| --- Log opened Thu Nov 22 00:00:24 2012 | ||
| -!- Juul [~Juul@50-0-146-204.dsl.dynamic.sonic.net] has quit [Read error: Connection reset by peer] | 00:03 | |
| -!- sylph_mako [~mako@103-9-42-1.flip.co.nz] has quit [Ping timeout: 246 seconds] | 00:11 | |
| -!- Juul [~Juul@50-0-146-204.dsl.dynamic.sonic.net] has joined ##hplusroadmap | 00:20 | |
| -!- sylph_mako [~mako@103-9-42-1.flip.co.nz] has joined ##hplusroadmap | 00:25 | |
| -!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has quit [Quit: Leaving] | 00:26 | |
| -!- Juul [~Juul@50-0-146-204.dsl.dynamic.sonic.net] has quit [Read error: Connection reset by peer] | 00:35 | |
| -!- Simurg [~Shehrazad@95.5.88.151] has joined ##hplusroadmap | 00:36 | |
| -!- Simurg [~Shehrazad@95.5.88.151] has quit [Changing host] | 00:36 | |
| -!- Simurg [~Shehrazad@unaffiliated/shehrazad] has joined ##hplusroadmap | 00:36 | |
| -!- ElixirVitae [~Shehrazad@unaffiliated/shehrazad] has quit [Ping timeout: 252 seconds] | 00:37 | |
| -!- sylph_mako [~mako@103-9-42-1.flip.co.nz] has quit [Ping timeout: 264 seconds] | 00:42 | |
| abetusk_ | is anyone up? | 00:48 |
|---|---|---|
| fenn | nobody here but us degenerates | 00:48 |
| fenn | that's pronounced "duh jean hurts" | 00:49 |
| fenn | ow | 00:50 |
| abetusk_ | fenn, are you involved in creating kanzure and nmz787's laser cutter? | 00:51 |
| fenn | yes i had some input on the motion stage design, fwiw | 00:51 |
| abetusk_ | do you know how it's going? | 00:52 |
| fenn | nathan is going to be testing some ideas using a FIB he has access to | 00:52 |
| fenn | there's quite a bit of process stuff that needs to be worked out, completely separate from the laser part | 00:53 |
| fenn | how to cast and bond layers of PDMS etc | 00:53 |
| abetusk_ | even getting a single layer cut would be a huge step forward | 00:53 |
| fenn | i was looking at some samples cut on a friend's universal laser systems .. er, laser. | 00:55 |
| fenn | it had awful aliasing on the curves, i was really surprised | 00:56 |
| fenn | it's like they didnt have enough bits in the position counter | 00:56 |
| abetusk_ | what was the system and what was the accuracy and resolution? | 00:56 |
| -!- sylph_mako [~mako@103-9-42-1.flip.co.nz] has joined ##hplusroadmap | 00:58 | |
| fenn | hm. sometimes the internet is not helpful | 01:02 |
| fenn | it was gray and red, at least ten years old. cutting size 1000 by 1500mm or so. it was probably an xl-9200 or xl-12000 | 01:03 |
| fenn | they both have resolution of 1016 dpi, but i don't believe it | 01:04 |
| fenn | he was saying the shopbot had the same software bug | 01:06 |
| fenn | it's something in the motion interpolation of arcs | 01:06 |
| fenn | they come out looking more like small 45 degree lines joined together | 01:06 |
| abetusk_ | interesting. All the more reason to use something like grbl or linuxcnc | 01:07 |
| fenn | about 0.2mm long i'd guess | 01:07 |
| abetusk_ | yeah, 10 mil is not so great for the type of work you want to be doing | 01:07 |
| fenn | nathan seems to dislike linuxcnc for reasons i can't understand | 01:10 |
| fenn | or perhaps he's just more excited about using the "propeller" at any cost | 01:11 |
| fenn | it always seemed like someone's first attempt at getting a chip fabbed, a neat idea but not something to seriously use in anything you want to work consistently | 01:12 |
| fenn | also i hate the whole basic stamp scene for how they dumb down everything | 01:12 |
| abetusk_ | propeller is still using basic? | 01:13 |
| fenn | anyway that's just my impression of it, i don't have any valid complaints | 01:13 |
| fenn | there's a c compiler for it. i'm not sure what the best way to program 8 cores is | 01:13 |
| fenn | they have asm and also "spin" | 01:14 |
| abetusk_ | I'm pretty agnostic. I like grbl, but that's because it's simple, cheap and I've used it. The idea of making a custom kernel just to communicate to the stepper drivers seems a little backward, but that's just my opinion | 01:14 |
| abetusk_ | I'm happy to use whatever tool is appropriate for the job | 01:14 |
| fenn | nah a lot of people misunderstand what linuxcnc is about | 01:14 |
| fenn | it's not really for sending pulses to stepper drivers, though most people use it for that | 01:15 |
| fenn | the idea is to be able to do complex calculations in real time such as you'd need to control non-cartesian robots, and to be able to use any mix of servos, steppers, galvanometers, what have you | 01:15 |
| fenn | also you have to remember when linuxcnc was first created, there were no microcontrollers capable of even running a cartesian robot | 01:16 |
| abetusk_ | ah interesting. Yes, that does make more sense | 01:17 |
| fenn | er, well they had some pretty ambitious motion planning algorithms as well | 01:18 |
| fenn | anyway it has a long history and may not be the best thing for running a small mill | 01:18 |
| abetusk_ | Yeah, for a beginner mill, grbl is probably just fine | 01:19 |
| fenn | but it's quite thoroughly tested and does what it's supposed to | 01:19 |
| fenn | i'm somewhat less enthused by this caveat: "Grbl renders circles and arcs by subdividing them into teeny tiny lines. You will probably never need to adjust this value – but if you find that your circles are too crude (really? one tenth of a millimeter is not precise enough for you? Are you in nanotech?) you may adjust this. Lower values gives higher precision but may lead to performance issues." | 01:22 |
