12 Aug
2006
12 Aug
'06
2 a.m.
On 8/12/06, woodelf <bfranchuk at jetnet.ab.ca> wrote:
Still I would like to see in surface mount ---
tiny Flip/Clips and see how small you can build
a PDP-{Favorate number}.
I _have_ thought about that for replacement -8/L and -8/i boards -
design a simple board with, say, 4 to 6 16-pin SO pads and decoupling
caps and a dual row of jumper pads to cross-connect the edge fingers
to the various chip pads to be able to replicate, say, an M111 vs an
M117. The "problem" is that I doubt you could get real TTL (not
LS-TTL) parts in SO. I haven't experimented with replacing TTL with
LS parts in any of my -8s, so I don't know if there would be any
issues or not. So far, I've always had enough of the right things on
hand when I go to replace a chip on an M-series module. I would think
that the dimensions could be on the order of 2 x the size of just the
area of the fingers.
The advantage, of course, is that it would be easy to make a large
number of these on one PCB panel, reducing per-unit costs. Using one
(or maybe two at most) base designs would also help quantity issues.
You could sit down one day and make a stack of a dozen M111s, then
make two dozen M113s the next day with the same PCB; just add the
right SO-parts and configure the jumpers in the right order, and there
you are.
-ethan
12 Aug
12 Aug
3:03 a.m.
From: "Ethan Dicks" <ethan.dicks at gmail.com>
On 8/12/06, woodelf <bfranchuk at jetnet.ab.ca>
wrote:
> Still I would like to see in surface mount ---
> tiny Flip/Clips and see how small you can build
> a PDP-{Favorate number}.
It should be possible to do a 3/4 scale or so model using readily
available 0.1" pitch edge connectors instead of the unobtanium
0.125" pitch that DEC used. The gold adds a lot to the card
costs, though. (You'd be money ahead to switch to pin headers.)
Since switches and other parts have gotten similarly smaller,
that scale would probably work out OK.
If you wanted to set a size record of some sort, you could use
.05" pitch connectors, but that would surely be harder to find
and cost more.
I _have_ thought about that for replacement -8/L and
-8/i boards -
design a simple board with, say, 4 to 6 16-pin SO pads and decoupling
caps and a dual row of jumper pads to cross-connect the edge fingers
to the various chip pads to be able to replicate, say, an M111 vs an
M117. The "problem" is that I doubt you could get real TTL (not
LS-TTL) parts in SO. I haven't experimented with replacing TTL with
LS parts in any of my -8s, so I don't know if there would be any
issues or not. So far, I've always had enough of the right things on
hand when I go to replace a chip on an M-series module. I would think
that the dimensions could be on the order of 2 x the size of just the
area of the fingers.
The advantage, of course, is that it would be easy to make a large
number of these on one PCB panel, reducing per-unit costs. Using one
(or maybe two at most) base designs would also help quantity issues.
You could sit down one day and make a stack of a dozen M111s, then
make two dozen M113s the next day with the same PCB; just add the
right SO-parts and configure the jumpers in the right order, and there
you are.
I have CAD drawings for a regular size flipchip that uses "D" packages to
do all the common NAND and inverter stuff in the Mxxx series. You just
install the correct row of components for whatever you need today :-).
I don't know of anywhere you can get straight TTL in SO packages, but I
have 74F family stuff in "D" packages.
Vince
I also have CAD drawings of enough of the Gxxx and Mxxx module families
(using thru-hole packages) to build a TC08 from scratch, if one had that
kind of time and money. I stopped just short (so far) of adding all of
the necessary modules to build an 8/i. Except the core planes, of course,
which I have no way to build out of modern parts :-/.
6:44 a.m.
"Ethan Dicks" wrote:
I _have_ thought about that for replacement -8/L and -8/i boards -
Just curious, are there any machine readable schematics or netlists
for the 8/I?
