9 Jan
2019
9 Jan
'19
11:23 a.m.
From: Bill Gunshannon
I have a number of different memory modules. Mostly
DEC but a couple
zthird party. Here's the problem. None of them are reflected in any of
the documentation I have been able to find so I can't configure them
away from their defaults! ...
Anybody got any pointers to help me configure some of this stuff?
M7551-AC - All the docs I can find seem to refer to
AA or AB and jumpers and switches are not
in the same locations.
You've looked in EK-MSV1Q-UG-002, right? That seems to cover a couple of
different revisions.
M8067-LB
M8067-LF
M8067-LJ - Same problem. I can find no documentation for any -L
boards and these don't even resemble the pictures I find.
The M8067-L variants are all MSV11-PL (512-Kbyte QBUS MOS memory). The
last letter usually indicates the vendor of the MOS chips used.
And then I have two non-DEC module that are unlikely
to
have any documentation still floating around for.
Camintonn CMV-1000
I too couldn't find any documentatin for this; there is the 'SMS 1000
Microcomputer System OEM Manual', which says how to configure one for a base
configuration. I started to work out what the configuration switches do, by
experiment, but I got distracted before I finished. I have found my notes
from this exercise, and can send them to you if they
are of any use.
Noel
9 Jan
9 Jan
11:28 a.m.
As far as I know about the 11/93 (or 11/94), all memory is on the CPU module.
The 11/9[3|4] is sort-of an 11/8[3|4] with all (PMI) memory on the CPU module as the cache
on the 8[3|4].
The 11/9[3|4] with 2 MB RAM can only be upgraded to 4 MB by adding RAM chips on the CPU
module.
However, all holes are filled with solder, so this upgrade is quite a lot of work.
Henk
11:41 a.m.
On 1/9/19 2:28 PM, Henk Gooijen via cctalk wrote:
As far as I know about the 11/93 (or 11/94), all memory is on the CPU module.
The 11/9[3|4] is sort-of an 11/8[3|4] with all (PMI) memory on the CPU module as the
cache on the 8[3|4].
The 11/9[3|4] with 2 MB RAM can only be upgraded to 4 MB by adding RAM chips on the CPU
module.
However, all holes are filled with solder, so this upgrade is quite a lot of work.
I was beginning to suspect as much as my attempts got nowhere.
My board has two empty spots with holes filled with solder.
That is not really a problem as I have a professional solder
sucker. :-) I assume these might have to do with addressing.
Then there are four that I suspect are the memory chips and
they are surface mount. Not so sure I want to delve into
that as my eyes aren't what they used to be. I guess the
first step would be to see if the chips are even available.
bill
11 Jan
11 Jan
9:28 a.m.
Well, it has been so long since I had to put together an
entire system I forgot what fun it can be.
With almost documentation I was able to configure 4 1meg
memory modules and I tested them all in my 11/23+ box.
KDJ11-B with 4 different (but similar) memory cards.
"MAP" option of the KDJ11 shows 4 meg (minus the I/O
Page) and all the right CSR's for the first four blocks.
Really wanted to make this a deskside so I moved the
cards to a MicroPDP box (Yes, I have a coule of the
deskside pedestals). Power on, Memory CSR error.
Move then back to the 11/23+ works fine. If I only
put the first three in the MicroPDP box it works.
But as soon as I try to put in a fourth module I
get the memory CSR error. Tried different cards
(although working in the 11/23+ box would make me
think it is not a problem with the card) no help.
Anybody care to take a stab at what might be causing
this problem?
bill
11:25 a.m.
Most Probable cause is interrupt grant is broken. For most microspheres backplanes the
first three slots are different than remaining.
Phoned this in!
On Jan 11, 2019, at 12:28 PM, Bill Gunshannon via cctalk <cctalk at classiccmp.org>
wrote:
Well, it has been so long since I had to put together an
entire system I forgot what fun it can be.
With almost documentation I was able to configure 4 1meg
memory modules and I tested them all in my 11/23+ box.
KDJ11-B with 4 different (but similar) memory cards.
"MAP" option of the KDJ11 shows 4 meg (minus the I/O
Page) and all the right CSR's for the first four blocks.
Really wanted to make this a deskside so I moved the
cards to a MicroPDP box (Yes, I have a coule of the
deskside pedestals). Power on, Memory CSR error.
Move then back to the 11/23+ works fine. If I only
put the first three in the MicroPDP box it works.