| fenn | that's for 0.8; i believe 0.7 interpolates arcs in realtime | 01:23 |
| fenn | why would you ever hardcode something like that? | 01:23 |
| fenn | i like linuxcnc's method of specifying (G64) the tolerance of the path, and letting the controller figure out how far it can cut corners | 01:24 |
| abetusk_ | well, the nice thing about grbl (and linuxcnc as well) is that you can extend it to your needs. | 01:25 |
| abetusk_ | I wanted feedback from my micro running grbl so I altered grbl and now it has a command that gives feedback | 01:25 |
| fenn | cool. what kind of feedback? | 01:26 |
| abetusk_ | The main feature was the positional feedback | 01:27 |
| abetusk_ | but it also reports what the current feed and seek rate are. | 01:28 |
| fenn | you have encoders or something? | 01:28 |
| abetusk_ | One of my annoyances with grbl was that you could set the default feed and seek rate but that's of course separate from the current feed and seek rate. | 01:28 |
| abetusk_ | I do height sensing. I need to make sure it's finished it's move in the z direction before I give it the next one | 01:29 |
| fenn | ah there should be a M code for that | 01:29 |
| fenn | i guess grbl doesnt do M codes | 01:29 |
| abetusk_ | hmm, not sure I looked all that closely. What m code is it supposed to be? | 01:30 |
| nmz787__ | hey skorket | 01:31 |
| nmz787__ | er abetusk_ | 01:31 |
| abetusk_ | heh, you got me | 01:31 |
| fenn | i think it's called a probe move, similar to homing with a switch | 01:31 |
| nmz787__ | know anyone that needs nanoscale milling done? I have access to a FIB now | 01:31 |
| nmz787__ | abetusk_: http://web.pdx.edu/~pmoeck/phy381/fib.pdf | 01:32 |
| nmz787__ | was you goal an overall mill, or just pcb fab? | 01:35 |
| fenn | looks like they haven't implemented probe moves https://github.com/grbl/grbl/wiki/Development-Path-and-Future-Needs | 01:36 |
| fenn | "Up to debate." | 01:36 |
| abetusk_ | nmz787__, nice | 01:37 |
| abetusk_ | I wanted both really. The focus was always on rapid prototyping for PCBs but I was also hoping to have a general mill | 01:38 |
| abetusk_ | I just bought a ShapeOko for the hackerspace that's starting up, so maybe I'll use that for courser stuff and general use and try to keep mine to just PCBs | 01:38 |
| abetusk_ | but we'll see | 01:38 |
| fenn | i'd like to move toward direct inkjet masking and surface mount for most things | 01:39 |
| fenn | through hole is actually denser if you can't do plated vias, but it's crusty looking | 01:40 |
| fenn | and you have to drill all those holes | 01:40 |
| fenn | better to do it in software with topo router :P | 01:40 |
| fenn | someone wrote a topo router for PCB (gEDA) but i haven't figured out how to use it | 01:41 |
| abetusk_ | ah, check this out: https://www.youtube.com/watch?v=PhH05jNyjCk | 01:41 |
| abetusk_ | and their solution to the via problem: https://www.youtube.com/watch?v=qnOVQR95NB0 | 01:41 |
| abetusk_ | problem with direct inkjet is that you still have to etch with nasty chemicals. Resolution is probably much better though | 01:42 |
| fenn | is it actually cutting copper with the laser? why do they raster like that? | 01:42 |
| abetusk_ | from what I understand, it's vaporizing the copper, so yes. | 01:43 |
| abetusk_ | You're asking why the laser seems to be going in a diagonal like that? | 01:43 |
| fenn | yeah why not just cut around the traces | 01:44 |
| fenn | i thought copper was too reflective and too heat dispersive to cut cleanly | 01:44 |
| abetusk_ | that's a good question. I do seem to remember reading somewhere that they're actually vaporizing spots on the copper. Maybe it's for heat dissipation? | 01:44 |
| fenn | you had to coat it with something black to get enough heat into ti | 01:44 |
| abetusk_ | I think they're using a Nd:YAG that's been doubled or quadrupled down to 250 or 500 nm | 01:45 |
| fenn | yeah the spot in the video was green | 01:45 |
| abetusk_ | past 500nm or so, copper reflects too much | 01:45 |
| abetusk_ | so I thought that too, but apparently it can be done | 01:45 |
| fenn | hmm. anyway, i dont mind an etching step. most people who are into lasers and mills dont even know about newer etchants like persulfate or cuprous chloride | 01:47 |
| fenn | or is it cupric chloride | 01:47 |
| fenn | anyway, add copper sulfate to vinegar and salt, bubble air into it | 01:48 |
| abetusk_ | the latter, I believe. Muriatic acid + peroxide, right? | 01:48 |
| fenn | yes | 01:48 |
| fenn | its the cuprous ion that does the actual etching, formed in equilibrium with the cupric ion | 01:48 |
| abetusk_ | I looked into using etchants, but all the nasty chemicals are a big hassle for me | 01:50 |
| fenn | one of these days some kid will come up with an automatic point to point wire bonder and we can forget about all this pcb nonsense | 01:50 |
| fenn | it's just bleach, i dont see what the big deal is. it doesn't even smell bad | 01:51 |
| fenn | then you add aluminum foil and get your copper back | 01:52 |
