| 00:30.49 | *** join/#edev bigbootay (n=bigboota@CPE0004e28a9289-CM000a73a14ed4.cpe.net.cable.rogers.com) |
| 00:50.23 | *** join/#edev ccjoe (n=ccjoe@www.clearcore.com) |
| 01:04.18 | *** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) |
| 02:34.35 | *** join/#edev redblue (i=star@ppp044.109-253-207.mtl.mt.videotron.ca) |
| 02:40.03 | MonMotha | jeesh, 98 components on a 3.75x4.75" 2 layer board |
| 03:59.38 | sunrunner20 | lol |
| 04:49.54 | MonMotha | sunrunner20: 98 components on that size of a board is really not that dense |
| 04:50.00 | MonMotha | however, on a 2 layer, it's getting there |
| 04:52.18 | dr-dotirc | are you routing your own pcb? |
| 04:53.16 | MonMotha | yeah |
| 04:53.22 | MonMotha | I do so frequently, actually |
| 04:53.33 | MonMotha | this is about the densest I can pull off with 2 layers, though I've still got a corner |
| 06:34.55 | *** join/#edev davidc__ (n=davidc@c-67-190-1-112.hsd1.co.comcast.net) |
| 07:34.19 | T0mW | MonMotha: multilayer costs too much |
| 07:41.16 | *** join/#edev DrZimmerman (n=theo@217-162-225-59.dclient.hispeed.ch) |
| 07:51.00 | MonMotha | T0mW: why do you think this is a 2 layer design? :) |
| 07:51.20 | MonMotha | biggest challenge gets to be routing power/ground in a manner that isn't likely to cause problems |
| 07:51.39 | T0mW | MonMotha: because you are cheap like me, and I've got a few boards that I can still cram some parts onto. heh |
| 07:52.01 | MonMotha | T0mW: I'm filling up the corner I spoke of earlier |
| 07:52.06 | T0mW | heh |
| 07:52.12 | MonMotha | "client" (aka the friend I'm doing this for) wants bass detection |
| 07:52.19 | MonMotha | is there a preferred way of doing that, btw? |
| 07:52.48 | MonMotha | I'm currently adding the stereo channels after rectification, low passing, comparing, using that to charge a leaky cap, and tagging a comparator off that |
| 07:52.59 | MonMotha | works well enough in simulation |
| 07:53.08 | MonMotha | haven't gotten a chance to actually build it yet |
| 07:56.29 | T0mW | MonMotha: http://www.openhardware.net/board.gif |
| 07:56.38 | T0mW | MonMotha: this evenings work, again. |
| 07:56.51 | MonMotha | fun fun |
| 07:57.10 | MonMotha | I haven't gone to TSSOPs on this board because 1) I don't need the space that badly and 2) they're harder to route since you can't sneak a trace between the pins |
| 07:57.16 | T0mW | MonMotha: it is, I've 512K of RAM on that with a 60MHz ARM7TDMI |
| 07:57.18 | MonMotha | so I'm mostly just using SOs. The micro is a QFP of course |
| 07:57.39 | T0mW | MonMotha: you using the stencils + toaster oven yet? |
| 07:59.01 | T0mW | MonMotha: hey, you using that $33 special at Advanced Circuits? If so, are you panelizing any boards? |
| 08:01.28 | MonMotha | no stencils or toaster oven yet |
| 08:01.39 | MonMotha | and yes, I am using the $33 special, and yes I panelize out the wazoo |
| 08:01.46 | MonMotha | I get my 70 sq. in. worth :) |
| 08:01.48 | dr-dotirc | what is it good for, a switch? |
| 08:01.53 | T0mW | MonMotha: do they cut them for you? |
| 08:02.25 | MonMotha | T0mW: no |
| 08:03.41 | T0mW | MonMotha: yeah, I've been faced with cutting them out myself. I just did an order for the Zipit Addon board and I had them put 10 (2X5) on a sheet. It was "interesting" cutting them out with a dremel diamond saw disk. Lots of dust. |
| 08:04.12 | T0mW | I got some disposable surgical masks + latex gloves, but it was still messy. |
| 08:04.19 | MonMotha | T0mW: http://gigapet.monmotha.net/~monmotha/board.pdf |
| 08:04.24 | MonMotha | that's my current layout |
| 08:04.33 | MonMotha | that's just a printout from layout, so it's not exactly pretty |
| 08:04.47 | MonMotha | T0mW: in all honesty...use a hacksaw |
| 08:05.01 | MonMotha | the cut isn't quite as clean, but the health hazard is much lower |
