21 Oct
2001
21 Oct
'01
3:46 p.m.
For the power and ground connections for each chip, a
friend suggested I
should solder 16-gauge wire from a tie point or the power connector
straight to each chip and/or its bypass cap. The only 16-gauge wire I have
is stranded, and it's a mess to work with -- too large to work with easily
and the strands mean it won't really bend and stay on the tie point or pin.
What kind of juice do those chips need? 16 gauge seems awfully heavy for
any ICs I have worked with. I assume you know guages, but 16-18 guage is
standard lamp cord and will let you draw 5-10 amps on 110 volt. This
sounds good for a main power input to a transformer, but all ICs I have
worked with only pull at max a few amps at much lower voltages, so you
should be able to use much thinner wire.
I would check your amp requirements, and move to an 18 or 20 guage solid
or even higher if you can... phone station cord is usually 24 guage, but
don't mix it up with the 26 guage multipair cord... the usual obvious
difference is, station cord is 4 colors red/green/yellow/black....
multipair cord is 2 (or more) pairs of striped wire (blue, orange, green,
brown, slate all striped with white for the first 6 pairs)
But that aside, if you strip the standed cord, tiwst the strands in your
finger, and then tin the end with your solder iron (heat the end, and let
it draw solder into it, until it has a nice coat), it will be MUCH more
managable (although it still in general won't stay where you bend it,
other than the tinned end, lightweight alligator clips work wonders for
holding things in place).
-chris
<http://www.mythtech.net>
21 Oct
21 Oct
8:35 p.m.
This sounded like overkill to me.
Best thing to do, is to have a WW board with pads and
busses. I've had no problems using ordinary 30 ga WW
wire for power and ground following these rules:
0. ALWAYS use "modified wrap" with one turn of
insulation as a strain relief.
1. No "christmas treeing" power from one chip to the
next.
2. 0.1-0.01 uF bypass caps on EVERY chip. Especially
if you use HTTL or TTL instead of LSTTL. Straight TTL
draws frightening amounts of current at the instant of
a switch.
3. No more than 1" from bus to chip on the +5 and GND.
4. Use ONLY machine tooled sockets. I like Jameco's,
they aren't the best but they are good, inexpensive
for what they are, and not surplus, and therefore work
(and look) consistently. Gold on the "cup" is
important-I don't think gold plated pins is as
important if you are wrapping correctly. Yeah, I know
side-wipe sockets are cheaper - but how much is your
time worth. If you are reading this, quite a bit, I'll
bet.
5. _Never_, _never_ reuse a wire wrap wire if you make
a mistake and unwrap it - even if you "just started".
It will definitely break, if not now (and gum up your
tool), maybe later, when you'll have to find it in a
jumble of hundreds of wires. Or worse, a bare piece
will float around on your board.
6. It's nice to do the +5/GND in another color, and do
it all first.
If you don't have a plated board, I would run 16 ga
_solid_ wire up and down parallel to the chips, power
them, and work from that. You can tie them to the
board with WW wire from the top if you can't solder
them.
I've seen WW style guides that say, "no wiring under
chips". I don't understand this. wiring across chips
brings noise from crosstalk down and shortens the
wiring considerably, also desirable. It looks visually
messy, but it looks good electrically on my scope.
Keep the wiring "short", but NEVER, NEVER "tight".
There should never be mechanical tension on any wire.
If you have a plated board (and have no intention of
reusing it), I have no problems soldering discretes to
the board, wrapping resistors, and then tack-soldering
to them to permanently connect them. Using headers
(expensive for some reason unless you know where to
find them), and sockets is extreme.
Don't make mistakes - be methodical. care in the
beginning will pay off with few or no nightmares in
the end. Check for mistakes _often_ and use a VOM with
a continuity tester to buzz out connections. You have
to get used to working backwards (the chips are
"backwards" looking from the bottom up compared to the
pinouts in a book). I don't find this kind of work the
kind where you can spend a 1/2 hour here, and 1/2 hour
there. Unless you are a wizard, no TV or radio for
distractions.
Instead of the smoke test, use the VOM to power the
board (with expensive chips pulled), red lead to +5,
blk to ground. On a medium complexity board,
resistance should be 100-500 ohms. I have yet to hurt
any TTL chip doing this, because the current is so low
on a modern meter.
Using these techniques, I've gotten 4004, 8008, and
8008 boards of 15-30 MOS and TTL chip complexity in
about 5 x 8" boards, running at decent speeds, to work
(first crack in some cases!!) and have good clean
signals (see 8080geek.freeservers.com).
I don't do this professionally, I have an English
degree, but these statements come from experience.
Comments?
__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com
22 Oct
22 Oct
10:07 a.m.