| fenn | oh also you might not know that you can run flexible kapton copper foil directly through a laser printer | 01:53 |
| fenn | now i'm wondering how good the alignment is on duplex printing.. | 01:54 |
| fenn | it's hard to beat lasers for drilling small holes though | 01:56 |
| fenn | need a FIB for that :P | 01:56 |
| nmz787__ | why am i not finding hits for fib mask mastering gold | 01:57 |
| nmz787__ | (relevant ones) | 01:57 |
| nmz787__ | i'm pretty sure someone has coated glass or quartz with gold and then tried ablating in a fib before | 01:57 |
| fenn | why gold in particular? | 01:59 |
| nmz787__ | there's a gold sputterer there | 01:59 |
| fenn | http://proceedings.spiedigitallibrary.org/proceeding.aspx?articleid=1310840 | 02:00 |
| fenn | not exactly a howto | 02:00 |
| fenn | http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4966198 | 02:01 |
| fenn | oh that's deposition | 02:02 |
| nmz787__ | i also learned of this company today, i may go visit them next week since i'm flying back to rochester (which is where they are) | 02:02 |
| nmz787__ | http://www.stampertech.com/ | 02:03 |
| nmz787__ | they were making 3 micron wide stuff that was around 100nm tall using some interference direct exposure thing | 02:03 |
| fenn | hey just convert your microfluidic design to polar coordinates | 02:04 |
| nmz787__ | some tech ultralord shit | 02:04 |
| fenn | UV laser cured? what's interference direct exposure supposed to mean? | 02:04 |
| nmz787__ | well apparently he can produce 8 micro bibles in under half an hour | 02:04 |
| nmz787__ | some version of this maybe http://www.stampertech.com/holo2005.html | 02:06 |
| nmz787__ | there's a patent | 02:06 |
| fenn | oh i see | 02:07 |
| fenn | it's just for disc graphics | 02:07 |
| fenn | because holograms can't be etched with a digital process | 02:08 |
| nmz787__ | http://www.google.com/patents?id=kqTZAAAAEBAJ&printsec=abstract&zoom=4#v=onepage&q&f=false | 02:08 |
| fenn | you need curvy slopes or the fourier spectrum is all whack | 02:08 |
| nmz787__ | huh? | 02:08 |
| fenn | ffs google.. fucking toolbar.. there's like 1/6th of the screen actually showing content, the rest is navigation and google plus cruft | 02:09 |
| nmz787__ | i'm seeing more like 1/3 content | 02:09 |
| nmz787__ | but anyway if he can make 8 micro bibles in under 30 mins that's pretty good | 02:10 |
| nmz787__ | so if its pretty cheap, i can make my macro to micro to nano manifold with that tech, and make all the reactor stuff in the nanoscale | 02:11 |
| nmz787__ | if the stampertech isn't right or cheap, then i'll just get some 8000dpi film printed for lithography for now | 02:12 |
| nmz787__ | at least i'll get to making some channels | 02:13 |
| nmz787__ | i'm not sure how to do my reaction center size optimization | 02:13 |
| fenn | bah why am i reading this patent, it's not meant to be understood | 02:14 |
| nmz787__ | like there are reagent costs which decrease, but manufacturing cost and probably dev time increase | 02:14 |
| nmz787__ | (as you go down in scale) | 02:14 |
| nmz787__ | and then there is fucking layer registration | 02:14 |
| fenn | yeah precision is easy, it's the accuracy that's a bitch | 02:15 |
| nmz787__ | that's one reason alone for not wanting to manufacture microfluidics | 02:15 |
| fenn | you can make features small if they dont interact with other layers | 02:15 |
| fenn | does it matter if, say, a via is 10x the size of the surrounding components? | 02:16 |
| fenn | hmm they're claiming a million dots per second? that's pretty fast | 02:18 |
| fenn | and yet burning a dvd is 60x faster still | 02:20 |
| fenn | hard to wrap your head around nano speeds | 02:21 |
| fenn | how many bits do you need to specify positions in a complete reactor? if one were to use galvanometers to move a laser spot around | 02:22 |
| fenn | or a FIB or something like that | 02:23 |
| nmz787__ | specify positions? | 02:23 |
| fenn | so if your resolution is 1mm and you have 8 bits you can make something up to 256 mm wide | 02:24 |
| nmz787__ | the reaction chamber is just a line with a few branching lines, one for each chemical needed, one end of the main line is rinse other end is waste | 02:24 |
| fenn | if resolution is 1nm and 8 bits maximum of 256nm wide, and so on | 02:25 |
| nmz787__ | ahh | 02:25 |
| fenn | it's a 2d object more or less, and you can trade bits on each axis, but no real reason to do that | 02:26 |
| nmz787__ | i'd say a reactor would be a few handuls of ten bytes then | 02:26 |
| -!- sylph_mako [~mako@103-9-42-1.flip.co.nz] has quit [Ping timeout: 255 seconds] | 02:26 | |
| nmz787__ | handfulls* | 02:26 |
| fenn | really? that seems way too high | 02:26 |
| fenn | please jog my memory, what's a reasonable channel width? | 02:27 |
| nmz787__ | well i think resolution is more like 10 or 20 nm | 02:27 |
| nmz787__ | and i think 100nm channels are small enough to start with | 02:27 |
| abetusk_ | ok, going to bed. Have a good TG | 02:27 |
| nmz787__ | 100nm wide that is | 02:27 |
| nmz787__ | i figure it will have to be 10 to 20 times as long as it is wide | 02:28 |
| nmz787__ | so width is under a byte | 02:29 |
| nmz787__ | and i guess at 20x width long thats still under a byte | 02:29 |
| nmz787__ | right? | 02:29 |
