soul-d | when you figure out other uses for tools you might think twice when to mention :) -> http://www.youtube.com/watch?v=5-8t6N4rk4M&feature=player_embedded | 09:19 |
---|---|---|
West0n | How do you design the origonal exposure patern? | 20:27 |
horizontally | don't use transparencies not meant for laser printers in laser printers | 20:27 |
West0n | What program do you design the artwork in? | 20:27 |
West0n | And how do you make sure its to scale? | 20:27 |
Thetawaves | they have a 5 pixel design rule or something | 20:29 |
West0n | What program do you design the artwork in? | 20:32 |
West0n | And make sure the dimensions are correct? | 20:32 |
West0n | Paint doesent realy have a dimension tool | 20:33 |
West0n | I am trying to use solidworks | 20:37 |
West0n | But I can't seem to print to scale | 20:37 |
Thetawaves | wait for azonenberg to nonafk | 20:41 |
azonenberg | Thetawaves: I do the original design in Glade (the IC layout tool, not the GUI designer) | 21:17 |
azonenberg | Then i convert the GDS to a DXF | 21:17 |
azonenberg | and print that to scale | 21:17 |
West0n | Is glade free azonenberg? | 21:19 |
West0n | I'm guessing its not | 21:19 |
azonenberg | West0n: Freeware but closed source | 21:19 |
West0n | What can I use for simple designs? | 21:19 |
West0n | I only need a 8mm ring | 21:19 |
West0n | That's 1mm wide | 21:19 |
azonenberg | West0n: This is what I use, thats all i can say | 21:20 |
azonenberg | http://i.imgur.com/4eYkX.png | 21:20 |
azonenberg | That's a design I did in it | 21:20 |
azonenberg | a full 2" wafer of the comb drive I'll be making soon | 21:20 |
azonenberg | You are going to want to use vector graphics | 21:22 |
azonenberg | Autocad might not be a bad choice if you have it already and know how to use it | 21:22 |
West0n | What program do you print dxf's from? | 21:22 |
West0n | And does glade itself convert gds to dxf? | 21:23 |
Thetawaves | how much force will you get with that? | 21:23 |
azonenberg | Right now? (still working on this part) | 21:23 |
azonenberg | I convert DXF to PNG using QCad | 21:23 |
azonenberg | and postprocess in gimp | 21:23 |
azonenberg | then do final printing in gimp | 21:23 |
West0n | I know how to use solidworks | 21:23 |
azonenberg | The toolchain does need improving | 21:23 |
West0n | And drew a 3d drawing | 21:23 |
West0n | Can't figure out how to export it as something usefull.... lol | 21:23 |
azonenberg | eventually i want to feed the GDS directly into a laser direct write system for mask fab | 21:23 |
azonenberg | Thetawaves: No idea | 21:23 |
azonenberg | The goal is, first off, to build it successfully | 21:24 |
azonenberg | Doesnt even matter if it works | 21:24 |
azonenberg | being able to build something off the mask set is a first step | 21:24 |
West0n | Solidworks won't let me export in 1:1 scale for dxf... | 21:26 |
azonenberg | Hmm, strange | 21:26 |
Thetawaves | i think you should just use inkscape | 21:28 |
Thetawaves | or another good vector graphics program | 21:28 |
Thetawaves | hand draw it like the 6502 | 21:28 |
azonenberg | That will work fine for simple masks | 21:29 |
azonenberg | the nice thing about glade is that it can export to standard formats like GDS | 21:29 |
azonenberg | and it supports cell based layout | 21:29 |
West0n | How do I print vector graphics with correct scale? | 21:29 |
azonenberg | meaning you can do hierarchal designs rather than flat | 21:29 |
azonenberg | Depends totally on your toolchain | 21:29 |
azonenberg | If you figure out a good way of doing it, let us know and we'll document it on the wiki | 21:29 |
azonenberg | the only way i know how to do to-scale printing is Gimp specifying physical size of the rendered image in the print set up page | 21:30 |
West0n | I'm gonna try google sketchup | 21:35 |
West0n | I think you can print to scale from it | 21:35 |
azonenberg | Do your own research for your tools | 21:37 |
azonenberg | I'd love to hear how it turns out but I don't know every tool around so i cant help you with everything | 21:37 |
Thetawaves | it would be better to laser->galvanometer->lense->wafer | 21:39 |
Thetawaves | convert dxf to gcode | 21:40 |