A non-text attachment was scrubbed...
Name: not available
Type: text
Size: 3697 bytes
Desc: not available
Url :
http://www.classiccmp.org/pipermail/cctalk/attachments/20011022/2e2b98f7/at…
23 Oct
23 Oct
8:48 a.m.
On Mon, Oct 22, 2001 at 08:07:57PM +0100, Tony Duell wrote:
Interesting, I've never had a problem using manual slit-n-wrap. I also
think that it's easier to trace than straight wire wrap unless you cut
each wire to the proper length.
john
--
************************************************************************
* * *
* John Ott * Email: jott(a)hamming.ee.nd.edu *
* Dept. Electrical Engineering * *
* 275 Fitzpatrick Hall * *
* University of Notre Dame * Phone: (219) 631-7752 *
* Notre Dame, IN 46556 USA * Fax: (219) 631-9924 *
* * *
************************************************************************
Best thing to do, is to have a WW board with pads and
busses. I've had no problems using ordinary 30 ga WW
wire for power and ground following these rules:
0. ALWAYS use "modified wrap" with one turn of
insulation as a strain relief.
And ignore those darn slit-n-wrap tools (unless you can afford one of the
professional motorised ones). The hand ones seem to give about 50% bad
connections. No thanks! 
22 Oct
22 Oct
12:21 p.m.
Thanks for all the feedback! I'll let y'all know how it turns out..
Gordon
1:21 p.m.
New subject: zaft@azstarnet.com
Since I've been in the business of proof-of-concept prototyping for the past 25
years, I thought I'd wait until everyone else weighs in on this matter. I have
found very few wirewrap cards capable of distributing power properly, so I have
made my own wherever practical. The boards I make have a continuous ground
plane on the wiring side of the board and a continuous Vcc plane on the
component side, with a ground strip around the exposed edges of the board, away
from the backplane connector. In cases where no
backplane is involved, the
ground strip on both sides is available at each edge.
This facilitates ground
reference attachment for 'scope and logic analyzer.
The holes are plated through and the board is coated with a dry-film solder
mask, thereby facilitating the use of bare-wire analog circuit construction
where it's needed. I've run fairly large synchronous circuits on a board of
this type, with as many as 250 bipolar flipflops toggling at more or less the
same time, i.e. from the same, albeit separately buffered, clock edge, without
harmful effects, all the while monitoring the power<=>ground noise to be <100mv.
The trick, of course, is that the attachment to the ground and power plane is
normally accomplished by means of a soldered-on capacitor of relatively small
value, typically 0.01 microfarad. Since one lead of a bypass capacitor is
attached to the power pin of each IC, the other can be attached very near the
capacitor by soldering it directly to the ground plane. So long as one keeps
the leads shorter than about 1mm, the maximal benefit from bypass will be
obtained.
I still use boards of this type, albeit seldom with the solder mask, since I buy
the big "colander" panels at RadioShack.com and cut them to fit. I still
isolate the outer edges of the continuous power plane on the top of the board by
relieving it from the main plane by means of a dremel too with a saw blade, and
then connect the edges to ground with a solder bridge wherever it's convenient.
That way I can still attach probe grounds wherever I need.
There are few ways, short of having them made up, which is what I did, to get
decent wire-wrap boards that will accomodate the connectors and IC packages
you'll need. The RS.com store located here in Denver has large panels of this
ground and power plane equipped perf board with plated-through holes, for about
$100, which is plenty, but the boards are big enough to warrant it.
regards,
Dick
----- Original Message -----
From: "Gordon Zaft" <zaft(a)azstarnet.com>
To: <classiccmp(a)classiccmp.org>
Sent: Monday, October 22, 2001 3:21 PM
Subject: Re: OT: power & ground for wirewrap boards
Thanks for all the feedback! I'll let y'all know how it turns out..
Gordon
6:33 p.m.
New subject: Anti Spam sites
Here are some links to anti spam sites
http://www.abuse.net/
http://members.hostedscripts.com/antispam.html link your webpage to this
site and spammers will get 100 new random character bad email addresses to
add to their database.
http://spamcop.net/
9:54 a.m.
A non-text attachment was scrubbed...
Name: not available
Type: text
Size: 1279 bytes
Desc: not available
Url :
http://www.classiccmp.org/pipermail/cctalk/attachments/20011022/513fd958/at…
8286
days inactive
8288
days old
7 comments
7 participants
participants (7)
-
ard＠p850ug1.demon.co.uk -
edick＠idcomm.com -
jott＠mastif.ee.nd.edu -
mikeford＠socal.rr.com -
mythtech＠Mac.com -
sloboyko＠yahoo.com -
zaft＠azstarnet.com