| fenn | 2000nm at 10nm resolution is about a byte, yes | 02:30 |
| nmz787__ | but since you'd need rows, you'd have a byte * 5 | 02:30 |
| nmz787__ | at 20nm res | 02:30 |
| fenn | rows? | 02:31 |
| nmz787__ | well its a raster image | 02:31 |
| nmz787__ | so umm | 02:31 |
| nmz787__ | err | 02:31 |
| fenn | i'm still not sure why it has to be a raster | 02:32 |
| nmz787__ | guess its just like a TV | 02:32 |
| nmz787__ | sortof | 02:33 |
| nmz787__ | or at least it needs an XY | 02:33 |
| fenn | ever used a vector oscilloscope? | 02:33 |
| nmz787__ | and PULSE command | 02:33 |
| nmz787__ | or FIRE | 02:33 |
| nmz787__ | etc | 02:33 |
| fenn | FIRE THE ION CANNON | 02:33 |
| nmz787__ | no | 02:34 |
| fenn | electron beams have no inertia so raster scan isn't any faster than some other pattern | 02:35 |
| fenn | it's the area you cover that matters | 02:35 |
| fenn | also there's some hysteresis lag i guess.. i've never built a TV so i'm sorta making it up as i go | 02:35 |
| fenn | either way, vector or raster, you're going to have some misalignment at the edges if you need to reposition part way through an etching | 02:36 |
| fenn | so it's best to do the whole thing in one go | 02:37 |
| fenn | but to do that you need lots of bits and very accurate DAC's | 02:37 |
| fenn | er, very low noise DACs at least | 02:37 |
| fenn | and low noise amplifiers, and power supplies, etc | 02:38 |
| fenn | it would be nice if we could do realtime alignment to the image itself | 02:40 |
| nmz787__ | i'm not sure how the zoom works exactly, I think its just a digital pre-amp after the DAC but before the electrostatic plates/lens | 02:40 |
| fenn | so it's not actually moving around? | 02:40 |
| nmz787__ | well i figure the smalles piece of tubing will probably need 1 sq mm by itself | 02:40 |
| nmz787__ | well the most far out zoom electrically is 1 sq mm visible, but then the steps for the beam are much bigger | 02:41 |
| fenn | i figured the sample was on some kind of xy stage | 02:41 |
| nmz787__ | and that is why the image is less-zoomed in | 02:41 |
| nmz787__ | it is also on an XY stage | 02:41 |
| nmz787__ | i'm not sure how accurate that is | 02:41 |
| fenn | so, maybe you can first etch a grid at the wide zoom setting, then zoom in and realign to the grid marks | 02:42 |
| nmz787__ | but you could maybe setup a grid of fiducials that you can just pan around in | 02:42 |
| nmz787__ | well i'm not sure the electric zoom stuff needs fiducials | 02:43 |
| nmz787__ | but i def think the stage positioning would | 02:43 |
| nmz787__ | and i figure the smallest piece of macroscale tubing is gonna require a whole sq mm by itself | 02:43 |
| fenn | oh but you're just connecting a few tubes together | 02:44 |
| fenn | macro -> micro -> nano | 02:44 |
| nmz787__ | so i think i'm gonna need like 5x5 1 mm sqs | 02:44 |
| fenn | a fan-out i believe it's called | 02:44 |
| nmz787__ | right but if i don't do that fan out (macro->micro manifold) with lithography then i have to do it with the fib | 02:45 |
| fenn | oh fan-out means something else apparently | 02:45 |
| nmz787__ | and if i don't figure out some stage fiducial system, something like that 10cm long grating you want would never happen | 02:45 |
| fenn | i didnt know about the electrical pan/zoom | 02:46 |
| fenn | i guess you can also image before you etch, so you can see if it's aligned properly before pressing the "FIRE THE ION BEAM" button | 02:47 |
| nmz787__ | yeah full zoom i think field of view was low microns | 02:47 |
| fenn | is there automatic alignment software? | 02:48 |
| nmz787__ | not sure how old/new it is, and how 'auto' it is | 02:48 |
| nmz787__ | the software FEI is showing now on their website looks decent | 02:48 |
| fenn | if you have to manually realign 500 times for each chip, that kinda sucks | 02:48 |
| nmz787__ | but i dunno how much of that is only with their newer equipment | 02:49 |
| nmz787__ | but if it has to be done manually, i think we could figure out a system | 02:49 |
| nmz787__ | or, fix it with automagic somehow | 02:49 |
| fenn | you should try doing two of those gratings side by side (2mm x 5mm) and see how well the lines match up | 02:49 |
| fenn | the big lines should be continuous for sure, i dunno about the small lines | 02:50 |
| nmz787__ | what do you mean? | 02:50 |
| fenn | the variable pitch "chirp mask" | 02:50 |
| fenn | to test alignment | 02:51 |
| fenn | if it's misaligned in Y the lines will be out of phase | 02:51 |
| nmz787__ | ahh | 02:51 |
| nmz787__ | why would you be sure of the big lines being aligned? | 02:52 |
| fenn | since it covers several length scales you can see how well it aligns.. say the 1 micron line looks okay but the 100 nm line are out of phase | 02:52 |
| nmz787__ | capillary microbore OD 360 micron avail | 02:53 |
| fenn | eh? | 02:53 |
| nmz787__ | well how many 360 micron tubes can you jam reliably into a sq mm | 02:54 |
| nmz787__ | 4 or 5 maybe? | 02:54 |
| fenn | how long is a string | 02:54 |
| fenn | (how long are the tubes) | 02:54 |
| nmz787__ | i found a paper eariler using long neck PDMS flanges to aid in stabilizing tubing | 02:54 |
| fenn | oh, perpendicularly | 02:55 |
| nmz787__ | well the tubing needs to go from the chip area into a bottle of reagent i guess | 02:55 |