Thetawaves | run the galvanometers with the gcode | 21:41 |
azonenberg | Thetawaves: Thats the idea | 21:41 |
azonenberg | except it would probably use GDSII as the native format | 21:41 |
azonenberg | or GDSII to gcode | 21:41 |
azonenberg | But that's the ultimate goal | 21:41 |
azonenberg | I'm talking with a couple of folks now about it | 21:41 |
Thetawaves | GDSII to gcode | 21:41 |
Thetawaves | ? | 21:41 |
Thetawaves | sounds like a bitch | 21:42 |
azonenberg | Thetawaves: I might do GDSII to gerber and then make gcode from that, or run the gerber natively | 21:43 |
azonenberg | i dont know enough about the file formats to know how CAM friendly they are | 21:43 |
azonenberg | But what i do know is i am going to have a pair of galvos on a mirror assembly with a bluray laser bouncing off it (405nm, right in the middle of my photoresist's peak sensitivity range) | 21:43 |
azonenberg | shining onto a metalized photoresist-coated microscope slide | 21:43 |
azonenberg | the goal being a chrome-on-glass mask | 21:44 |
Thetawaves | are you going to make the galvanometer? | 21:44 |
azonenberg | it wont have alignment capability so i cant do direct write on the wafer | 21:44 |
azonenberg | And not sure, this is a longer term project | 21:44 |
Thetawaves | yes after some consideration | 21:53 |
Thetawaves | i WOULD be fun to build a galvo | 21:53 |
Thetawaves | it* | 21:53 |
azonenberg | how about one that does 2D on one mirror? | 21:54 |
azonenberg | So you can do x/y raster scan | 21:54 |
Thetawaves | procuring magnets could be irritating | 21:54 |
azonenberg | i'm not sure how that would be done | 21:54 |
Thetawaves | i would not attempt it | 21:54 |
Thetawaves | procuring magnets is always irritating | 21:55 |
Thetawaves | azonenberg, have you figured out a way to do 2d on one mirror? | 21:57 |
azonenberg | Lol | 21:57 |
azonenberg | If i had, i'd have built the thing already :p | 21:58 |
azonenberg | I was thinking some kind of a ball joint at the center of the mirror | 21:58 |
azonenberg | sprung to keep it centered | 21:58 |
azonenberg | then electromagnets on all four corners | 21:58 |
Thetawaves | no | 21:58 |
azonenberg | Why? | 21:58 |
Thetawaves | the electromagnet is the mirror | 21:58 |
Thetawaves | you have magnets on the outside | 21:58 |
Thetawaves | north/south | 21:59 |
azonenberg | Why does it have to be done like that? | 21:59 |
azonenberg | remember i dont need super fast panning | 21:59 |
Thetawaves | i'm no expert, but if you look at the way regular galvanometers work... | 21:59 |
azonenberg | its not like i'm trying to do realtime video projection etc | 21:59 |
azonenberg | i'll be doing a brief ~100ns pulse | 21:59 |
Thetawaves | you don't want to control two separate coils when you can do one | 21:59 |
azonenberg | then panning by a small amount and doing it again | 21:59 |
azonenberg | And the thing is, if i do 2 axis | 22:00 |
azonenberg | i will need at least 2 nad possibly 4 | 22:00 |
Thetawaves | you're talking about 4 signals vs 2 | 22:00 |
azonenberg | in fact, i almost wonder if i could do it with digital servomotors | 22:00 |
azonenberg | similar to how they do swashplates on RC helicopters | 22:00 |
Thetawaves | you probably could | 22:00 |
azonenberg | i just would like to make it a little bit faster | 22:00 |
azonenberg | And also more precise | 22:00 |
azonenberg | because i would like to have micron-scale positioning accuracy lol | 22:01 |
Thetawaves | you could probably do better with some belts and a high def encoder | 22:01 |
azonenberg | Probably | 22:01 |
azonenberg | Again, these are just ideas and i havent done any real research on it | 22:02 |
azonenberg | Right now i have more immediate problems at hand | 22:02 |
azonenberg | Like the 100-odd exams i need to have graded ASAP | 22:02 |
Thetawaves | <- machine tool enthusiast | 22:02 |
azonenberg | I want to get better at machining too, i'm competent but no expert | 22:02 |
azonenberg | thing is, i dont hav ethe time right now :( | 22:03 |
Thetawaves | what a shame | 22:03 |
--- Tue Nov 22 2011 | 00:00 |
Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!