10 Dec
2002
10 Dec
'02
9 a.m.
Hi all,
Well, it looks like I've finally destroyed the Ace. I've just spent the
past hour trying to desolder the dead RAMs and buffers from the Ace's main
PCB. Unfortunately it looks like the board was designed to self-destruct
when anyone tried to repair it.
The pads appear to have been designed to peel off on the application of
heat, they're less than 5 mils around the hole (what do you think that
means?) and they don't even seem to be through-hole plated. The tin plating
was applied straight on top of oxidised copper - I've had to retin some pads
and tracks courtesy of that major screwup.
In my opinion, the Jupiter Ace is one of the most appallingly-made
machines I've ever tried to repair. Heck, the Commodore 64 was bad, but at
least the pads were easy enough to desolder. It looks to me like Jupiter
Cantab's PCB designer was either *VERY* inexperienced or just wanted to make
sure that no-one could fix an Ace if it failed. It's also beginning to look
like the ROMs are stuffed, but that shouldn't be too hard to sort out --
I've just bid on some 2532 EPROMs on eBay from someone in Austria.
Does anyone know how I could rescue this machine? It looks like the RAMs
are definetly fried, along with some of the logic as well. Font RAM and
Video RAM are still not being loaded on startup so the output of the video
generator is still 100% noise, however it *is* changing when the machine is
powered off and then back on again. I'm shotgunning all the RAMs (there's
only six of them) and the bus muxes.
Has anyone here either repaired one of these machines or got a spare Ace
to sell me? I've got a proper PSU now, with only one connector (the jack
plug the Ace uses), so I can say with near absolute certainty that the same
mistake will not occur again.
Later.
--
Phil.
philpem(a)dsl.pipex.com
http://www.philpem.dsl.pipex.com/
10 Dec
10 Dec
9:16 a.m.
Phil,
Most likely the board was designed for low cost. If there is no
soldermask, larger annular rings increase the chance of solder bridges
during manufacturing. But more likely, smaller annular rings were used to
reduce the amount of solder needed, thus lowering the cost. I don't believe
they Ace nor the ZX81 nor any of that genre of computers was designed to be
repaired. They were designed to be produced as cheaply as possible.
--John
-----Original Message-----
From: cctalk-admin(a)classiccmp.org [mailto:cctalk-admin@classiccmp.org]On
Behalf Of Philip Pemberton
Sent: Tuesday, December 10, 2002 15:04
To: cctalk(a)classiccmp.org
Subject: One more screwup with the Ace...
Hi all,
Well, it looks like I've finally destroyed the Ace. I've just spent the
past hour trying to desolder the dead RAMs and buffers from the Ace's main
PCB. Unfortunately it looks like the board was designed to self-destruct
when anyone tried to repair it.
The pads appear to have been designed to peel off on the application of
heat, they're less than 5 mils around the hole (what do you think that
means?) and they don't even seem to be through-hole plated. The tin plating
was applied straight on top of oxidised copper - I've had to retin some pads
and tracks courtesy of that major screwup.
In my opinion, the Jupiter Ace is one of the most appallingly-made
machines I've ever tried to repair. Heck, the Commodore 64 was bad, but at
least the pads were easy enough to desolder. It looks to me like Jupiter
Cantab's PCB designer was either *VERY* inexperienced or just wanted to make
sure that no-one could fix an Ace if it failed. It's also beginning to look
like the ROMs are stuffed, but that shouldn't be too hard to sort out --
I've just bid on some 2532 EPROMs on eBay from someone in Austria.
Does anyone know how I could rescue this machine? It looks like the RAMs
are definetly fried, along with some of the logic as well. Font RAM and
Video RAM are still not being loaded on startup so the output of the video
generator is still 100% noise, however it *is* changing when the machine is
powered off and then back on again. I'm shotgunning all the RAMs (there's
only six of them) and the bus muxes.
Has anyone here either repaired one of these machines or got a spare Ace
to sell me? I've got a proper PSU now, with only one connector (the jack
plug the Ace uses), so I can say with near absolute certainty that the same
mistake will not occur again.
Later.
--
Phil.
philpem(a)dsl.pipex.com
http://www.philpem.dsl.pipex.com/
10:18 a.m.
Hi all,
Well, it looks like I've finally destroyed the Ace...
