>Date: Sat, 12 Aug 2006 16:08:16 +1200
>From: "Ethan Dicks" <ethan.dicks at gmail.com>
>Subject: SMT hobby projects (was Re: group buy for homebrew CPUs?)
>On 8/12/06, der Mouse <mouse at rodents.montreal.qc.ca> wrote:
>> > That said, the notion that surface mount is somehow "harder to
>> > solder" is nothing but a load of crap. It is DIFFERENT...not harder.
>>
>> I've never tried surface-mount chips. But I've tried soldering things
>> of similar size, and for me, yes, it *is* harder.
>
>While parts can slip or be pulled by solder wicking things around, a
>few simple tricks (soldering corners of QFPs first, using scotch tape
>to hold down a 1206 SMT resistor while soldering one end...) make
>things a bit easier.
>I put together an IOB6120 (QFP FPGA, 0.5mm-pitch TSOP FLASH...) with a
>Weller iron and no special tools... just solder wick for clean-up, and
>a flux pen to help everything flow nice and cleanly. I've also done a
>few CF connectors for a couple of Spare Time Gizmos projects (Elf2000
>disk board, MP3 player...) To be fair, I wasn't doing it with over-40
>eyes, but I expect that someday, I'll have to start using my bench
>lens/light for everyday stuff.
I agree with Ethan. Surface mount is different but not necessarily
harder. I find desoldering through-hole to be many times harder than
desoldering surface mount. I *hate* desoldering through-hole
components, if I need to do it non-destructively. If I can destroy
the component in the process it's not as bad, just tedious--very very
tedious.
I've desoldered and resoldered up to 208 pin QFPs so far. I've only
done a couple of those. Mostly I did 160 pin QFPs. But I have not
had a failure yet. That is, the chips always remained functional.
There were one or two times where I had to touch up the soldering job
to get the board fucntional--I sometimes remove too much solder
during clean up.
The trick to surface mount soldering is being aware of how surface
tension can work for you and using it. It's a different set of
habits and way of thinking from through-hole.
I did my 208 pin QFP desolder/resolder with some Chip Quik alloy, a
Milwaukee brand heat gun, modeling clay (to keep surrounding
components from blowing away) a 40 watt and a 15 watt soldering
pencil from Radio Shack and a little bottle of soldering flux with a
brush.
Oh, I guess there was a roll of solder and some spray can flux
cleaner, as well as swabs and alcohol involved. But nothing
expensive or exotic.
Up to about 40 or 50 pins, I don't need the heat gun. The soldering
pencils are sufficient.
For surface mount resistors, I usually tin one pad. Then heat that
pad with a pencil in one hand and place the resistor with tweezers in
the other hand. Hold in place with the tweezers and remove the
pencil so that the solder cools to hold the resistor in place.
For magnification I use a 10X magnifying cup and hold the big end in
my eye socket by squishing it into the skin around the eye, sort of
like a monacle. The last time I checked these were available in the
jewelry section of Ebay for under $5.
Jeff Walther
>Date: Fri, 11 Aug 2006 15:41:31 -0400
>From: "Evan Koblentz" <evan at snarc.net>
> Ah ... that's a relief, LOL I thought you were saying you'd never heard of
>WD!
>
Well, there's a company called Western Design. They design ICs as
far as I can tell. I don't know if there's any relationship to
Western Digital, but they seem to go by WD also.
Jeff Walther
LCD panels have two connections: A data cable (20 to 60 pins), and the
backlight.
The data cables are totally non-standard and unique, unless you got a panel
that takes VGA or video inputs (this is extremely rare and not the norm).
Almost every single panel is different from almost every other panel, and
the connectors are impossible to get in small quantities.
The backlight is just the two ends of a cold-cathode Fluorescent lamp tube 6
to 16 inches long, depending on the size of the LCD panel. It goes to a
high voltage inverter mounted in the lid of the laptop with the LCD panel
which produces 400 to 2,000 volts to start and run the lamp. [some larger
LCD panels have multiple lamps].
The chances of being able to successfully use an LCD panel from a laptop for
any other purpose are near zero.
On 8/12/06, der Mouse <mouse at rodents.montreal.qc.ca> wrote:
> > That said, the notion that surface mount is somehow "harder to
> > solder" is nothing but a load of crap. It is DIFFERENT...not harder.
>
> I've never tried surface-mount chips. But I've tried soldering things
> of similar size, and for me, yes, it *is* harder.
While parts can slip or be pulled by solder wicking things around, a
few simple tricks (soldering corners of QFPs first, using scotch tape
to hold down a 1206 SMT resistor while soldering one end...) make
things a bit easier.
> Admittedly, that could be a lack of tools. Given a stereo loupe,
> pantographic waldos, and a really fine-point soldering iron, it quite
> possibly wouldn't be all that tough. But those ain't cheap, and I
> think needing significantly more expensive equipment counts as
> "harder".
