21 Sep
2008
21 Sep
'08
2:02 a.m.
There is the ocasional homebuilt, TTL CPU thread on here. I wanted to make known this
guy, John Plutorak, who reproduced the guidance computer in his basement:
http://klabs.org/history/build_agc/
Awsome project, take a look. I intend to build one too...
Randy Dawson
_________________________________________________________________
Get more out of the Web. Learn 10 hidden secrets of Windows Live.
http://windowslive.com/connect/post/jamiethomson.spaces.live.com-Blog-cns!5…
21 Sep
21 Sep
6:51 a.m.
Randy Dawson wrote:
There is the ocasional homebuilt, TTL CPU thread on
here. I wanted to make known this guy, John Plutorak, who reproduced the guidance
computer in his basement:
http://klabs.org/history/build_agc/
Awsome project, take a look. I intend to build one too...
I've seen this before and it's very interesting.
Here's a piece of trivia I picked up while researching material for my
last project at VCF east about Floating Point Math. I talked with Frank
O'Brian, the caretaker of the Apollo Guidance Computer exhibit at
InfoAge in NJ about this subject. He provided this tidbit of info that I
thought I could share here. While looking at the original code listings
for this computer, (at the exhibit) it's quite amazing how they
accomplished all this math within such a constrained environment given
the technology of it's day. I feel this would make some people scratch
their heads :)
What type of arithmetic operations were used to perform all the
calculations on the Apollo Guidance Computer ?
a) Floating Point arithmetic operations
b) Fixed Point Arithmetic operations
c) Integer Arithmetic operations
d) Slide Rule Arithmetic operations
e) all of the above
=Dan
[ Pittsburgh 250th --- http://www2.applegate.org/~ragooman/ ]
7:21 a.m.
I think a lot of this was done at Draper Lab, part of MIT then.
One person who worked on the code you are interested in was Richard Warren.
Steve
Dan wrote:
Randy Dawson wrote:
There is the ocasional homebuilt, TTL CPU thread
on here. I wanted
to make known this guy, John Plutorak, who reproduced the guidance
computer in his basement:
http://klabs.org/history/build_agc/
Awsome project, take a look. I intend to build one too...
I've seen this before and it's very interesting.
Here's a piece of trivia I picked up while researching material for my
last project at VCF east about Floating Point Math. I talked with
Frank O'Brian, the caretaker of the Apollo Guidance Computer exhibit
at InfoAge in NJ about this subject. He provided this tidbit of info
that I thought I could share here. While looking at the original code
listings for this computer, (at the exhibit) it's quite amazing how
they accomplished all this math within such a constrained environment
given the technology of it's day. I feel this would make some people
scratch their heads :)
What type of arithmetic operations were used to perform all the
calculations on the Apollo Guidance Computer ?
a) Floating Point arithmetic operations
b) Fixed Point Arithmetic operations
c) Integer Arithmetic operations
d) Slide Rule Arithmetic operations
e) all of the above
=Dan
[ Pittsburgh 250th --- http://www2.applegate.org/~ragooman/ ]
6:28 p.m.
Dan wrote:
I feel this would make some people scratch their heads
:)
What type of arithmetic operations were used to perform all the
calculations on the Apollo Guidance Computer ?
a) Floating Point arithmetic operations
b) Fixed Point Arithmetic operations
c) Integer Arithmetic operations
d) Slide Rule Arithmetic operations
e) all of the above
what ? everyone still scratching their heads on this one :)
Ok, the answer would be d) slide rule
If anyone remembers using their slide rule, you happen to be using
Fractional Arithmetic. If I wrote this down for answer d) it may have
been obvious to some, so I thru a little twist into it. The Apollo
programmers always coded their calculations as if they were using a
slide rule(this was the early 60's remember). They would have to keep
track on paper of the equation's dynamic range(also anticipate how large
this range could become) since the design didn't incorporate exponents
within the answer(as would be found in a FPU).
BTW, the HP-65 calculator was used as the backup computer on later
missions, since it was too programmable.
=Dan
[ Pittsburgh 250th --- http://www2.applegate.org/~ragooman/ ]
8:55 a.m.
On 21 Sep 2008 at 4:02, Randy Dawson wrote:
There is the ocasional homebuilt, TTL CPU thread on
here. I wanted to
make known this guy, John Plutorak, who reproduced the guidance
computer in his basement:
http://klabs.org/history/build_agc/
Awsome project, take a look. I intend to build one too...
I looked at a couple of documents, fully expecting to see ROM-on-a-
rope and noticed that the implementation was (mostly) SSI/MSI TTL.
You could probably simplify things in terms of construction
tremendously by using FPGA or CPLD methods. The result should be no
more or less accurate to the original than the TTL version.
Cheers,
Chuck
9:42 a.m.
Chuck Guzis wrote:
I looked at a couple of documents, fully expecting to
see ROM-on-a-
rope and noticed that the implementation was (mostly) SSI/MSI TTL.
You could probably simplify things in terms of construction
tremendously by using FPGA or CPLD methods. The result should be no
more or less accurate to the original than the TTL version.
But that is what made it special - the components at the time.
