12 Feb
2016
12 Feb
'16
6:32 p.m.
Hi,
Seems I have bits 4 and 3 sticking on my Clearpoint QRAM-2-SAB-1 88b
4MB memory in my pdp11/73.
Can anyone offer hints as to how to identify which component is broken
and how to go about repairing this?
It's the only memory board in this machine, so I guess the problem
might actually be a bus or processor board, right? I have no other q-bus
memory to test with, so can't do swapping / process of elimination to be
sure.
Here's the manual:
http://www.arclightindustries.com/docs/Clearpoint-88B.pdf (which I probably
should add to manx or archive.org or something).
Here's a snippet of the VMJA diags run illustrating bits 4 and 3
sticking. During the next VMJA run, all addresses were showing up as
errored instead of just the ones ending in xxx000xx, so I guess it's
getting worse!
@173000g
Starting system
BOOTING UP XXDP-XM EXTENDED MONITOR
XXDP-XM EXTENDED MONITOR - XXDP V2.5
REVISION: F0
BOOTED FROM DL0
124KW OF MEMORY
NON-UNIBUS SYSTEM
RESTART ADDRESS: 152000
TYPE "H" FOR HELP !
.R VMJA??
VMJAB0.BIC
CVMJAB0 ECC/PARITY MEMORY DIAGNOSTIC
11/83 CACHE AVAILABLE
SWR = 000000 NEW = 000040
CSR MAP
CSR 0 1 2 3 4 5 6 7 8 9 A B C D E F
MEMTYPE P
CSR NUMBER 0 CONTROLS TOO MANY BANKS
2044K OF Q-BUS PARITY MEMORY
2044K WORDS OF MEMORY TOTAL
MEMORY CONFIGURATION MAP
16K WORD BANKS
1 2 3 4 5 6 7
012345670123456701234567012345670123456701234567012345670123
ERRORS
MEMTYPE PPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP
CSR 000000000000000000000000000000000000000000000000000000000000
PROTECT PP
1 1 1 1 1 1 1
0 1 2 3 4 5 6
456701234567012345670123456701234567012345670123456701234567
ERRORS
MEMTYPE PPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP
CSR 000000000000000000000000000000000000000000000000000000000000
PROTECT
1
7
01234567
ERRORS
MEMTYPE PPPPPPPP
CSR 00000000
PROTECT
MEMORY DATA ERROR
PC BANK VADD PADD GOOD BAD XOR CSR MTYP INT PAT
027606 10 060000 01000000 000010 000030 000020 0 P 27
027606 10 060002 01000002 000010 000030 000020 0 P 27
027606 10 060004 01000004 000010 000030 000020 0 P 27
027606 10 060006 01000006 000010 000030 000020 0 P 27
027606 10 060010 01000010 000010 000030 000020 0 P 27
027606 11 060000 01100000 000011 000031 000020 0 P 27
027606 11 060002 01100002 000011 000031 000020 0 P 27
027606 11 060004 01100004 000011 000031 000020 0 P 27
027606 11 060006 01100006 000011 000031 000020 0 P 27
027606 11 060010 01100010 000011 000031 000020 0 P 27
027606 12 060000 01200000 000012 000032 000020 0 P 27
027606 12 060002 01200002 000012 000032 000020 0 P 27
027606 12 060004 01200004 000012 000032 000020 0 P 27
027606 12 060006 01200006 000012 000032 000020 0 P 27
027606 12 060010 01200010 000012 000032 000020 0 P 27
027606 13 060000 01300000 000013 000033 000020 0 P 27
027606 13 060002 01300002 000013 000033 000020 0 P 27
027606 13 060004 01300004 000013 000033 000020 0 P 27
027606 13 060006 01300006 000013 000033 000020 0 P 27
027606 13 060010 01300010 000013 000033 000020 0 P 27
027606 14 060000 01400000 000014 000034 000020 0 P 27
027606 14 060002 01400002 000014 000034 000020 0 P 27
027606 14 060004 01400004 000014 000034 000020 0 P 27
027606 14 060006 01400006 000014 000034 000020 0 P 27
027606 14 060010 01400010 000014 000034 000020 0 P 27
027606 15 060000 01500000 000015 000035 000020 0 P 27
027606 15 060002 01500002 000015 000035 000020 0 P 27
027606 15 060004 01500004 000015 000035 000020 0 P 27