(from my primative understanding an 8/i was made from a few different
flip chips and a *lot* of backplane wirewrap - just wondering if that's
machine readable anywhere)
On another topic, I do SMT all the time. You do need a few tools - a
nice maginifying lens w/light, tweezers, a very very fine tip (a low end
metcal is very helpful). A special hot air gun is also helpful for
removing parts with lots of pins. And if your pcb's are finished with
"hasl" - hot air solder layer I find you don't need flux as much (but it
does help).
-brad
9:13 a.m.
On 8/12/06, Brad Parker <brad at heeltoe.com> wrote:
On another topic, I do SMT all the time. You do need
a few tools - a
nice maginifying lens w/light, tweezers, a very very fine tip (a low end
metcal is very helpful). A special hot air gun is also helpful for
removing parts with lots of pins. And if your pcb's are finished with
"hasl" - hot air solder layer I find you don't need flux as much (but it
does help).
I picked up a surplus Hakko 850 hot air rework station for around $60,
the nozzles run around $15-25 on ebay and MPJA has a hot tweezer setup
for around $60. MPJA's price for a Hakko 851 clone is $99, so you can
set up a fair SMT work area for a couple of hundred at the most. Some
careful surplus and ebay shopping can minimize the cost if you aren't
in a big hurry.
James
--
www.blackcube.org - The Texas State Home for Wayward and Orphaned Computers
www.blackcube.org/personal/index.html - Personal web page
12:51 p.m.
From: "Brad Parker" <brad at heeltoe.com>
"Ethan Dicks" wrote:
Not that I am aware of. (I thought of creating one, but so far haven't
convinced myself to do it.)
I _have_ thought about that for replacement -8/L
and -8/i boards -
Just curious, are there any machine readable schematics or netlists
for the 8/I? (from my primative understanding an 8/i was made from
a few different
flip chips and a *lot* of backplane wirewrap - just wondering if that's
machine readable anywhere)
If it were, it would make the job of replacing all that wire-wrap with
a PCB tractable. It would also make it a lot easier to do things like
make an 8/i in an FPGA :-).
Vince
1 p.m.
Ethan Dicks wrote:
I _have_ thought about that for replacement -8/L and
-8/i boards -
design a simple board with, say, 4 to 6 16-pin SO pads and decoupling
caps and a dual row of jumper pads to cross-connect the edge fingers
to the various chip pads to be able to replicate, say, an M111 vs an
M117. The "problem" is that I doubt you could get real TTL (not
LS-TTL) parts in SO. I haven't experimented with replacing TTL with
LS parts in any of my -8s, so I don't know if there would be any
issues or not. So far, I've always had enough of the right things on
hand when I go to replace a chip on an M-series module. I would think
that the dimensions could be on the order of 2 x the size of just the
area of the fingers.
The advantage, of course, is that it would be easy to make a large
number of these on one PCB panel, reducing per-unit costs. Using one
(or maybe two at most) base designs would also help quantity issues.
You could sit down one day and make a stack of a dozen M111s, then
make two dozen M113s the next day with the same PCB; just add the
right SO-parts and configure the jumpers in the right order, and there
you are.
A variation on this theme might be to build a very small number of
different types of boards, but with a CPLD on each one...a small, cheap
one like a Xilinx XC9536 (something like $3/ea qty. 1)...then have a
JTAG connector on each board, and squirt the "personality" into the
board that way. I've not had a chance to work closely with
FlipChip-based machines (unfortunately) so I don't know how much
consistency there is on things like power pins and such, so I don't know
if this'd be practical, but I think it's an interesting idea.
-Dave
--
Dave McGuire
Cape Coral, FL
1:15 p.m.
Dave McGuire wrote:
Ethan Dicks wrote:
I think that skirts the appeal of a *real* "miniture PDP".
E.g., why not just do the whole design in an FPGA? It would
be a pretty boring accomplishment... :-(
OTOH, having the exact same design done in "miniature
technology" would be cool!
I've a friend who had a miniature pinball machine (intended
as a "bar top" device). Roughly 10"x24" with everything
scaled down (and made in stainless!). *That* was cool!