| nmz787__ | so 3-12 inches i'd say | 02:55 |
| fenn | does it need to go into the center of the chip? how about inserting them parallel to the chip layers? | 02:56 |
| fenn | you'd have much more area in contact with the tubing that way | 02:56 |
| fenn | say your layer thickness is 360 microns, then you just cut a 360 micron wide channel | 02:57 |
| nmz787__ | hmm | 02:58 |
| nmz787__ | well there is some thickness to the tubing | 02:58 |
| nmz787__ | the wall i mean | 02:59 |
| fenn | you're worried about leaking around the corners? | 02:59 |
| nmz787__ | that seems like it would only be worse if I only had to use the FIB for everything | 02:59 |
| fenn | channel cross section is square; i assume the tubing is circular | 02:59 |
| nmz787__ | because that would be more sq mm | 02:59 |
| nmz787__ | i think the long neck flanges just made a thickier PDMS layer | 03:00 |
| fenn | what's the limiting factor in FIB etching? | 03:00 |
| fenn | is it beam power? | 03:01 |
| nmz787__ | slow to get more than 50nm down into silicon | 03:01 |
| nmz787__ | yeah | 03:01 |
| nmz787__ | depends on the material | 03:01 |
| nmz787__ | if i wanted to master molds in silicon, it could take an hour to mill 10x20x10 micron rectanguloid | 03:01 |
| fenn | have you thought about etching other materials like, say, kapton | 03:01 |
| nmz787__ | hmm, no | 03:02 |
| fenn | it tends to sublimate rather than melt | 03:02 |
| nmz787__ | i am looking at some diamond stuff now | 03:02 |
| nmz787__ | in a FIB? | 03:02 |
| fenn | melting probably isnt a problem with FIB though | 03:02 |
| -!- ThomasEgi [~thomas@gw-ko-kostr2.inka-online.net] has joined ##hplusroadmap | 03:03 | |
| -!- ThomasEgi [~thomas@gw-ko-kostr2.inka-online.net] has quit [Changing host] | 03:03 | |
| -!- ThomasEgi [~thomas@panda3d/ThomasEgi] has joined ##hplusroadmap | 03:03 | |
| fenn | kapton's used in laser cutting because of how clean it cuts | 03:03 |
| nmz787__ | so the slow silicon etch rate is why i was looking for lithography mask mastering with the FIB | 03:03 |
| nmz787__ | because i figure you only need a 50-200nm of gold or so to block light | 03:04 |
| nmz787__ | i dont actually know, but that seems like plenty of atoms | 03:04 |
| fenn | heh | 03:05 |
| nmz787__ | so i would think mastering a mask would be quick to FIB, then just do lithography | 03:05 |
| nmz787__ | for the microscale stuff | 03:05 |
| nmz787__ | the nano would still need to be directly milled | 03:05 |
| fenn | you need to block the etchant i'd think, not light | 03:05 |
| nmz787__ | because its smaller than light :P | 03:05 |
| fenn | reactive ion etching | 03:05 |
| nmz787__ | huh? | 03:06 |
| nmz787__ | but its a beam of ions | 03:06 |
| fenn | basically you have an oxygen plasma that howls away at the entire surface of silicon | 03:06 |
| nmz787__ | when? | 03:06 |
| fenn | during litho etching | 03:06 |
| nmz787__ | no | 03:06 |
| nmz787__ | not all | 03:06 |
| fenn | http://en.wikipedia.org/wiki/Reactive-ion_etching | 03:07 |
| nmz787__ | i meant ablate gold from glass or quartz, then do contact lithography on SU-8 resist | 03:07 |
| fenn | why transfer the pattern around like that | 03:07 |
| nmz787__ | develop that, then pour silicone over that | 03:07 |
| nmz787__ | to avoid hours on the FIB | 03:08 |
| fenn | it's the same surface area | 03:08 |
| nmz787__ | or i make one master macro-micro-nano *breakout* board | 03:08 |
| nmz787__ | and then just make tiny iterations of nano boards | 03:09 |
| nmz787__ | that press up against the adapter | 03:09 |
| fenn | hmm | 03:09 |
| nmz787__ | seems like that adapter could/should be steel or glass | 03:09 |
| nmz787__ | but i guess if its easy to replicate, PDMS is fine too | 03:10 |
| fenn | yeah that's a good idea | 03:10 |
| nmz787__ | but i'm gonna check out this holographic guy first | 03:11 |
| fenn | you might even be able to bridge the micro-nano part on the nano chip, and then you only have to do macro-micro on the breakout board, which can be done with 3d printing or laser cutting | 03:11 |
| nmz787__ | it would be great if he could just make me a few adapter masters in less than a half hour | 03:12 |
| nmz787__ | since they don't need to go below micron scale | 03:12 |
| nmz787__ | yeah | 03:12 |
| fenn | i'm picturing tweezering some capillary tubing into place under a microscope | 03:13 |
| nmz787__ | like it just needs to accept some of those 360 micron or 510 micron OD tubes, and bring them down to like a 1-5 micron sized hole | 03:13 |
| nmz787__ | i think it'd be more like threading a needle | 03:14 |
| nmz787__ | you could probably use a scope | 03:14 |
| fenn | want to make sure the tube doesn't get clogged or broken | 03:14 |
| fenn | might even need glue | 03:14 |
| nmz787__ | alternatively, maybe the adapter is PDMS, and a tray hinges down on top, holding an array or capillary barbs | 03:14 |
| nmz787__ | like an IC holder | 03:15 |
| nmz787__ | socket | 03:15 |
| nmz787__ | i saw some 510 micron OD steel | 03:15 |
| fenn | that uses a lot of area on your chip though | 03:15 |
| fenn | you need pi*(510/2)^2 microns per tube | 03:16 |
| fenn | at least | 03:17 |
| fenn | twice that realistically | 03:17 |
| fenn | is 510 microns "micro" or "macro"? | 03:18 |