They have 3 input NOR's in tiny logic, so do smaller PCB's
if you want a smaller system. :)
Cheers,
Chuck
11:23 a.m.
On 21 Sep 2008 at 10:42, bfranchuk at jetnet.ab.ca wrote:
But that is what made it special - the components at
the time.
They have 3 input NOR's in tiny logic, so do smaller PCB's
if you want a smaller system. :)
One could also make use of transistor arrays to recreate the original
RTL part. Given that there were some 5000 of these packages in the
original (IIRC), that might make for a project a bit too large to
chew. Given that integration is used to decrease this package count
by 9/10ths, what's the problem with substituting some programmable
logic? How is that "less authentic" than substituting MSI TTL for
SSI RTL? One could use the extra space in the rack to store a six-
pack of one's favorite beverage. :)
Cheers,
Chuck
12:46 p.m.
Chuck Guzis wrote:
One could use the extra space in the rack to store a
six-
pack of one's favorite beverage. :)
I bet the astronauts back then wouldn't have mind that extra convenience :)
=Dan
[ Pittsburgh 250th --- http://www2.applegate.org/~ragooman/ ]
1 p.m.
On Sun, Sep 21, 2008 at 03:46:08PM -0400, Dan wrote:
Chuck Guzis wrote:
>One could use the extra space in the rack to store a six-
>pack of one's favorite beverage. :)
>
I bet the astronauts back then wouldn't have mind
that extra convenience :)
When did Tang come in six-packs? ;-)
-ethan
--
Ethan Dicks, A-333-S Current South Pole Weather at 21-Sep-2008 at 20:00 Z
South Pole Station
PSC 468 Box 400 Temp -72.4 F (-58.0 C) Windchill -119.8 F (-84.3 C)
APO AP 96598 Wind 20.4 kts Grid 70 Barometer 675.8 mb (10787 ft)
Ethan.Dicks at usap.gov http://penguincentral.com/penguincentral.html
12:56 p.m.
Dan wrote:
Yes, but I think part of this hobby is also having
some level of
authenticity.
Although many of the original parts can't be available, some level of
hardware consistency is maintained.
Another group redesigned the ENIAC into a FPGA chip.
Part of an engineer's job is to design within constraints, you lose that
when using advanced technology.
Eliminating this factor when trying to replicate a system is making it
too easy.
...
Besides what fun is for all to stand around the table
and look down on
one lonely FPGA chip ?
Chuck Guzis wrote:
On 21 Sep 2008 at 10:42, bfranchuk at jetnet.ab.ca wrote:
I'd say it's just a matter of preference of where one wants to draw the line.
Just for the sake of discussion, I'd suggest using SMD CMOS 4025 triple 3-input
NOR gates, or 7410 triple 3-input NAND gates with negative logic, to fairly
closely mimic the original implementation. (Might depend on whether wire-AND
connections were used much in the original.)
But that is what made it special - the components
at the time.
They have 3 input NOR's in tiny logic, so do smaller PCB's
if you want a smaller system. :)
One could also make use of transistor arrays to recreate the original
RTL part. Given that there were some 5000 of these packages in the
original (IIRC), that might make for a project a bit too large to
chew. Given that integration is used to decrease this package count
by 9/10ths, what's the problem with substituting some programmable
logic? How is that "less authentic" than substituting MSI TTL for
SSI RTL? One could use the extra space in the rack to store a six-
pack of one's favorite beverage. :)
9:46 a.m.
Chuck Guzis wrote:
On 21 Sep 2008 at 4:02, Randy Dawson wrote:
Yes, but I think part of this hobby is also having some level of
authenticity.
Although many of the original parts can't be available, some level of
hardware consistency is maintained.
Another group redesigned the ENIAC into a FPGA chip.
Part of an engineer's job is to design within constraints, you lose that
when using advanced technology.
Eliminating this factor when trying to replicate a system is making it
too easy.
You can also say writing an emulator is a form of a reproduction.
Since it's still running on hardware but with a different level of
programming, versus say, Verilog/Vhdl/C
Besides what fun is for all to stand around the table and look down on
one lonely FPGA chip ?
just my 2cents(melted down to 5cents of copper)
=Dan
--
[ Pittsburgh 250th --- http://www2.applegate.org/~ragooman/ ]
There is the ocasional homebuilt, TTL CPU thread
on here. I wanted to
make known this guy, John Plutorak, who reproduced the guidance
computer in his basement:
http://klabs.org/history/build_agc/
Awsome project, take a look. I intend to build one too...
I looked at a couple of documents, fully expecting to see ROM-on-a-
rope and noticed that the implementation was (mostly) SSI/MSI TTL.
You could probably simplify things in terms of construction
tremendously by using FPGA or CPLD methods. The result should be no
more or less accurate to the original than the TTL version
5936
days inactive
5937
days old
10 comments
7 participants
participants (7)
-
bfranchuk＠jetnet.ab.ca -
cclist＠sydex.com -
ethan.dicks＠usap.gov -
hilpert＠cs.ubc.ca -
ragooman＠comcast.net -
rdawson16＠hotmail.com -
steve＠radiorobots.com