I _have_ thought about that for replacement -8/L
and -8/i boards -
design a simple board with, say, 4 to 6 16-pin SO pads and decoupling
caps and a dual row of jumper pads to cross-connect the edge fingers
to the various chip pads to be able to replicate, say, an M111 vs an
M117. The "problem" is that I doubt you could get real TTL (not
LS-TTL) parts in SO. I haven't experimented with replacing TTL with
LS parts in any of my -8s, so I don't know if there would be any
issues or not. So far, I've always had enough of the right things on
hand when I go to replace a chip on an M-series module. I would think
that the dimensions could be on the order of 2 x the size of just the
area of the fingers.
The advantage, of course, is that it would be easy to make a large
number of these on one PCB panel, reducing per-unit costs. Using one
(or maybe two at most) base designs would also help quantity issues.
You could sit down one day and make a stack of a dozen M111s, then
make two dozen M113s the next day with the same PCB; just add the
right SO-parts and configure the jumpers in the right order, and there
you are.
A variation on this theme might be to build a very small number of
different types of boards, but with a CPLD on each one...a small, cheap
one like a Xilinx XC9536 (something like $3/ea qty. 1)...then have a
JTAG connector on each board, and squirt the "personality" into the
board that way. I've not had a chance to work closely with
FlipChip-based machines (unfortunately) so I don't know how much
consistency there is on things like power pins and such, so I don't know
if this'd be practical, but I think it's an interesting idea.
3:56 p.m.
Don wrote:
Huh? I don't think so, at least not for me. I'm totally in love
with my SBC6120, which is at least a move in that direction. I can run
it 24/7 and just have it available on a serial port which I can get to
when I'm not home, for some OS/8 hacking when I get the itch.
That said, though, it'll never replace my "real" -8 machines. I have
several 8/e and one 8/m system, sadly no others...anybody got one (8/i,
8/l maybe) that might be up for trade?
A variation
on this theme might be to build a very small number of
different types of boards, but with a CPLD on each one...a small,
cheap one like a Xilinx XC9536 (something like $3/ea qty. 1)...then
have a JTAG connector on each board, and squirt the "personality" into
the board that way. I've not had a chance to work closely with
FlipChip-based machines (unfortunately) so I don't know how much
consistency there is on things like power pins and such, so I don't
know if this'd be practical, but I think it's an interesting idea.
I think that skirts the appeal of a *real* "miniture PDP".
E.g., why not just do the whole design in an FPGA? It would
be a pretty boring accomplishment.... :-( OTOH, having the exact same design done in
"miniature
technology" would be cool!
I've a friend who had a miniature pinball machine (intended
as a "bar top" device). Roughly 10"x24" with everything
scaled down (and made in stainless!). *That* was cool!
Wow, that does sound cool!
-Dave
--
Dave McGuire
Cape Coral, FL
4:07 p.m.
Dave McGuire wrote:
Don wrote:
Huh? I don't think so, at least not for me. I'm totally in love with
my SBC6120, which is at least a move in that direction. I can run it
24/7 and just have it available on a serial port which I can get to when
I'm not home, for some OS/8 hacking when I get the itch.
But, you can make the move all the way "in that direction"
and just run an emulator. :> I.e. the appeal is having
the actual hardware. And, while PDP's are reasonably
commonplace, having a *tiny* one (imagine a 10x10x10 8/i
sitting on your COFFEE TABLE :> ) would be truly unique.
A
variation on this theme might be to build a very small number of
different types of boards, but with a CPLD on each one...a small,
cheap one like a Xilinx XC9536 (something like $3/ea qty. 1)...then
have a JTAG connector on each board, and squirt the "personality"
into the board that way. I've not had a chance to work closely with
FlipChip-based machines (unfortunately) so I don't know how much
consistency there is on things like power pins and such, so I don't
know if this'd be practical, but I think it's an interesting idea.
I think that skirts the appeal of a *real* "miniture PDP".