| nmz787__ | uhh, well | 03:18 |
| nmz787__ | is has the micro prefix | 03:18 |
| fenn | can you get smaller tubing? | 03:18 |
| fenn | like 50 microns? | 03:18 |
| nmz787__ | quick search showed 360 | 03:18 |
| nmz787__ | 50 micron ID 360 OD | 03:19 |
| -!- devrandom [~devrandom@gateway/tor-sasl/niftyzero1] has quit [Remote host closed the connection] | 03:20 | |
| fenn | ideally you'd have one piece that goes from finger-grabbable tubing to nano holes.. how about an inkjet head? | 03:21 |
| -!- devrandom [~devrandom@gateway/tor-sasl/niftyzero1] has joined ##hplusroadmap | 03:21 | |
| nmz787__ | those are pretty big though | 03:22 |
| fenn | 50 micron | 03:22 |
| nmz787__ | i dont think i could very easily position something less than 50 microns by hand | 03:22 |
| fenn | well, 10 to 100 micron | 03:22 |
| fenn | no probably not | 03:23 |
| nmz787__ | well i mean the whole thing is big | 03:23 |
| nmz787__ | 3 x 5 cm? | 03:23 |
| fenn | ah but the chip part is only a couple mm on a side | 03:23 |
| nmz787__ | can you disconnect them? | 03:23 |
| fenn | it's glued to a kapton manifold much like flex pcb | 03:23 |
| nmz787__ | from the ... reservoirs?? | 03:23 |
| nmz787__ | hmm | 03:24 |
| fenn | HP’s latest generation of printhead technology, ink chambers, fluidic features and nozzles are formed monolithically using photo-imageable thick-film polymers. All-lithographic patterning coupled with new advances in silicon surface and bulk micromachining techniques eliminates the need for mechanical alignment processes and results in fluidic features that have the precision and integration | 03:26 |
| fenn | scale of semiconductor devices and circuits. | 03:26 |
| fenn | internet's not giving up the info i want | 03:27 |
| nmz787__ | heh | 03:28 |
| fenn | unfortunately "inkjet chip" refers to the stupid DRM-like ink volume counter they use to force you to buy brand name ink cartridges | 03:29 |
| nmz787__ | arduino inkshield?? | 03:29 |
| nmz787__ | oh biocurios has that bioprinter group | 03:30 |
| fenn | well anyway the point is there's some way to couple macro scale tubing to 10 micron holes etched in silicon | 03:31 |
| -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap | 03:31 | |
| fenn | so then you just need a way to go from a flat silicon surface with holes in it, to your nanofluidic chip | 03:31 |
| fenn | which seems to be mostly a problem of alignment | 03:31 |
| fenn | they're going to drag me out of bed and force feed me in 6 hours :\ | 03:34 |
| fenn | and ask all sorts of probing questions about what i'm doing with my life | 03:34 |
| nmz787__ | are you home with parents? | 03:35 |
| fenn | yeah | 03:36 |
| fenn | " Polymer processing with UV excimer lasers occurs without any heat affected zones and yields particularly clean ablation results. " in polyimide (kapton) | 03:37 |
| fenn | repeatable 300 nm +- 5nm holes | 03:37 |
| fenn | in 50 micron kapton film | 03:38 |
| fenn | er, 500 nm holes | 03:38 |
| fenn | +- 100 nm my bad | 03:38 |
| nmz787__ | yeah but i no have UV laser | 03:39 |
| nmz787__ | err | 03:39 |
| nmz787__ | wait | 03:39 |
| nmz787__ | actually i think this company does have one but its in NM | 03:40 |
| fenn | the wavelength is 308nm so you can probably expect lambda * 1.5 hole size | 03:40 |
| fenn | i'm sure there's a formula for that | 03:40 |
| fenn | anyway you can FIB the hole | 03:41 |
| fenn | and then do PCB etching for the fluid manifold | 03:41 |
| nmz787__ | well its past bed time | 03:43 |
| nmz787__ | g2g | 03:43 |
| fenn | looks like "high precision" pcb etching is 25 microns or so | 03:43 |
| fenn | night | 03:44 |
| ThomasEgi | 25 microns.. bout one mil. that's pretty small for pcb's already | 03:44 |
| fenn | that was what lpkf claimed | 03:47 |
| ThomasEgi | main problem with etching pcb's is that your copper usualy is prety thick. so etching a thin trace would etch out the volume below the trace,too. | 03:47 |
| ThomasEgi | if you only etch a few microns in depth, going smaller should be no real problem | 03:48 |
| fenn | the channel doesnt need to be full copper thickness | 03:49 |
| fenn | even laser printers are getting up there in resolution these days.. 2400 dpi = 10 micron is common | 03:50 |
| fenn | pcb photo plotters have 8600 dpi resolution | 03:51 |
| fenn | http://mitspcb.com/edoc/bungard/photoplotter.htm | 03:51 |
| ThomasEgi | you can use the usual minify process to project a big plot onto a smaller surface | 03:52 |
| fenn | i think you can even get things 3D printed this small with SLA | 03:52 |
| fenn | these syringe pumps are so much overkill | 03:59 |
| fenn | the laser bubble pump is a much better idea | 04:00 |
| fenn | i get worried when a company's catalog says 'tubing diameter 1/16" um OD' | 04:01 |
| ThomasEgi | wha?.. 1/16 inch micrometer? | 04:10 |
| -!- strages_home [~strages@adsl-98-67-175-176.shv.bellsouth.net] has quit [Quit: Lost terminal] | 05:40 | |
| archels | http://www.ncbi.nlm.nih.gov/pubmed/22237065 | 06:39 |
| archels | The inverse calorie intake-LTL association is consistent with trial data showing beneficial effects of calorie restriction on aging biomarkers. Further exploration of energy intake and LTL dynamics in the young is needed. | 06:39 |
| archels | ltl = leukocyte telomere length | 06:39 |