Ow!
Your problems were probably caused by aggressive engineering cost
reduction. "Quality" boards are made of fiberglass and resin, have
plated vias and a solder mask. Many low-end products are missing
one or more of those features (mind you, a $100 TV set counts as low
end in terms of board quality).
If you were using a temperature controlled iron, one way to minimize
thermal damage is to have the iron barely above the melting point
of your solder. If your iron is a fixed temp, you can select a
lighter element (fewer watts to pump into the board). Practice
can also help. I remember obliterating boards as a kid. My Quest
Elf has a number of scorch marks and blobby solder joints. I keep
thinking about scanning it for an example, then cleaning it off and
repopulating it with decent sockets. If I ever finish the external
front panel and the computer doesn't come right up, that's probably
what I'll do - deconstruct and reconstruct with better parts.
Since you are past that point, if you can't smooth things out by
hand, they do make replacement eyes and lands and traces, but
they aren't cheap. Another Ace might be cheaper.
I've just bid on some 2532 EPROMs on eBay from
someone in Austria.
If you don't win the bid, I can probably help you with a couple of
2532s. If you get me the contents as a binary or Intel Hex file, I
can program them first.
B.G. Micro has 2532s for $2.00 each or 10 for $17.50.
Does anyone know how I could rescue this machine?
It looks like the
RAMs are definetly fried, along with some of the logic as well...
If it were mine and I didn't have a logic probe or oscilloscope and
had to resort to "shotgunning" RAMs, I'd at least put machined-pin
sockets in every place I pulled a chip from. It's not an awesome price,
it's kindof OK, but B.G. Micro has a variety of sized (14-pin to 40-pin
for 3/$1.00). At least then, you won't be lifting any _more_ pads. If
visual access is an issue (i.e., you want to see under the socket to
ensure there isn't any hidden damage), you can install machined pin strips
or break individual pins from the sockets and solder them in without a
plastic frame. It's harder to insert an IC, but you can see the board.
Has anyone here either repaired one of these
machines or got a spare
Ace to sell me?
Sorry. Never seen an Ace, just the ZX-80 and ZX-81. From all the
discussion, I'm getting a good picture in my head (and I even toyed
with the idea of building one from scratch), but for now, I have to
focus on getting my SBC6120 built.
Cheers,
-ethan
__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com
11:15 a.m.
Ethan Dicks wrote:
I once made the mistake of saying to a Motorola Plant manager something
similar to what you just said (this plant made single sided boards on
CEM material.) His comment was that they made high quality inexpensive
(not cheap) boards. Also, plated thru vias guarantee nothing; it is the
amount of copper plating in the holes that make the difference. Less
than 1 mil of copper plating can result in blowholes and unsatisfactory
reflowing of the tin-lead into solder. Etc, etc, etc :).
BTW, the circuit boards for the Atari 2600 were made of CEM and were
laid up 8 boards per panel. The process control was the same as higher
cost boards.
Hi all,
Well, it looks like I've finally destroyed the Ace...
Your problems were probably caused by aggressive engineering cost
reduction. "Quality" boards are made of fiberglass and resin, have
plated vias and a solder mask. Many low-end products are missing
one or more of those features (mind you, a $100 TV set counts as low
end in terms of board quality). 
12:04 p.m.
J.C.Wren wrote:
Phil,
Most likely the board was designed for low cost. If there is no
soldermask, larger annular rings increase the chance of solder bridges
during manufacturing. But more likely, smaller annular rings were used to
reduce the amount of solder needed, thus lowering the cost. I don't believe
they Ace nor the ZX81 nor any of that genre of computers was designed to be
repaired. They were designed to be produced as cheaply as possible.
Or any computers made since then! Cheap Cheap Cheap.!
1:56 p.m.
A non-text attachment was scrubbed...
Name: not available
Type: text
Size: 1985 bytes
Desc: not available
Url :
http://www.classiccmp.org/pipermail/cctalk/attachments/20021210/8a0dd4ba/at…
7896
days inactive
7896
days old
5 comments
6 participants
participants (6)
-
ard＠p850ug1.demon.co.uk -
bfranchuk＠jetnet.ab.ca -
erd_6502＠yahoo.com -
jcwren＠jcwren.com -
marvin＠rain.org -
philpem＠dsl.pipex.com