| 08:05.12 | T0mW | MonMotha: no, couldn't it had to be a precise cut |
| 08:05.18 | MonMotha | ah, ok |
| 08:05.24 | MonMotha | you panelized it yourself, right? |
| 08:05.29 | MonMotha | or did they panelize it for you on a $33/ea? |
| 08:05.37 | dr-dotirc | did you use orcad or pcad? |
| 08:05.39 | MonMotha | the $33/ea is pretty minimal in terms of service |
| 08:05.42 | MonMotha | dr-dotirc: that's orcad |
| 08:06.18 | T0mW | MonMotha: Oh, you've got tons of room! |
| 08:06.22 | T0mW | heh |
| 08:06.24 | MonMotha | T0mW: in some places, yes |
| 08:06.39 | MonMotha | there is still room on the bottom |
| 08:06.54 | MonMotha | the issue as I said is routing everything on a 2 layer board |
| 08:07.04 | MonMotha | the routing is becoming the density limiter |
| 08:07.10 | T0mW | MonMotha: I don't bother with drill charts in a few years now. They never look at them, they use their own software to make a drill drawing (if ever). |
| 08:07.23 | MonMotha | T0mW: I don't send the drill chart. OrCAD just puts it on there by default |
| 08:07.31 | T0mW | MonMotha: goto 5mil trace + 5mil spacing. |
| 08:07.45 | MonMotha | when I send it, I send the gerbers for top/bot/smt/smb/sst/ssb/outline and the excellon drill file |
| 08:07.55 | MonMotha | $33/ea only gets you 8mil/8mil |
| 08:07.59 | MonMotha | so that tends to be my spec |
| 08:08.00 | T0mW | NO! |
| 08:08.08 | T0mW | MonMotha: 5/5 with $33 |
| 08:08.12 | MonMotha | they do? |
| 08:08.16 | T0mW | yessir |
| 08:08.22 | MonMotha | that's new |
| 08:08.35 | T0mW | been ordering them that way for the past couple of boards (3 so far) |
| 08:08.42 | MonMotha | front page says 6mils |
| 08:08.50 | MonMotha | Min. 0.006" line/space |
| 08:08.58 | MonMotha | which is still 2mils better than 8, I guess |
| 08:09.17 | MonMotha | perhaps my 8 was the "I really want this to work and not have manufacturing bugs" thing |
| 08:09.36 | T0mW | MonMotha: uuuh, I stand corrected, it is 6, I've 5mil on this board display, that is what confused me. But 6 mil is better than 8 |
| 08:09.36 | MonMotha | my DRC is set to 9 mils on this design currently, with the knowledge that I can bump it down if needed |
| 08:09.46 | MonMotha | yes, 6 is better than 8 |
| 08:09.56 | T0mW | you can cram a lot more traces in 6/6 |
| 08:10.04 | MonMotha | yes, youc an |
| 08:10.43 | MonMotha | this is about the densest board I've done so far. Every design I do I find new ways to make things denser without significantly altering my manufacturing specs |
| 08:11.00 | MonMotha | I try to save turning my DRC down to the absolute minimum for times when it's absolutely required |
| 08:11.02 | MonMotha | same with via size |
| 08:11.29 | MonMotha | those are actually 12 mil traces on that board |
| 08:11.31 | MonMotha | quite large |
| 08:13.24 | T0mW | MonMotha: this is a 6/6 I had done last month: http://www.openhardware.net/Quad485_3U.gif |
| 08:14.06 | MonMotha | even at 6 mils, I still don't think you can sneak a trace between TSSOP pads, can you? |
| 08:14.35 | T0mW | MonMotha: definately had to have a JTAG connector on that board, no more soldering freakin wires to a pin header to get a jtag connection. |
| 08:14.44 | T0mW | MonMotha: no |
| 08:15.13 | MonMotha | heh, I always bring out JTAG. It's not always to a dedicated header, but you can always access JTAG from a single connector on my boards |
| 08:15.45 | T0mW | MonMotha: if you study the ratsnest enough, you'll usually find a nice flow to the layout. It took me a few tries to get that Quad485 CPU + RAM wiring to flow |
| 08:16.32 | MonMotha | I've noticed that |
| 08:16.41 | MonMotha | and of course, placement is critical |
| 08:16.59 | MonMotha | as can be re pinning devices that allow that (such as using a different gate on some glue logic if you have any) |
| 08:17.18 | MonMotha | like I said, every new board I make ends up way denser than I previously thought I was capable of |