| -!- Simurg [~Shehrazad@unaffiliated/shehrazad] has quit [Ping timeout: 244 seconds] | 07:18 | |
| -!- Simurg [~Shehrazad@95.5.88.151] has joined ##hplusroadmap | 07:26 | |
| -!- Simurg [~Shehrazad@95.5.88.151] has quit [Changing host] | 07:26 | |
| -!- Simurg [~Shehrazad@unaffiliated/shehrazad] has joined ##hplusroadmap | 07:26 | |
| -!- delinquentme [~asdfasdf@c-71-236-101-39.hsd1.pa.comcast.net] has joined ##hplusroadmap | 07:31 | |
| delinquentme | so is there a general consensus in here yet on what technology we're banking on for life extension :D ?? | 07:57 |
| Mariu | ! | 07:58 |
| Mariu | no idea, I'm curious about that myself | 07:59 |
| archels | try to hang in there using a healthy lifestyle and minor nutritional supplementation until we invent something that *actually works* | 07:59 |
| strangewarp | replicators, connected to nuclear reactors, stored in the chest cavity | 08:59 |
| Mariu | :o | 09:01 |
| -!- Coornail [~Coornail@li66-97.members.linode.com] has quit [Quit: ZNC - http://znc.sourceforge.net] | 09:03 | |
| -!- Coornail [~Coornail@li66-97.members.linode.com] has joined ##hplusroadmap | 09:03 | |
| -!- audy [~audy@heyaudy.com] has quit [Read error: Operation timed out] | 09:06 | |
| -!- Coornail [~Coornail@li66-97.members.linode.com] has quit [Read error: Operation timed out] | 09:07 | |
| -!- Coornail [~Coornail@li66-97.members.linode.com] has joined ##hplusroadmap | 09:08 | |
| -!- Coornail is now known as Guest95905 | 09:09 | |
| -!- audy [~audy@heyaudy.com] has joined ##hplusroadmap | 09:11 | |
| -!- chevbird [~chevbird@pool-108-34-246-28.prvdri.fios.verizon.net] has quit [Ping timeout: 244 seconds] | 09:23 | |
| -!- chevbird [~chevbird@pool-108-34-246-28.prvdri.fios.verizon.net] has joined ##hplusroadmap | 09:28 | |
| -!- strages_home [~strages@adsl-98-67-175-176.shv.bellsouth.net] has joined ##hplusroadmap | 09:31 | |
| kanzure | delinquentme: yes http://diyhpl.us/~bryan/papers2/longevity/ | 09:37 |
| Mariu | thanks, kanzure | 09:37 |
| -!- sylph_mako [~mako@103-9-42-1.flip.co.nz] has joined ##hplusroadmap | 09:48 | |
| -!- chevbird [~chevbird@pool-108-34-246-28.prvdri.fios.verizon.net] has quit [Quit: Leaving] | 09:50 | |
| -!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has joined ##hplusroadmap | 09:55 | |
| -!- EnLilaSko [~Nattzor@unaffiliated/enlilasko] has joined ##hplusroadmap | 10:23 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has quit [Ping timeout: 255 seconds] | 10:24 | |
| archels | http://turingbirds.com/temp/Science-2012-Langecker-932-6.pdf | 10:25 |
| archels | holy mother of awesome | 10:25 |
| archels | "Synthetic Lipid Membrane Channels Formed by Designed DNA Nanostructures" | 10:25 |
| Mariu | :p | 10:26 |
| fenn | seems like a bad link, not sure what's wrong | 10:27 |
| fenn | oh it's just slow | 10:28 |
| archels | it's 7MB | 10:28 |
| fenn | eta 10 min | 10:28 |
| archels | well find it on your own if you don't want to wait :P | 10:29 |
| yashgaroth | no no I'm invested in this download now | 10:29 |
| archels | haha | 10:29 |
| fenn | what he said | 10:30 |
| yashgaroth | oh they just stuck cholesterol on it? well that's cheating | 10:33 |
| yashgaroth | "and applied voltage pulses to facilitate incorporation into the membrane" damn straight *cough* | 10:35 |
| -!- jordyd [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has joined ##hplusroadmap | 10:36 | |
| -!- jordyd [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has quit [Changing host] | 10:36 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has joined ##hplusroadmap | 10:36 | |
| archels | http://turingbirds.com/temp/Langecker.SM.pdf | 10:44 |
| archels | supplementary material | 10:44 |
| archels | n fact, we seem to observe spontaneous insertion into the lipid bilayer (as discussed | 10:46 |
| archels | in SOM text S2) only for POPC lipids, whereas for DPhPC bilayers the additional | 10:46 |
| archels | application of voltage pulses is required (see below). | 10:46 |
| yashgaroth | oh no doubt, I'm just a fan of electricity re moving DNA across membranes | 10:48 |
| archels | anyway, the voltage and current for the necessary pulse are comparable to biological values, so probably not lethal to the cell | 10:48 |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has quit [Ping timeout: 244 seconds] | 11:22 | |
| -!- jordyd [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has joined ##hplusroadmap | 11:36 | |
| -!- jordyd [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has quit [Changing host] | 11:36 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has joined ##hplusroadmap | 11:36 | |
| -!- EnLilaSko [~Nattzor@unaffiliated/enlilasko] has quit [Ping timeout: 244 seconds] | 12:14 | |
| -!- EnLilaSko [~Nattzor@unaffiliated/enlilasko] has joined ##hplusroadmap | 12:15 | |
| -!- augur [~augur@208.58.5.87] has quit [Remote host closed the connection] | 12:22 | |
| -!- augur [~augur@208.58.5.87] has joined ##hplusroadmap | 12:23 | |
| -!- augur [~augur@208.58.5.87] has quit [Ping timeout: 244 seconds] | 12:27 | |
| -!- Skei [~Quentin@87-194-112-220.bethere.co.uk] has joined ##hplusroadmap | 12:39 | |
| -!- Skei [~Quentin@87-194-112-220.bethere.co.uk] has quit [Quit: Leaving] | 12:51 | |