But as soon as I try to put in a fourth module I
get the memory CSR error. Tried different cards
(although working in the 11/23+ box would make me
think it is not a problem with the card) no help.
Anybody care to take a stab at what might be causing
this problem?
bill
11:55 a.m.
On 1/11/19 2:25 PM, Allison Parent via cctalk wrote:
Most Probable cause is interrupt grant is broken. For
most microspheres backplanes the first three slots are different than remaining.
Yes, that's true. But the problem doesn't occur until the 5th slot.
And a quick look at the C-D edges of the board shows that the BUS
Grant lines are jumpered. So the 2 A-B slots should pass the grants
just fine. I saw no mention about these not working in A-B/A-B
backplanes. The only mention the docs make at all is that the
only lines used on C-D slots is power.
bill
Phoned this in!
On Jan 11, 2019, at 12:28 PM, Bill Gunshannon via cctalk <cctalk at classiccmp.org>
wrote:
Well, it has been so long since I had to put together an
entire system I forgot what fun it can be.
With almost documentation I was able to configure 4 1meg
memory modules and I tested them all in my 11/23+ box.
KDJ11-B with 4 different (but similar) memory cards.
"MAP" option of the KDJ11 shows 4 meg (minus the I/O
Page) and all the right CSR's for the first four blocks.
Really wanted to make this a deskside so I moved the
cards to a MicroPDP box (Yes, I have a coule of the
deskside pedestals). Power on, Memory CSR error.
Move then back to the 11/23+ works fine. If I only
put the first three in the MicroPDP box it works.
But as soon as I try to put in a fourth module I
get the memory CSR error. Tried different cards
(although working in the 11/23+ box would make me
think it is not a problem with the card) no help.
Anybody care to take a stab at what might be causing
this problem?
bill
12:41 p.m.
On 01/11/2019 02:55 PM, Bill Gunshannon via cctalk wrote:
On 1/11/19 2:25 PM, Allison Parent via cctalk wrote:
That is generally true... however there were more than a few odd backplanes
including a few that were wired for Core and some design to suit a
different
memory and incompatible (smoke releasing!).? Others are unique as the RL11
two board set require a CD backplane slot pair (will not work if ABAB).
You need to look at the backplane and find out what one it is.
I have a few boxes I got without a backplane at all and as a result they
have
what I find was most useful and very non standard.
The CSR complaint is because if you talk to the board and don't get a
int vector
or can't set a working one (floating)? the CSR is broke.? This is a
behavior that
is much different from unibus.? It also happens if an expected CSR is
non standard
for nominal drivers.? An example is my 11/73 as it has 2 RQDX3 so one
has to be
at a new and unusual CSR same for the second DLV11J serial.? Some memory
(parity or ECC) also have CSRs and if they interupt and the drive is not
there
I've heard havoc ensues (of all my memory on hand none have that).
Irs a side effect of the Q bus starting in the 70s (LSI-11) and
continuing with
some evolution though the 90s (11/83 and 11/9x).
Allison
Most Probable cause is interrupt grant is broken.
For most microspheres backplanes the first three slots are different than remaining.
Yes, that's true. But the problem doesn't occur until the 5th slot.
And a quick look at the C-D edges of the board shows that the BUS
Grant lines are jumpered. So the 2 A-B slots should pass the grants
just fine. I saw no mention about these not working in A-B/A-B
backplanes. The only mention the docs make at all is that the
only lines used on C-D slots is power.
bill
12:54 p.m.
On 1/11/19 3:41 PM, allison via cctalk wrote:
On 01/11/2019 02:55 PM, Bill Gunshannon via cctalk
wrote:
Not only that, it will kill the RL11.
On 1/11/19 2:25 PM, Allison Parent via cctalk
wrote:
That is generally true... however there were more than a few odd backplanes
including a few that were wired for Core and some design to suit a
different
memory and incompatible (smoke releasing!).? Others are unique as the RL11
two board set require a CD backplane slot pair (will not work if ABAB). Most Probable cause is interrupt grant is broken.
For most microspheres backplanes the first three slots are different than remaining.
Yes, that's true. But the problem doesn't occur until the 5th slot.
And a quick look at the C-D edges of the board shows that the BUS
Grant lines are jumpered. So the 2 A-B slots should pass the grants
just fine. I saw no mention about these not working in A-B/A-B
backplanes. The only mention the docs make at all is that the
only lines used on C-D slots is power.
bill
You need to look at the backplane and find out what one it is.