I put together an IOB6120 (QFP FPGA, 0.5mm-pitch TSOP FLASH...) with a
Weller iron and no special tools... just solder wick for clean-up, and
a flux pen to help everything flow nice and cleanly. I've also done a
few CF connectors for a couple of Spare Time Gizmos projects (Elf2000
disk board, MP3 player...) To be fair, I wasn't doing it with over-40
eyes, but I expect that someday, I'll have to start using my bench
lens/light for everyday stuff.
Just this week, I did a hybrid through-hole/SMT project with an Atmel
90S2313 on a Dick Smith protoboard (Datak 12-611B for the curious) -
the CPU and other chips were through-hole, but I used some SMT caps to
pack enough goodies on a 24-pin-long protoboard. It was the first
time I used SMT parts on a protoboard, but it was so effective that it
won't be the last.
I guess I just got lucky enough to get into SMT projects in the early
1990s and didn't really worry about it. I can recommend that if you
are going to be doing a lot of them that the proper tools really make
things easier. I did a stint with a Dremmel-sized hot-air soldering
gun and an RF soldering iron and a foot-activated solder paste pump,
and I have to say that it was _nice_, but I don't own those tools
myself (wish I did). Even so, you can still do a lot with SOTs and
1206 resistors and a 1/32" soldering iron tip and tools that you
_ought_ to own for through-hole projects.
-ethan
e.stibler pontificated:
>Yup, get ready to solder .2 mm balls on 800 balls BGA.
>Get your toaster-oven ready ;-)
The really annoying thing about BGA is that I haven't figured a way to
desolder and reuse chips. Very vexing.
Dave McGuire wrote: Chuck Guzis wrote:
I suspect that the legacy DIP stuff will get harder to buy new through
"normal" channels, given RoHS. In the long run, I suppose we'd all better
get used to working with surface mount. :(
Nah, not anytime soon...unless I'm hallucinating, plenty of DIPs are
available as RoHS compliant parts now. There's nothing "legacy" about DIPs
just yet.
That said, the notion that surface mount is somehow "harder to solder" is
nothing but a load of crap. It is DIFFERENT...not harder. In fact, I find
all but the very smallest SMT parts EASIER to work with than through-hole
components.
I suspect that someone, way back when, was afraid of it and said "it's to
haaaard" and people have been propagating that rumor ever since.
-Dave
Dave,
Close up SMD work when your pushing 40 or 50 is tricky. I can easely handle
the finest pitch when I'm looking through a binoc microscope. It's quite
amusing to see a 30 gauge blue wire and solder the size of a truck while
trying to mod a board. Without optical assistance, I might as well be trying
to neuter a fly with a butcher knife.
Jim Davis.
Dennis Boone <drb at msu.edu> wrote:
> LOGINOUT exists:
LOGINOUT is the program that logs users and batch jobs in or out of VMS. If that didn't exist, I doubt very much that VMS would start up at all.
Look for a command file in TCPware that starts it all up. There will be a reference there to some device that is causing your problem: wrong disk, hardware/network device not functioning, etc. There should be a log file from the startup job. What does it say? It should point you to the source of the problem.
Jonas
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.405 / Virus Database: 268.10.9/417 - Release Date: 2006-08-11
On Fri, 11 Aug 2006 12:01:13 -0700, "Chuck Guzis" <cclist at sydex.com>
wrote:
> On 8/11/2006 at 11:08 AM Don wrote:
>
>> Didn't SDS make something like that?
>
> As in Scientific Data Systens, then XDS, (Xerox), then oblivion? I
> wasn't
> aware that they did. At one time async logic was a hot topic.
IIRC the PDP 9 (and possibly the 5 and 7) were async machines. That
was my first "personal" computer (nobody else wanted to touch it...)
and I spent many hour programming and keeping the beast running. When
the thing failed it was a real pain to troubleshoot. You would find a
signal on the backplane that was out of spec, but when you put the
card on the extension board to troubleshoot the damn thing would
start working. That and the unique control "memory" that would stop
working when things got damp...
CRC
Roy replied:
> Probably possible if you don't care at all about the board...
>
>Just take it outside, and hit the back side of the board with a torch. :-)
Tried the 400F oven trick. Works wonders with through-hole and many SMDs, but BGAs
either loose all the balls or the solder balls melt into a sheet. Not fun.
Small resistors and capacitors are hard to do this way too, they don't have enough mass to dislodge
with the shake.
Both that and the fourth reference seem to be talking about X.25, whatever
that is. (I probably knew what that was at some point but the brain ain't
what it used to be. :-)
X.25 - packet switched network standard. Most times you'd come across AX.25 - packet radio.