E.g., why not just do the whole design in an FPGA? It would
be a pretty boring accomplishment.... :-( That said, though, it'll never replace my
"real" -8 machines. I have
several 8/e and one 8/m system, sadly no others...anybody got one (8/i,
8/l maybe) that might be up for trade?
Quite interesting as the ballistics are very different
as you scale things down. And, the mechanisms need to be
completely reengineered as they don't scale well, either!
OTOH, having the exact same design done in
"miniature
technology" would be cool!
I've a friend who had a miniature pinball machine (intended
as a "bar top" device). Roughly 10"x24" with everything
scaled down (and made in stainless!). *That* was cool!
Wow, that does sound cool! 
4:29 p.m.
Don wrote:
> Huh? I don't think so, at least not for me.
I'm totally in love
> with my SBC6120, which is at least a move in that direction. I can
> run it 24/7 and just have it available on a serial port which I can
> get to when I'm not home, for some OS/8 hacking when I get the itch.
But, you can make the move all the way "in that
direction"
and just run an emulator. :> I.e. the appeal is having
the actual hardware. And, while PDP's are reasonably
commonplace, having a *tiny* one (imagine a 10x10x10 8/i
sitting on your COFFEE TABLE :> ) would be truly unique.
The SBC1620 is tiny -- the PCB is 6" x 4 1/4". My problem is
all I got is Pee-cee parts like a power supply and case.
That is the parts that are harder to find here.
> That said, though, it'll never replace my
"real" -8 machines. I
> have several 8/e and one 8/m system, sadly no others...anybody got one
> (8/i, 8/l maybe) that might be up for trade?
Boy that shows our age ... PDP-8's are modern machines still. :)
Well at least you know what speed the machine ran at, with todays
marketing double talk and bloatware who knows what they run at.
PS.
Anybody got a DEAD 8 I stuff my SBC1620 in Canada?
4:46 p.m.
You wrote:
Eh. I use emulators frequently, but there's more to classic
computing than software.
Well it depends on your definition of "modern". There are people who
seem to think x86 is "modern"...PDP-8s were still in their heyday when
that train wreck of an architecture was designed!
Huh? I
don't think so, at least not for me. I'm totally in love
with my SBC6120, which is at least a move in that direction. I can
run it 24/7 and just have it available on a serial port which I can
get to when I'm not home, for some OS/8 hacking when I get the itch.
But, you can make the move all the way "in that direction"
and just run an emulator. :> I.e. the appeal is having the actual hardware.
Exactly. :)
And, while PDP's are reasonably
commonplace, having a *tiny* one (imagine a 10x10x10 8/i
sitting on your COFFEE TABLE :> ) would be truly unique.
Absolutely!
The SBC1620 is tiny -- the PCB is 6" x 4
1/4". My problem is
all I got is Pee-cee parts like a power supply and case.
That is the parts that are harder to find here.
I think it's time to take up a collection for you. ;)
That said,
though, it'll never replace my "real" -8 machines. I
have several 8/e and one 8/m system, sadly no others...anybody got
one (8/i, 8/l maybe) that might be up for trade?
Boy that shows our age ... PDP-8's are modern machines still. :) Well at least you know what speed the machine ran at,
with todays
marketing double talk and bloatware who knows what they run at.
Yup. :-(
-Dave
--
Dave McGuire
Cape Coral, FL
13 Aug
13 Aug
4:04 a.m.
Don wrote:
But, you can make the move all the way "in that direction"
and just run an emulator. :> I.e. the appeal is having
the actual hardware. And, while PDP's are reasonably
commonplace, having a *tiny* one (imagine a 10x10x10 8/i
sitting on your COFFEE TABLE :> ) would be truly unique.
If you're going to build a miniature system out of modern components, why not
just use a FPGA? :-)
Once the decision's been made to not stick with the original hardware, then
surely any construction technique is fair game?
--
A. Because it destroys the natural flow of conversation.
Q. What's wrong with top posting ?
I think that skirts the appeal of a *real*
"miniture PDP".