| -!- juri_ [juri@funkykitty.net] has joined ##hplusroadmap | 13:51 | |
| -!- delinquentme_ [~asdfasdf@c-71-236-101-39.hsd1.pa.comcast.net] has joined ##hplusroadmap | 14:00 | |
| -!- delinquentme [~asdfasdf@c-71-236-101-39.hsd1.pa.comcast.net] has quit [Ping timeout: 252 seconds] | 14:03 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has quit [Ping timeout: 245 seconds] | 14:04 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has joined ##hplusroadmap | 14:18 | |
| -!- user123abc [~sally@c-67-171-79-251.hsd1.pa.comcast.net] has joined ##hplusroadmap | 14:23 | |
| -!- augur [~augur@65.199.61.222] has joined ##hplusroadmap | 14:27 | |
| -!- EnLilaSko [~Nattzor@unaffiliated/enlilasko] has quit [Quit: - nbs-irc 2.39 - www.nbs-irc.net -] | 14:28 | |
| -!- Tabrenus [~Tabrenus@213.211.132.86.static.edpnet.net] has joined ##hplusroadmap | 14:42 | |
| chido | electroporation: a natural choice | 14:58 |
| yashgaroth | :D | 14:59 |
| -!- Tabrenus [~Tabrenus@213.211.132.86.static.edpnet.net] has quit [Quit: Tabrenus] | 15:03 | |
| -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Leaving] | 15:04 | |
| -!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has quit [Quit: turkey] | 15:04 | |
| -!- user123abc [~sally@c-67-171-79-251.hsd1.pa.comcast.net] has quit [Quit: Leaving] | 15:22 | |
| -!- augur [~augur@65.199.61.222] has quit [Ping timeout: 244 seconds] | 15:23 | |
| nmz787__ | hey how do i getifconfig and iwconfig | 15:27 |
| nmz787__ | on a fresh debian install | 15:27 |
| sylph_mako | Apparently it's being obsoleted in favor of iw | 15:36 |
| -!- yolo [~doomboxsa@75.105.12.23] has quit [Read error: Connection reset by peer] | 16:28 | |
| -!- sivoais [~zaki@199.19.225.239] has quit [Ping timeout: 260 seconds] | 16:39 | |
| -!- sivoais [~zaki@199.19.225.239] has joined ##hplusroadmap | 16:42 | |
| -!- curt1s [~curtis@sol.whatbox.ca] has quit [Ping timeout: 248 seconds] | 16:49 | |
| -!- curt1s [~curtis@sol.whatbox.ca] has joined ##hplusroadmap | 16:50 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has quit [Remote host closed the connection] | 17:45 | |
| -!- jordyd [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has joined ##hplusroadmap | 17:49 | |
| -!- jordyd [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has quit [Changing host] | 17:49 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has joined ##hplusroadmap | 17:49 | |
| -!- jordyd_ [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has joined ##hplusroadmap | 17:49 | |
| -!- jordyd_ [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has quit [Remote host closed the connection] | 17:50 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has quit [Read error: Connection reset by peer] | 17:51 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has joined ##hplusroadmap | 17:51 | |
| -!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Remote host closed the connection] | 17:58 | |
| -!- delinquentme_ [~asdfasdf@c-71-236-101-39.hsd1.pa.comcast.net] has quit [Read error: Connection reset by peer] | 17:59 | |
| -!- ParahSailin [~parah@unaffiliated/parahsailin] has quit [Read error: Connection reset by peer] | 18:20 | |
| -!- ParahSailin [~parah@adsl-69-151-150-200.dsl.hstntx.swbell.net] has joined ##hplusroadmap | 18:20 | |
| -!- ParahSailin [~parah@adsl-69-151-150-200.dsl.hstntx.swbell.net] has quit [Changing host] | 18:20 | |
| -!- ParahSailin [~parah@unaffiliated/parahsailin] has joined ##hplusroadmap | 18:20 | |
| -!- delinquentme [~asdfasdf@c-71-236-101-39.hsd1.pa.comcast.net] has joined ##hplusroadmap | 18:55 | |
| -!- technicus [~technicus@50.124.104.66] has joined ##hplusroadmap | 19:57 | |
| juri_ | so, i'm interested in printing living things. anyone have a howto? ;) | 20:01 |
| -!- He||eshin [~talinck@cpe-174-101-208-182.cinci.res.rr.com] has quit [Ping timeout: 252 seconds] | 20:07 | |
| kanzure | juri_: http://diyhpl.us/wiki/diybio/faq | 20:14 |
| kanzure | nmz787__: sudo apt-get install iw | 20:14 |
| -!- technicus [~technicus@50.124.104.66] has left ##hplusroadmap [] | 20:14 | |
| -!- Helleshin [~talinck@cpe-174-101-208-182.cinci.res.rr.com] has joined ##hplusroadmap | 20:20 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has quit [Remote host closed the connection] | 20:25 | |
| -!- jordyd [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has joined ##hplusroadmap | 20:30 | |
| -!- jordyd [~jordyd@99-177-65-98.lightspeed.wepbfl.sbcglobal.net] has quit [Changing host] | 20:30 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has joined ##hplusroadmap | 20:30 | |
| -!- antifas [~safitan@75.105.12.23] has joined ##hplusroadmap | 20:44 | |
| juri_ | kanzure: thanks. | 21:00 |
| -!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has joined ##hplusroadmap | 22:10 | |
| -!- jordyd [~jordyd@unaffiliated/jordyd] has quit [Quit: Leaving] | 22:40 | |
| -!- delinquentme [~asdfasdf@c-71-236-101-39.hsd1.pa.comcast.net] has quit [Ping timeout: 244 seconds] | 23:16 | |
| -!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has quit [Ping timeout: 256 seconds] | 23:27 | |
| -!- augur [~augur@c-75-74-129-116.hsd1.fl.comcast.net] has joined ##hplusroadmap | 23:49 | |
| --- Log closed Fri Nov 23 00:00:25 2012 | ||
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!