I want to thank everybody again for the interest and good information regarding this admittedly semi-OT topic. :)
I hope I'm not stretching everybody's patience now, but given your interest, maybe I can elaborate a bit on where I'm coming from (if you're pressed for time and only interested in issues immediately related to OpenBSD, you can safely press Del now ;): 2009/4/6 J.C. Roberts <list-...@designtools.org>: >> > For Do-It-Yourself PCB's, you *really* want postscript support. Yep. But then, I want PostScript anyway. Badly. I've never owned a (working) PostScript printer in my life and I'm sick of raster image printers. Sick of 'em. Sick of 'em all. ;) >> > Color support is not necessary, and you can easily get away with >> > finding a free, used, office laser printer. As odd as it might >> > seem, some of the old laser printers are actually "better" in the >> > sense of they were built to last and you can still get parts for >> > most of them. > On Mon, 6 Apr 2009 15:05:14 -0400 Jason Dixon <ja...@dixongroup.net> > wrote: >> >> If the above is correct (and I believe JCR) then I can highly >> recommend the Brother HL-2170W. It's inexpensive and has worked >> great for me with OpenBSD. Comes with wireless *and* wired >> networking. >> >> http://www.brother-usa.com/Printer/ModelDetail.aspx?ProductID=hl2170W 2009/4/6 J.C. Roberts <list-...@designtools.org>: > If the real reason for buying a laser printer is PCB work... It isn't. And actually, I *do* want colour. Currently my only printer is a Canon BJC-85. That is an *EXTREMELY* slow colour inkjet, and while it prints ok text documents, its images are fairly atrocious. Also, it's *EXTREMELY* tedious to refill. Think hunching over the desk and using a syringe to slllllloooooooooowwwwwly drip-feed refill ink into its cartridge tanks. Or pay an arm and a leg for consumables that just aren't worth it. My existing printer's only advantage? It's portable. But I haven't even got a laptop anymore (other than an ancient PowerBook that has neither a USB nor a parallel port). So I could kind of use a new printer. And since my objective it so replace my current printer, I want colour, because otherwise I'd be tempted to keep the old one around. So having noticed that colour lasers have become relatively affordable, I want to kill the maximum number of birds with the one stone. Here's what I can "afford": I can put one printer on my birthday wish list. ;) I cannot afford buying a b/w laser AND a colour whatever, however cheap those two printers might be. Discovering the laser printer DIY PCB stuff just put another bird in my PETA-hating slingshot crosshairs. Some of you may remember that a long time ago I amateurishly cobbled together a serial shut down button for a headless OpenBSD box: http://marc.info/?l=openbsd-misc&m=111177380010679&w=2 (The sled software download no longer works, you can either find a prior version here: http://www.linuxfocus.org/~guido/#sled or maybe google and see if you can find the sled-0.4.tar.gz version somewhere. Being able to verify that the code doesn't wipe your hard drive would also help. No promises.) Anyway, while I got things to work, I never told you just how terrible a mess of leftovers and unshielded wires my worse-than-a-ratsnest Frankencircuit was. Want to see something truly frightening? Behold: http://imgur.com/J650Q.jpg http://imgur.com/J66KA.jpg So of course when I learned that I could bake my very own pretty PCBs, I damn near wet myself. Imagine: No more weirdly and reluctantly twisted wires held in place by solder. A neat properly printed circuit board. That I can lay out any way I darn well please. Heaven. But wait, there's more: Because I also recently discovered this: http://www.teuthis.com/daisy/ Portable MP3 sounds kinda tempting. Hey, I still use a cassette tape Walkman. (I bought 2 cheap MP3 players in the past but returned them both, because their Flash memory had errors.) They do sell these as kits here http://www.makershed.com/ProductDetails.asp?ProductCode=MKTET1 , but $115 USD? Nah. So I installed Cadsoft Eagle on my Ubuntu box, opened the board file, and selectively extracted only the actual top and bottom circuits and converted that and came up with this: http://imgur.com/J7JY2.png http://imgur.com/J7RNU.png (If you want to print this, set the ppi to 1200x1200, otherwise it won't be the correct size.) The idea is that I could (a) buy all the parts: - PIC microcontroller part number PIC18F45J10-I/P cf. http://www.microchip.com/wwwproducts/Devices.aspx?dDocName=en024619 and http://www.microchipdirect.com/productsearch.aspx?Keywords=PIC18F45j10 - http://www.vlsi.fi/en/products/vs1011.html http://verkkokauppa.planeetta.net/epages/Planeetta.sf/en_GB/?ObjectPath=/Shop s/vlsi/Products/N6N9480 - etc, etc.; (b) use a laminating machine to transfer the toner to both the top and bottom of double layer PCB stock; (c) etch the thing; (d) make me on of these, so I can SMT-solder: http://www.engadget.com/2006/03/07/how-to-make-a-surface-mount-soldering-iron / (e) dremel and wire and/or solder through