I have a few boxes I got without a backplane at all and as a result they
have
what I find was most useful and very non standard.
The CSR complaint is because if you talk to the board and don't get a
int vector
or can't set a working one (floating)? the CSR is broke.? This is a
behavior that
is much different from unibus.? It also happens if an expected CSR is
non standard
for nominal drivers.? An example is my 11/73 as it has 2 RQDX3 so one
has to be
at a new and unusual CSR same for the second DLV11J serial.? Some memory
(parity or ECC) also have CSRs and if they interupt and the drive is not
there
I've heard havoc ensues (of all my memory on hand none have that).
Irs a side effect of the Q bus starting in the 70s (LSI-11) and
continuing with
some evolution though the 90s (11/83 and 11/9x).
At this point I am going with your comment about bus grant.
I don't know why but that seems to be the likely culprit.
More to follow if I get this working.
bill
12:34 p.m.
On 1/11/19 11:28 AM, Bill Gunshannon via cctalk wrote:
Well, it has been so long since I had to put together
an
entire system I forgot what fun it can be.
With almost documentation I was able to configure 4 1meg
memory modules and I tested them all in my 11/23+ box.
KDJ11-B with 4 different (but similar) memory cards.
"MAP" option of the KDJ11 shows 4 meg (minus the I/O
Page) and all the right CSR's for the first four blocks.
Really wanted to make this a deskside so I moved the
cards to a MicroPDP box (Yes, I have a coule of the
deskside pedestals). Power on, Memory CSR error.
Move then back to the 11/23+ works fine. If I only
put the first three in the MicroPDP box it works.
But as soon as I try to put in a fourth module I
get the memory CSR error. Tried different cards
(although working in the 11/23+ box would make me
think it is not a problem with the card) no help.
Anybody care to take a stab at what might be causing
this problem?
1) You might want to check your power supply loading.?? You have a
lot
of memory of chips to feed. ? The H7864 output at +5V is 36 amps.?
Between the KDJ11-B and 4 memory boards you may be using between 20-25
amps. Add a? disk controller, disk drive and any other Q-bus devices
might bring you close to the limits for either +5V or +12V from an aging
supply.
2) Are the CSR addresses between 17772100-06?? Have you tried to move
them above this range just to see if the behavior changes?
? Jerry
12:48 p.m.
On 1/11/19 3:34 PM, Jerry Weiss wrote:
On 1/11/19 11:28 AM, Bill Gunshannon via cctalk
wrote:
Nothing but CPU and memory at this point. And there will never be
drives as I will be using SCSI and 6 of them sit in a box of their
own just waiting for the PDP-11 to wake up and smell the coffee.
Well, it has been so long since I had to put
together an
entire system I forgot what fun it can be.
With almost documentation I was able to configure 4 1meg
memory modules and I tested them all in my 11/23+ box.
KDJ11-B with 4 different (but similar) memory cards.
"MAP" option of the KDJ11 shows 4 meg (minus the I/O
Page) and all the right CSR's for the first four blocks.
Really wanted to make this a deskside so I moved the
cards to a MicroPDP box (Yes, I have a coule of the
deskside pedestals). Power on, Memory CSR error.
Move then back to the 11/23+ works fine. If I only
put the first three in the MicroPDP box it works.
But as soon as I try to put in a fourth module I
get the memory CSR error. Tried different cards
(although working in the 11/23+ box would make me
think it is not a problem with the card) no help.
Anybody care to take a stab at what might be causing
this problem?
1) You might want to check your power supply loading.?? You have a
lot
of memory of chips to feed. ? The H7864 output at +5V is 36 amps.
Between the KDJ11-B and 4 memory boards you may be using between 20-25
amps. Add a? disk controller, disk drive and any other Q-bus devices
might bring you close to the limits for either +5V or +12V from an aging
supply.
2) Are the CSR addresses between 17772100-06?? Have you tried to move
them above this range just to see if the behavior changes?
CSR's are right. As I said, the combo worked until I moved it
to a different backplane. I think Allison has the idea. I am
going to go back and look into bus grants. But I would have
thought they would be alright as the likelihood of these being
in a serpentine A-B box seems much to likely for them to not
be capable of it.
bill
2217
days inactive
2219
days old
9 comments
5 participants
participants (5)
-
allisonportable＠gmail.com -
bill.gunshannon＠hotmail.com -
henk.gooijen＠hotmail.com -
jnc＠mercury.lcs.mit.edu -
jsw＠ieee.org