8 Feb
2016
8 Feb
'16
12:43 p.m.
From: Guy Sotomayor
I also wanted to get folk's opinion on the need to
actually produce an
S[P]C form factor board. ... is it OK to have the MEM11 be outside of
the 11/20 chassis and connect via BC11A (my replica) cables?
Well, that's going to up the cost; for some people, that might be an issue.
Also, I dunno if there are people out there with table-top 11/15's-20's (they
did exist BITD, I worked with a table-top one), but for them, an additional
box might be a hassle too.
That's assuming of course that the power
requirements for the MEM11
can be fulfilled by a single SPC slot. ... in the worst case, it may
require splitting the MEM11 functionality across multiple boards.
I guess I don't see the harm in making it two SPC (quad) boards? A flat cable
or two to connect across (I dunno how extensive the interconnect requirements
between the halves would be, and I have forgotten what the inter-slot
interconnect capabilities of an SPC backplane are - ISTR that it has some
bussing on the F section pins) would be easy and cheap.
Noel
8 Feb
8 Feb
1:04 p.m.
On Feb 8, 2016, at 12:43 PM, Noel Chiappa <jnc at
mercury.lcs.mit.edu> wrote:
The reality (unfortunately) is that no matter what solution I take, it?s not
going to be inexpensive. The reality is that an SPC board will be more
expensive because of the gold edge fingers.
The other thing is that the boards will be fully assembled. Except for some
connectors and the UNIBUS transceivers, everything else is surface mount.
I?m still crossing my fingers that I can keep within the 208 pin PQFP for the
FPGA and not have to move into a BGA part.
From: Guy Sotomayor
I also wanted to get folk's opinion on the
need to actually produce an
S[P]C form factor board. ... is it OK to have the MEM11 be outside of
the 11/20 chassis and connect via BC11A (my replica) cables?
Well, that's going to up the cost; for some people, that might be an issue.
Also, I dunno if there are people out there with table-top 11/15's-20's (they
did exist BITD, I worked with a table-top one), but for them, an additional
box might be a hassle too.
See above re: gold edge fingers. I was originally thinking that if I do have to
split the board up, that I?d make them completely independent. But that has
the issue of requiring 2x the number of UNIBUS transceiver parts (which are
all but unobtainium as of now). One of the things that drives up the power (and
board area) are said transceivers (and level shifters, etc). If I could come up
with a reasonable alternative for the SPC version, that may work. But that?s
all in the future at the moment.
TTFN - Guy
That's assuming of course that the power
requirements for the MEM11
can be fulfilled by a single SPC slot. ... in the worst case, it may
require splitting the MEM11 functionality across multiple boards.
I guess I don't see the harm in making it two SPC (quad) boards? A flat cable
or two to connect across (I dunno how extensive the interconnect requirements
between the halves would be, and I have forgotten what the inter-slot
interconnect capabilities of an SPC backplane are - ISTR that it has some
bussing on the F section pins) would be easy and cheap. 
1:08 p.m.
On Mon, Feb 8, 2016 at 1:04 PM, Guy Sotomayor <ggs at shiresoft.com> wrote:
issue.
The reality (unfortunately) is that no matter what solution I take, it?s
not
going to be inexpensive. The reality is that an SPC board will be more
expensive because of the gold edge fingers.
The other thing is that the boards will be fully assembled. Except for
some
connectors and the UNIBUS transceivers, everything else is surface mount.
I?m still crossing my fingers that I can keep within the 208 pin PQFP for
the
FPGA and not have to move into a BGA part.
A thought: would a second quad board necessarily need transceivers? I'm
thinking of the top-block connectors used in the PDP-8, and the top-plugged
ribbon cables for e.g., MicroVAX II CPU-to-memory connection. You might
still want to grab power through a few fingers, but that's an
implementation detail. -- Ian
--
Ian S. King, MSIS, MSCS, Ph.D. Candidate
The Information School <http://ischool.uw.edu>
Dissertation: "Why the Conversation Mattered: Constructing a Sociotechnical
Narrative Through a Design Lens
Archivist, Voices From the Rwanda Tribunal <http://tribunalvoices.org>
Value Sensitive Design Research Lab <http://vsdesign.org>
University of Washington
There is an old Vulcan saying: "Only Nixon could go to China."
On Feb 8, 2016, at 12:43 PM, Noel Chiappa <jnc
at mercury.lcs.mit.edu>
wrote:
From: Guy Sotomayor
I also wanted to get folk's opinion on the
need to actually produce an
S[P]C form factor board. ... is it OK to have the MEM11 be outside of
the 11/20 chassis and connect via BC11A (my replica) cables?
Well, that's going to up the cost; for some people, that might be an
Also, I dunno if there are people out there with table-top 11/15's-20's
(they
did exist BITD, I worked with a table-top one),
but for them, an
additional
box might be a hassle too.
cable
That's assuming of course that the power
requirements for the MEM11
can be fulfilled by a single SPC slot. ... in the worst case, it may
require splitting the MEM11 functionality across multiple boards.
I guess I don't see the harm in making it two SPC (quad) boards? A flat or two to connect across (I dunno how extensive
the interconnect
requirements
between the halves would be, and I have forgotten
what the inter-slot
interconnect capabilities of an SPC backplane are - ISTR that it has some
bussing on the F section pins) would be easy and cheap.
See above re: gold edge fingers. I was originally thinking that if I do
have to
split the board up, that I?d make them completely independent. But that
has
the issue of requiring 2x the number of UNIBUS transceiver parts (which are
all but unobtainium as of now). One of the things that drives up the
power (and
board area) are said transceivers (and level shifters, etc). If I could
come up
with a reasonable alternative for the SPC version, that may work. But
that?s
all in the future at the moment.
TTFN - Guy
1:18 p.m.
On Feb 8, 2016, at 1:08 PM, Ian S. King <isking at
uw.edu> wrote:
I?m not a particular fan of that because some of the signals I?m running are
pretty fast between the FPGA and some of the other components (~40ns
cycle time for the FRAMs for example). I wouldn?t want to run those signals
very far and certainly not across any sort of cabling.
It?s not clear to me (yet) how I could partition the design across 2 SPC boards.
TTFN - Guy
See above re: gold edge fingers. I was originally thinking that if I do
have to
split the board up, that I?d make them completely independent. But that
has
the issue of requiring 2x the number of UNIBUS transceiver parts (which are
all but unobtainium as of now). One of the things that drives up the
power (and
board area) are said transceivers (and level shifters, etc). If I could
come up
with a reasonable alternative for the SPC version, that may work. But
that?s
all in the future at the moment.
TTFN - Guy
A thought: would a second quad board necessarily need transceivers? I'm
thinking of the top-block connectors used in the PDP-8, and the top-plugged
ribbon cables for e.g., MicroVAX II CPU-to-memory connection. You might
still want to grab power through a few fingers, but that's an
implementation detail. ? Ian
3240
days inactive
3240
days old
3 comments
3 participants
participants (3)
-
ggs＠shiresoft.com -
isking＠uw.edu -
jnc＠mercury.lcs.mit.edu