the vias (http://en.wikipedia.org/wiki/Via_(electronics) ) to connect them; (f) and SMT-solder the shit out of these motherfucking chips on this motherfucking PCB. My biggest concern is that some of the paths are so thin, I'd be concerned whether I get them reliably masked with enough toner to have a complete connection -- and whether I can avoid having the acid eat up the thin circuit pathways from the side during the etching, breaking the circuit. But hey, if it doesn't work the first time, I should be able to edit the mask images and make these anorexic pathways a little healthier. Sure it's a challenge. If I pull it off it would be a quantum leap up from the horrors shown above. But hey, if I fail miserably, it's nobody's loss but my own, right? And actually, in an ideal world, I'd much rather create an audio player based on the VS1053 ( http://www.vlsi.fi/en/products/vs1053.html ) than the VS1011, because MP3 is patent-encumbered and OGG Vorbis isn't. But to do that I'd need to re-learn electronics, and probably learn embedded programming. Translation: Way, *way* WAY outta my league. In my wet dreams, perhaps. One day. Years from now. Maybe. 2009/4/6 J.C. Roberts <list-...@designtools.org>: > > You will probably never be building extremely high-speed PCB designs in > your garage, so a resolution of 300dpi or better should suffice. Even > the axis-based mixed-resolution printers (i.e. "600x300" Horizonal/Vert) > should work fine. If I really want to print out stuff like the above and have a chance of different circuits not running into each other, then I will need the highest dpi resolution I can get. I exported the mask PNGs so they'd be 1200x1200ppi. That should tell you something. > The things I like about my XEROX DocuPrint N17 are: > 1.) 1200 dpi resolution That would be super. > 2.) a paper path for thick card stock > 3.) network interface > 4.) postscript support (multiple levels) > 5.) duplexing > 6.) very cheap to run > You need to realize laser printers are *not* the only way to do PCB's, > and some of the inkjet printers are at least equal if not better for > this task. > > http://blog.makezine.com/archive/2006/08/how_to_direct_to_pcb_ink_jet_r.html > > (MakeZine has a number of other Home-PCB articles) > Some people argue that using inkjets is more accurate than laser > printers since the possible human errors in resist mask alignment are > eliminated. Also, with inkjet printers, you can even do "silkscreens" > of sorts on your home built boards. > > "Fashion Is My Only Conscience" ;-) You're absolutely right, and thank you very much for the link. However, having looked at the way they're essentially sacrificing a printer to the cause, and generally seeing the merits and limitations of both methods, I think that for now, the laser printer toner transfer method is the one for me. I did however see that Xerox has a series of solid ink printers, and I'm going to ask them what they think about the prospects of masking PCB stock with this. Should be fun. I could tell you if I get an answer if you're interested. Printer-wise, here's what I've looked at a bit closer: Xerox: Phaser 6110 USB version 6110V/B: ~199 / RJ45 version 6110V/N: ~259 IE ~129 ~139 DE ~# 85 ~# 95 UK Lexmark Lexmark C540n ~259 DE (Brother) HL-2400C ~$125 US (NB: WTF, Xerox?! Irish people get to pay 120 more for the exact same friggin printer?! We all know about "Ripoff Ireland" ( http://en.wikipedia.org/wiki/Rip-Off_Republic ), but this is _fucking ridiculous_. What's that? Taxes? Oh, like a 120 euro extra in taxes shit sandwich?!? Seriously? I don't believe you, Xerox.) > Lastly, if you're not in a rush, or if you're working on high-speed > designs, you should talk to your local PCB fabrication house. The silly > part is many of their customers do not use the entire "blank" so your > small design can easily be tossed in an unused corner of the blank and > be manufactured in parallel with orders from other customers. This is > really cheap to do, since without your order, the wasted space on the > blank would be tossed out. The cool part about this is you can get > multi-layer (>2) PCB designs --something you can't do at home-- done > very cheaply if you're patient, as well as get the benefit of DRC > testing, X-RAY analysis, Bed-Of-Nails (aka "clamshell") testing, ... > > As long as you can "Tape-Out" your design into a supported format (i.e. > gerber), your local fab house will most likely be *real* friendly since > it will save them wasted materials, and of course, they just never know > where the "little business" you offer may lead in the future. That's a real nice hint. I'll keep it in mind. Of course the free demo of Eagle can design only really small circuits, or open/print existing larger files. But Wikipedia tells me that both gEDA and KiCad also grok gerber files, so this might just work: http://en.wikipedia.org/wiki/Gerber_File Again, thank you everybody for your interest, and I hope this has not been too annoying a distraction from our regular scheduled programming. Thanks and regards, --ropers