| 08:17.27 | MonMotha | I'm getting better at it quickly, but I don't get to do much while in school |
| 08:17.38 | T0mW | http://www.openhardware.net/Quad.gif |
| 08:17.55 | T0mW | That's a snapshot of the eagle layout of the Quad485 board |
| 08:19.33 | T0mW | MonMotha: one trick I've learned that makes routing soooo much easier is to put the GND on the top layer and the +3.3volt on the bottom layer. Attach the power to a copper pour. |
| 08:19.51 | MonMotha | yeah, that works often times |
| 08:19.56 | T0mW | MonMotha: then when you need a +3v3, drop a via |
| 08:20.14 | MonMotha | yeah, basically you're pretending thaty ou have a 4 layer board |
| 08:20.42 | MonMotha | for various reasons, this particular board is actually 5v (mostly) |
| 08:20.50 | MonMotha | there's also some 12, and some 3.5 (of all things) |
| 08:21.09 | T0mW | of course, then you have to go through the routine of highlighting the power signal to make sure you get a good flow of current around the board, no high current areas trapped by traces . |
| 08:21.24 | MonMotha | yeah |
| 08:21.32 | MonMotha | and I've also gotten in trouble with that method when it comes to EMI |
| 08:21.38 | MonMotha | you can end up with some big loops |
| 08:21.51 | T0mW | MonMotha: how do you test them for EMI? |
| 08:21.53 | MonMotha | even if you don't care about passing regulatory, the susceptibility can be killer |
| 08:22.25 | MonMotha | T0mW: http://www.rhventures.org/ has a screenroom and a DC-70GHz spectrum analyzer |
| 08:22.27 | MonMotha | comes in handy |
| 08:22.35 | T0mW | nice |
| 08:23.12 | MonMotha | failing that, you can take the largest loop, and do some calculations |
| 08:23.45 | MonMotha | treat it as a loop antenna at your clock frequency (and a few harmonics that happen to be near where the antenna is ideal) and see what your emissions are likely to be |
| 08:23.54 | MonMotha | that can also give you some ideas for susceptibility |
| 08:24.19 | MonMotha | did you see my question on the bass detection btw? |
| 08:24.21 | T0mW | yeah, but, that is a fat antenna, lots of wavelengths could fit anywhere |
| 08:24.24 | MonMotha | does my method sound reasonable? |
| 08:24.45 | T0mW | no, didn't read it |
| 08:24.53 | MonMotha | and yeah, the loop method isn't perfect, but it generally yeilds a ballpark figure that's better than knowing nothing at all |
| 08:24.58 | MonMotha | oh |
| 08:25.09 | MonMotha | I have need to do bass detection on an audio signal (line level, for what it's worth) |
| 08:25.42 | MonMotha | my current solution is to rectify both channels, add, low-pass, compare with a threshold, use that to charge a leaky cap and compare that |
| 08:25.53 | T0mW | MonMotha: I don't do well with analog, they have special people that love that crap^H^H^H^H stuff |
| 08:25.59 | MonMotha | heh |
| 08:26.07 | MonMotha | this is pretty simple, but nor do I :) |
| 08:26.11 | MonMotha | I'm a digital guy |
| 08:26.16 | T0mW | nod |
| 08:26.25 | *** join/#edev linac_ (n=lin@121.71.56.121) |
| 08:27.04 | T0mW | I leave the analog stuff to those that actually enjoy it. It is cool to listen to them work and tell you what they're doing with Minority Carriers and stuff like that. |
| 08:27.16 | MonMotha | heh |
| 08:28.22 | T0mW | I've still got to implement a watermark CTS flow control in the LPC2000 software yet... |
| 08:29.15 | T0mW | I'm trying to build a decent foundation for this board, then I can move onto some interesting stuff to do with it, like getting the SPI EN28J60 ethernet working. |
| 08:56.48 | T0mW | MonMotha: one nice thing about surface mount. You can put all the legends and jumper settings text on the bottom silk layer |
| 08:57.43 | MonMotha | T0mW: indeed you can |
| 08:57.56 | MonMotha | though I often design boards for production without bottom silk since it's an extra expense |