E.g., why not just do the whole design in an FPGA? It would
be a pretty boring accomplishment.... :-(
Huh? I don't think so, at least not for me. I'm totally in love
with my SBC6120, which is at least a move in that direction. I can
run it 24/7 and just have it available on a serial port which I can
get to when I'm not home, for some OS/8 hacking when I get the itch.
3:36 a.m.
Jules Richardson wrote:
Don wrote:
But, you can make the move all the way "in that direction"
and just run an emulator. :> I.e. the appeal is having
the actual hardware. And, while PDP's are reasonably
commonplace, having a *tiny* one (imagine a 10x10x10 8/i
sitting on your COFFEE TABLE :> ) would be truly unique.
If you're going to build a miniature system out of modern components,
why not just use a FPGA? :-)
That was the point I made elsewhere. The novelty of
making a miniature system out of the same components
(more or less) that the original device was made with
just using newer packaging (e.g., SMT).
If an FPGA is fair game, then why not just use an
emulation?
Same reason that a miniature pinball machine is
"cool" while a "video pinball machine" quickly
becomes a yawner...
I think that skirts the appeal of a *real*
"miniture PDP".
E.g., why not just do the whole design in an FPGA? It would
be a pretty boring accomplishment.... :-(
Huh? I don't think so, at least not for me. I'm totally in love
with my SBC6120, which is at least a move in that direction. I can
run it 24/7 and just have it available on a serial port which I can
get to when I'm not home, for some OS/8 hacking when I get the itch. Once the decision's been made to not stick with
the original hardware,
then surely any construction technique is fair game?
10:23 a.m.
On Aug 13, 2006, at 3:36 AM, Don wrote:
If an FPGA is fair game, then why not just use an
emulation?
This is a very good point, and (personally) I don't have an answer
for it, so I think it just boils down to personal preference. I am
bored to tears with emulators...but an FPGA implementation of a classic
machine is something that I find very drool-worthy.
I think it's a matter of the feeling of "fakeness". The FPGA
implementation is REAL HARDWARE...Real logic gates implementing an
architecture, instead of emulation...which is SIMulation.
-Dave
--
Dave McGuire
Cape Coral, FL
4:02 p.m.
Dave McGuire wrote:
On Aug 13, 2006, at 3:36 AM, Don wrote:
<Devil'sAdvocateMode>
Except that each one of the logic operations emulated are executed on a
real ALU, using similar devices.
For example, when you run an AND opcode through an emulator, at some
point, the real CPU running the emulator does run it through a set of
AND gates.
When you do an ADD, or SUB that ADD or SUB goes through an ALU running a
similar instruction.
When you do a read or store from memory, that read or write goes on a
memory bus, and so on.
So it's not quite simulation. It also runs on real hardware. Granted,
the instructions for how to do the emulation are in software, but they
run on real hardware. You could look at this as being "interpreted" on
a real computer, and "compiled" on an FPGA.
Sort of. :-)
</Devil'sAdvocateMode>
If an FPGA is fair game, then why not just use
an
emulation?
This is a very good point, and (personally) I don't have an answer
for it, so I think it just boils down to personal preference. I am
bored to tears with emulators...but an FPGA implementation of a
classic machine is something that I find very drool-worthy.
I think it's a matter of the feeling of "fakeness". The FPGA
implementation is REAL HARDWARE...Real logic gates implementing an
architecture, instead of emulation...which is SIMulation.
4:43 p.m.
On 8/13/2006 at 4:02 PM Ray Arachelian wrote:
So it's not quite simulation. It also runs on real
hardware. Granted,
the instructions for how to do the emulation are in software, but they
run on real hardware. You could look at this as being "interpreted" on
a real computer, and "compiled" on an FPGA.
Yes, but to my mind, absent the periperals, the difference is scant enough
not to matter. Emulation has been with us almost as long as computers.