027606 15 060006 01500006 000015 000035 000020 0 P 27
027606 15 060010 01500010 000015 000035 000020 0 P 27
027606 16 060000 01600000 000016 000036 000020 0 P 27
027606 16 060002 01600002 000016 000036 000020 0 P 27
027606 16 060004 01600004 000016 000036 000020 0 P 27
027606 16 060006 01600006 000016 000036 000020 0 P 27
027606 16 060010 01600010 000016 000036 000020 0 P 27
027606 17 060000 01700000 000017 000037 000020 0 P 27
027606 17 060002 01700002 000017 000037 000020 0 P 27
027606 17 060004 01700004 000017 000037 000020 0 P 27
027606 17 060006 01700006 000017 000037 000020 0 P 27
027606 17 060010 01700010 000017 000037 000020 0 P 27
027606 20 060000 02000000 000020 000030 000010 0 P 27
027606 20 060002 02000002 000020 000030 000010 0 P 27
027606 20 060004 02000004 000020 000030 000010 0 P 27
027606 20 060006 02000006 000020 000030 000010 0 P 27
027606 20 060010 02000010 000020 000030 000010 0 P 27
027606 21 060000 02100000 000021 000031 000010 0 P 27
027606 21 060002 02100002 000021 000031 000010 0 P 27
027606 21 060004 02100004 000021 000031 000010 0 P 27
027606 21 060006 02100006 000021 000031 000010 0 P 27
027606 21 060010 02100010 000021 000031 000010 0 P 27
027606 22 060000 02200000 000022 000032 000010 0 P 27
027606 22 060002 02200002 000022 000032 000010 0 P 27
027606 22 060004 02200004 000022 000032 000010 0 P 27
027606 22 060006 02200006 000022 000032 000010 0 P 27
027606 22 060010 02200010 000022 000032 000010 0 P 27
027606 23 060000 02300000 000023 000033 000010 0 P 27
027606 23 060002 02300002 000023 000033 000010 0 P 27
027606 23 060004 02300004 000023 000033 000010 0 P 27
027606 23 060006 02300006 000023 000033 000010 0 P 27
027606 23 060010 02300010 000023 000033 000010 0 P 27
027606 24 060000 02400000 000024 000034 000010 0 P 27
027606 24 060002 02400002 000024 000034 000010 0 P 27
027606 24 060004 02400004 000024 000034 000010 0 P 27
027606 24 060006 02400006 000024 000034 000010 0 P 27
027606 24 060010 02400010 000024 000034 000010 0 P 27
027606 25 060000 02500000 000025 000035 000010 0 P 27
027606 25 060002 02500002 000025 000035 000010 0 P 27
027606 25 060004 02500004 000025 000035 000010 0 P 27
027606 25 060006 02500006 000025 000035 000010 0 P 27
027606 25 060010 02500010 000025 000035 000010 0 P 27
027606 26 060000 02600000 000026 000036 000010 0 P 27
027606 26 060002 02600002 000026 000036 000010 0 P 27
027606 26 060004 02600004 000026 000036 000010 0 P 27
027606 26 060006 02600006 000026 000036 000010 0 P 27
027606 26 060010 02600010 000026 000036 000010 0 P 27
027606 27 060000 02700000 000027 000037 000010 0 P 27
027606 27 060002 02700002 000027 000037 000010 0 P 27
027606 27 060004 02700004 000027 000037 000010 0 P 27
027606 27 060006 02700006 000027 000037 000010 0 P 27
027606 27 060010 02700010 000027 000037 000010 0 P 27
027606 30 060000 03000000 177747 177767 000020 0 P 27
027606 30 060002 03000002 177747 177767 000020 0 P 27
027606 30 060004 03000004 177747 177767 000020 0 P 27
027606 30 060006 03000006 177747 177767 000020 0 P 27
027606 30 060010 03000010 177747 177767 000020 0 P 27
027606 31 060000 03100000 177746 177766 000020 0 P 27
027606 31 060002 03100002 177746 177766 000020 0 P 27
027606 31 060004 03100004 177746 177766 000020 0 P 27
027606 31 060006 03100006 177746 177766 000020 0 P 27
027606 31 060010 03100010 177746 177766 000020 0 P 27
027606 32 06000
12 Feb
12 Feb
9:32 p.m.