| 09:02.43 | T0mW | MonMotha: depends on what your cost margins are. |
| 09:04.53 | T0mW | I do have a board like that though, the manufacturing cost of the board has to be kept around $26 in 1000 lot. Too bad though, I wish I could put another processor on the board and slap an external RAM on there too. But it is too "expensive" ($8 more) |
| 09:06.04 | T0mW | ok, this one is done. Until I get the isolated supply design to cram on there too... |
| 09:12.16 | T0mW | MonMotha: I wish I had a good technical college around here to take some classes |
| 09:24.29 | *** join/#edev leoncamel (n=leoncame@219.238.107.107) |
| 09:31.03 | MonMotha | well, it depends on what you want to learn |
| 09:45.40 | MonMotha | T0mW: I must also admit that I often overengineer my PCB designs. The PCB is often the most expensive part of the project due to small quantities and the wonders of free samples, so getting it right is important |
| 09:45.58 | MonMotha | especially since getting it wrong too many times can mean actually having to BUY parts, which are rarely available in the quantity I need |
| 09:46.26 | T0mW | MonMotha: agreed, rule #1 of engineering: "once released, you can never change it" |
| 09:46.41 | MonMotha | on the bright side, that usually means that they work correctly the first time, or at the minimum require minor changes that can easily be done with change wires |
| 09:46.58 | T0mW | they will let you cut & jumper it, but you cannot "improve" it. |
| 09:48.14 | MonMotha | this is the first design I've done with a SMPS on the board |
| 09:48.18 | MonMotha | so I'm kinda antsy |
| 09:48.20 | T0mW | crap |
| 09:48.38 | MonMotha | usually I leave the power up to somebody else (i.e. the magical wall wart) |
| 09:48.39 | T0mW | I put the wrong crystal on this board: 12.5pf instead of 6pf |
| 09:48.46 | MonMotha | whoops |
| 09:48.55 | MonMotha | can you just use different caps? |
| 09:49.03 | T0mW | well, darned RTC was gaining time... |
| 09:49.14 | MonMotha | something tells me from your reaction that the frequency difference is not acceptable in this case |
| 09:49.16 | MonMotha | evidently not... |
| 09:49.57 | T0mW | I'll allow up to a minute a month, worst case. I'd like it to be only a few seconds a month though. This is running about 4 minutes per month. |
| 09:50.30 | MonMotha | bah, just hook a GPS up to it and sync off that :) |
| 09:51.41 | T0mW | not a big deal as when the board is connected to the system, it gets a time update periodically. But, if the board is ever configured to run without 2way network communication (in a NEMA box someplace with an RF transmitter), I'd like to know that the RTC is pretty darned accurate. |
| 09:54.16 | MonMotha | what's this board do? |
| 09:54.44 | T0mW | oh well, I wasn't sure what crystal I put in it when I assembled it, so it has a 12.5pf now. I'll let it run a week or two and then see. |
| 09:55.19 | T0mW | MonMotha: Security + Life Safety: monitors Burgular detectors and Fire Alarm Control Panels. |
| 09:55.38 | MonMotha | ah, ok |
| 09:55.58 | MonMotha | a little bit more real world than the design I'm doing |
| 09:56.29 | T0mW | heh, some of the customers get real picky: like DoD, DoE, etc. |
| 09:56.35 | MonMotha | well yeah |
| 09:56.50 | T0mW | pencil pushers have nothing better to do than look for problems |
| 09:57.55 | T0mW | MonMotha: the thing that bothers me about using the ARM7TDMI is the data abort possibility... I have yet to test out the watchdog on this MPU |
| 10:01.01 | MonMotha | see, this is why I don't handle the software for systems of that nature |
| 10:01.56 | T0mW | MonMotha: heh, some day, some where, some one will come up to you and say "will you do it for the money?" |
| 10:02.18 | MonMotha | true |
| 10:02.52 | T0mW | a that point in your career, you'll think yourself so darned good at writing code... |