IIRC, the IBM 705 had an emulation program for the 650. S/360 could
emulate a number of older mainframes. I've often wondered if anyone
"nested' emulations on their S/360.
But it's all still digital electronics--electrons and holes and bits.
Mostly the packaging has changed. And if the packaging's REALLY the thing,
then we should be content with mock-ups with PeeCees behind the fa?ade
running the blinkenlights and sampling the switches to complete the
effect.
Where I find older equipment to be useful is running older peripherals.
All the rest is nostalgia. Since my salad days were spent with the really
big iron, I'm not likely to recreate them anytime soon. Those who spent
their good times with minicomputers have more of a shot at encapsulating
their happy times in amber.
Cheers,
Chck
5:10 p.m.
Ray Arachelian wrote:
<Devil'sAdvocateMode>
Except that each one of the logic operations emulated are executed on a
real ALU, using similar devices.
For example, when you run an AND opcode through an emulator, at some
point, the real CPU running the emulator does run it through a set of
AND gates.
When you do an ADD, or SUB that ADD or SUB goes through an ALU running a
similar instruction.
Using ADD or SUB gates, as appropriate... <snicker>
You could look at this as being
"interpreted" on
a real computer, and "compiled" on an FPGA.
Sort of. :-)
Actually, not a bad analogy!
12 Aug
12 Aug
1:23 p.m.
From: "Dave McGuire" <mcguire at neurotica.com>
A variation on this theme might be to build a very
small number of
different types of boards, but with a CPLD on each one...a small, cheap
one like a Xilinx XC9536 (something like $3/ea qty. 1)...then have a
JTAG connector on each board, and squirt the "personality" into the
board that way. I've not had a chance to work closely with
FlipChip-based machines (unfortunately) so I don't know how much
consistency there is on things like power pins and such, so I don't know
if this'd be practical, but I think it's an interesting idea.
The good news is that there are at least a dozen flipchips I can think
of where that approach should work fine (provided your CPLD uses TTL
levels).
The bad news is that those flipchips are the most common ones. (Several
are on ePay at the moment for $6+shipping.)
The problem that I see, is nearly every device made out of flipchips
uses one (usually more) of the harder to find flipchips. Worst are
the green-handled Gxxx series, which typically have custom linear
circuits on them. Also problematic are about half the Mxxx series,
the ones that use open-collector outputs, negative logic voltage
levels, etc. But much of my gear is looking for one or another of
these "special" modules.
Of course, the older Rxxx/Sxxx modules also use the negative levels,
making them more difficult to do with modern logic components. (The
Wxxx are a mixed bag; a little of everything, including simple connectors.)
Vince
3:59 p.m.
vrs wrote:
The good news is that there are at least a dozen
flipchips I can think
of where that approach should work fine (provided your CPLD uses TTL
levels).
The bad news is that those flipchips are the most common ones. (Several
are on ePay at the moment for $6+shipping.)
The problem that I see, is nearly every device made out of flipchips
uses one (usually more) of the harder to find flipchips. Worst are
the green-handled Gxxx series, which typically have custom linear
circuits on them. Also problematic are about half the Mxxx series,
the ones that use open-collector outputs, negative logic voltage
levels, etc. But much of my gear is looking for one or another of
these "special" modules.
Of course, the older Rxxx/Sxxx modules also use the negative levels,
making them more difficult to do with modern logic components. (The
Wxxx are a mixed bag; a little of everything, including simple connectors.)
Hmm yes, that would be problematic. :-( I don't know of a way around
that problem offhand.
-Dave
--
Dave McGuire
Cape Coral, FL
6559
days inactive
6560
days old
18 comments
10 participants
participants (10)
-
bfranchuk＠jetnet.ab.ca -
brad＠heeltoe.com -
cclist＠sydex.com -
ethan.dicks＠gmail.com -
james.rice＠gmail.com -
julesrichardsonuk＠yahoo.co.uk -
mcguire＠neurotica.com -
ray＠arachelian.com -
THX1138＠dakotacom.net -
vrs＠msn.com