induction oven, also would require cooling
<whitequark> I could, like, connect it to cold tap water, but that's just annoying to run around
<chiggs> is all of that hobby or work related? Sounds like quite an impressive project
<whitequark> I also kind of wanted to make a refrigerator
<whitequark> I do not have hobbies.
<cr1901_modern> Your interest in homecmos (well, _whitelogger_ monitors the room anyway) isn't a hobby?
<whitequark> I dunno. I'm not the one trying to classify it
<kyak> if you do it in your leisure time, and you enjoy it, this is hobby
<sb0_> rjo, we're not going to use the built-in AD9914 PLL, right?
sb0__ has joined #m-labs
_florent_ has joined #m-labs
sb0_ has quit [Ping timeout: 246 seconds]
sb0__ has quit [Ping timeout: 264 seconds]
sb0 has joined #m-labs
_florent_ has quit [Ping timeout: 246 seconds]
_florent_ has joined #m-labs
whitequark has quit [Ping timeout: 248 seconds]
whitequark has joined #m-labs
ylamarre has joined #m-labs
ylamarre1 has joined #m-labs
ylamarre1 has quit [Client Quit]
ylamarre1 has joined #m-labs
ylamarre has quit [Ping timeout: 264 seconds]
key2 has joined #m-labs
ylamarre1 has quit [Ping timeout: 246 seconds]
ylamarre has joined #m-labs
_florent_ has quit [Ping timeout: 255 seconds]
<sb0> rjo, in programmable modulus mode, the maximum denominator you can have is (2**32-1)*2**32, not 2**64
<GitHub20> [artiq] sbourdeauducq pushed 1 new commit to master: http://git.io/vqwzF
<GitHub20> artiq/master 34aacd3 Sebastien Bourdeauducq: complete AD9914 support (no programmable modulus, untested)
<sb0> we can do a 63-bit FTW, but I'm still unsure about the problems of integer overflow/loss of precision when computing it...
<sb0> well, the current floating point FTW equations would produce correct but quite imprecise results with 63 bits. I guess that's good enough?
travis-ci has joined #m-labs
<travis-ci> m-labs/artiq#298 (master - 34aacd3 : Sebastien Bourdeauducq): The build has errored.
travis-ci has left #m-labs [#m-labs]
ylamarre has quit [Read error: Connection reset by peer]
ylamarre has joined #m-labs
key2 has quit [Ping timeout: 246 seconds]
<whitequark> annnnnnnd the pump works
<whitequark> at least, spins up to ultimate rpm and bottoms out my pirani
<sb0> so this turbopump is using mechanical bearings? no fancy magnetic levitation?
<sb0> I got an ebara turbopump from the dumpster once, which did have magnetic bearings... and dozens of pins for control, no controller, and as with most japanese companies a very polite "fuck you" is the answer (if any) you get when asking for docs
travis-ci has joined #m-labs
<travis-ci> m-labs/artiq#298 (master - 34aacd3 : Sebastien Bourdeauducq): The build has errored.
travis-ci has left #m-labs [#m-labs]
mumptai has joined #m-labs
ylamarre has quit [Quit: ylamarre]
balrog has quit [Ping timeout: 245 seconds]
balrog has joined #m-labs
<rjo> sb0: no. the ad9858 also has a pll. nobody is using it even though it is an extremely cool feature (noce pfd in there).
<rjo> sb0: take 1<<63 as the denominator. the 64-bit behavious is not a problem if you take machine units to be w.r.t. a 1<<64 denominator, set the lsb to zero for accurate calculations and then just shift right before writing the numerator.
<sb0> exposing 63 is much easier
<sb0> integers are signed in py2llvm, so 64-bit unsigned is a problem
<rjo> fine. my vague guess is (havent checked) that the negative ftws will alias nicely.
<rjo> for a 1<<63 denominator.
<rjo> i remember having checked that for a similar problem many years ago but i would have to go through the math again.
<rjo> ... 1<<63 denom ftw on a 64 bit signed integer.
sb0 has quit [Read error: Connection reset by peer]