On Fri, Feb 12, 2016 at 3:32 PM, Jacob Ritorto <jacob.ritorto at gmail.com> wrote:
Hi,
Seems I have bits 4 and 3 sticking on my Clearpoint QRAM-2-SAB-1 88b
4MB memory in my pdp11/73.
Can anyone offer hints as to how to identify which component is broken
and how to go about repairing this?
Are the 256Kx1 DRAM chips in sockets? If they are it shouldn't be too
hard to come up with a scheme for exchanging some of the DRAM chips to
see if the stuck bits move and then work out which chip positions map
to which bits and then which DRAM chips need to be replaced.
9:55 p.m.
On 2/12/2016 5:32 PM, Jacob Ritorto wrote:
Hi,
Seems I have bits 4 and 3 sticking on my Clearpoint QRAM-2-SAB-1 88b
4MB memory in my pdp11/73.
Can anyone offer hints as to how to identify which component is broken
and how to go about repairing this?
It's the only memory board in this machine, so I guess the problem
might actually be a bus or processor board, right? I have no other q-bus
memory to test with, so can't do swapping / process of elimination to be
sure.
Here's the manual:
http://www.arclightindustries.com/docs/Clearpoint-88B.pdf (which I probably
should add to manx or archive.org or something).
Here's a snippet of the VMJA diags run illustrating bits 4 and 3
sticking. During the next VMJA run, all addresses were showing up as
errored instead of just the ones ending in xxx000xx, so I guess it's
getting worse!
@173000g
Starting system
BOOTING UP XXDP-XM EXTENDED MONITOR
XXDP-XM EXTENDED MONITOR - XXDP V2.5
REVISION: F0
BOOTED FROM DL0
124KW OF MEMORY
NON-UNIBUS SYSTEM
RESTART ADDRESS: 152000
TYPE "H" FOR HELP !
.R VMJA??
VMJAB0.BIC
CVMJAB0 ECC/PARITY MEMORY DIAGNOSTIC
11/83 CACHE AVAILABLE
SWR = 000000 NEW = 000040
CSR MAP
CSR 0 1 2 3 4 5 6 7 8 9 A B C D E F
MEMTYPE P
CSR NUMBER 0 CONTROLS TOO MANY BANKS
2044K OF Q-BUS PARITY MEMORY
2044K WORDS OF MEMORY TOTAL
MEMORY CONFIGURATION MAP
16K WORD BANKS
1 2 3 4 5 6 7
012345670123456701234567012345670123456701234567012345670123
ERRORS
MEMTYPE PPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP
CSR 000000000000000000000000000000000000000000000000000000000000
PROTECT PP
1 1 1 1 1 1 1
0 1 2 3 4 5 6
456701234567012345670123456701234567012345670123456701234567
ERRORS
MEMTYPE PPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP
CSR 000000000000000000000000000000000000000000000000000000000000
PROTECT
1
7
01234567
ERRORS
MEMTYPE PPPPPPPP
CSR 00000000
PROTECT
MEMORY DATA ERROR
PC BANK VADD PADD GOOD BAD XOR CSR MTYP INT PAT
027606 10 060000 01000000 000010 000030 000020 0 P 27
027606 10 060002 01000002 000010 000030 000020 0 P 27
027606 10 060004 01000004 000010 000030 000020 0 P 27
027606 10 060006 01000006 000010 000030 000020 0 P 27
027606 10 060010 01000010 000010 000030 000020 0 P 27
<< SNIP >>
Well clearly it is only affecting certain address bits - or the
diagnostic would not run at all - note that it is starting at 010000000,
so that points to the memory, rather than the processor or bus, at
least as a first approximation. No guarantees, but I'd sure start with
that as a working theory.
Another sign: this is right at the boundary between two rows.