| 10:02.54 | T0mW | lol |
| 10:11.17 | T0mW | MonMotha: I didn't know what I was getting into 20 years ago, I needed a job at the time and this thing came along. |
| 10:51.52 | *** join/#edev ico2 (n=ico2@modem-504.cheetah.dialup.pol.co.uk) |
| 11:01.19 | *** join/#edev ico2 (n=ico2@modem-504.cheetah.dialup.pol.co.uk) |
| 14:01.37 | *** join/#edev LyosNorezel (n=LyosNore@cpe-75-179-138-54.woh.res.rr.com) |
| 14:10.21 | *** join/#edev ico2 (n=ico2@modem-504.cheetah.dialup.pol.co.uk) |
| 14:12.00 | *** join/#edev redblue (i=star@ppp199.108-253-207.mtl.mt.videotron.ca) |
| 14:22.28 | *** join/#edev leoncamel (n=leoncame@219.142.129.211) |
| 14:48.18 | *** join/#edev GPSFan (n=kenm@12.10.255.248) |
| 14:48.18 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:56.50 | *** join/#edev ico2 (n=ico2@modem-504.cheetah.dialup.pol.co.uk) |
| 16:05.17 | *** join/#edev ico2_ (n=ico2@modem-1712.chimpanzee.dialup.pol.co.uk) |
| 16:05.40 | *** join/#edev mbuf (n=mbuf@59.92.49.44) |
| 17:53.30 | sunrunner20 | anybody interested in a superh 32-bit RISC eval kit? |
| 18:05.49 | *** join/#edev ico2__ (n=ico2@modem-486.cheetah.dialup.pol.co.uk) |
| 18:18.36 | Crofton | probably, but they might be afk :) |
| 18:29.42 | *** join/#edev ccjoe (n=ccjoe@www.clearcore.com) |
| 18:41.44 | sunrunner20 | yea |
| 18:44.34 | chouimat | sound interesting |
| 19:03.03 | *** join/#edev ico2_ (n=ico2@modem-3968.cougar.dialup.pol.co.uk) |
| 19:14.06 | jbevren | sunrunner20: superH kit? |
| 19:14.31 | sunrunner20 | yea |
| 19:16.08 | sunrunner20 | lol |
| 19:16.13 | sunrunner20 | intrested possibley? |
| 19:16.17 | jbevren | possibly |
| 19:17.43 | jbevren | is this something still being developed? |
| 19:18.10 | sunrunner20 | no |
| 19:18.15 | sunrunner20 | its an offical offering |
| 19:18.17 | jbevren | I'd actually love to get uboot onto my dreamcast with the dead laser :X |
| 19:18.22 | jbevren | some day.. lol |
| 19:18.29 | jbevren | so, whatcha got |
| 19:18.50 | sunrunner20 | i'll have to look latter lol |
| 19:18.55 | jbevren | ah ok |
| 19:19.01 | sunrunner20 | its ouside |
| 19:19.03 | sunrunner20 | well |
| 19:19.07 | sunrunner20 | its accross the lawn |
| 19:19.19 | sunrunner20 | whcih just happens tobe soaking weet |
| 19:19.20 | sunrunner20 | and muddy |
| 19:19.28 | jbevren | hehe |
| 19:19.36 | jbevren | any idea on a price? |
| 19:19.54 | jbevren | my budget is tignt atm, but I'll pay for a reasonable board ;) |
| 19:20.21 | sunrunner20 | shipping |
| 19:20.21 | sunrunner20 | plus a few bucks lol |
| 19:20.21 | jbevren | I'm in! |
| 19:20.21 | jbevren | hehe |
| 19:20.28 | jbevren | specially if it has some docs |
| 19:20.47 | jbevren | <PROTECTED> |
| 19:31.32 | *** join/#edev T0mW (n=Tom@24.238.68.20.res-cmts.sth.ptd.net) |
| 20:10.21 | sunrunner20 | price is pretty much open bid :p |
| 20:10.34 | sunrunner20 | i need money for my laptop :) |
| 20:21.16 | jbevren | so why dontcha ebay it :) |
| 20:32.07 | *** join/#edev undef (i=undef@unaffiliated/undef) |
| 20:32.20 | undef | hey philv |
| 20:51.38 | *** join/#edev ccjoe (n=ccjoe@www.clearcore.com) |
| 21:03.30 | *** join/#edev ico2__ (n=ico2@modem-3568.chameleon.dialup.pol.co.uk) |
| 21:04.30 | sunrunner20 | jbevren: because nobody on ebay will buy :o |
| 21:04.33 | sunrunner20 | its an older model |
| 21:04.52 | sunrunner20 | and I wan't to to go to somebody in know well enough that it won't go to waste |
| 22:23.02 | *** part/#edev undef (i=undef@unaffiliated/undef) |
| 22:25.21 | jbevren | ahh |
| 22:25.27 | jbevren | what cpu is on it? |
| 22:43.09 | *** join/#edev redblue (i=star@ppp057.110-253-207.mtl.mt.videotron.ca) |
| 23:04.11 | *** join/#edev ico2_ (n=ico2@modem-1710.cougar.dialup.pol.co.uk) |
| 23:09.52 | *** join/#edev JoseJX (n=jjezak@c-71-61-58-178.hsd1.pa.comcast.net) |