If you can't find a schematic, you can use the address to identify the
address lines on the bus (See Table 3, page 1-5), and trace them on the
board to find the relevant row of chips. Then use the bits the same way
to identify the specific chips.
If the chips are in sockets, you could always pull them one at a time to
find the relevant place in the array, as well.
I expect this is not the issue, but, one has to ask these things, just
in case. Did this memory work before? I ask because this is the 19th
bit. If the backplane is a Q18 backplane, that might be a problem?
Along those same lines, make sure that jumper P07 is positioned
correctly for a 22 bit address, and that P05 and P06 are set correctly
for the memory size.
Are you seeing the parity error light when this occurs?
Anyway, once the relevant chip(s) are identified, if they are in sockets
you can swap them with other bits or the same bits in other rows to
confirm. Otherwise you get to unsolder the suspects, and put in new ones.
JRJ
13 Feb
13 Feb
10:53 a.m.
On 02/12/2016 6:55 PM, Jay Jaeger wrote:
On 2/12/2016 5:32 PM, Jacob Ritorto wrote:
An old trick we use for testing soldered in DRAM is to simply jam a
known-to-be-good DRAM on top of the suspect one (legs bent in to make
good contact). DRAM normally fail bits high and so putting a good one on
top causes nothing different to happen if the suspect is good, but if
the suspect is bad then the top DRAM will drive the output and your RAM
test will pass.
Of course you wedge the good one on the suspect when the power is off.
Unless you are in a rush, and willing to possibly kill your test DRAM.
As a side note - there appears to be an error message: "CSR NUMBER 0
CONTROLS TOO MANY BANKS" Or is that irrelevant? I know nothing about
the PDP-11 test messages...
John :-#)#
Hi,
Seems I have bits 4 and 3 sticking on my Clearpoint QRAM-2-SAB-1 88b
4MB memory in my pdp11/73.
Can anyone offer hints as to how to identify which component is broken
and how to go about repairing this?
It's the only memory board in this machine, so I guess the problem
might actually be a bus or processor board, right? I have no other q-bus
memory to test with, so can't do swapping / process of elimination to be
sure.
Here's the manual:
http://www.arclightindustries.com/docs/Clearpoint-88B.pdf (which I probably
should add to manx or archive.org or something).
Here's a snippet of the VMJA diags run illustrating bits 4 and 3
sticking. During the next VMJA run, all addresses were showing up as
errored instead of just the ones ending in xxx000xx, so I guess it's
getting worse!
@173000g
Starting system
BOOTING UP XXDP-XM EXTENDED MONITOR
XXDP-XM EXTENDED MONITOR - XXDP V2.5
REVISION: F0
BOOTED FROM DL0
124KW OF MEMORY
NON-UNIBUS SYSTEM
RESTART ADDRESS: 152000
TYPE "H" FOR HELP !
.R VMJA??
VMJAB0.BIC
CVMJAB0 ECC/PARITY MEMORY DIAGNOSTIC
11/83 CACHE AVAILABLE
SWR = 000000 NEW = 000040
CSR MAP
CSR 0 1 2 3 4 5 6 7 8 9 A B C D E F
MEMTYPE P
CSR NUMBER 0 CONTROLS TOO MANY BANKS
2044K OF Q-BUS PARITY MEMORY
2044K WORDS OF MEMORY TOTAL
MEMORY CONFIGURATION MAP
16K WORD BANKS
1 2 3 4 5 6 7
012345670123456701234567012345670123456701234567012345670123
ERRORS
MEMTYPE PPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP
CSR 000000000000000000000000000000000000000000000000000000000000
PROTECT PP
1 1 1 1 1 1 1
0 1 2 3 4 5 6
456701234567012345670123456701234567012345670123456701234567
ERRORS
MEMTYPE PPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP
CSR 000000000000000000000000000000000000000000000000000000000000
PROTECT
1
7
01234567
ERRORS
MEMTYPE PPPPPPPP
CSR 00000000
PROTECT
MEMORY DATA ERROR
PC BANK VADD PADD GOOD BAD XOR CSR MTYP INT PAT
027606 10 060000 01000000 000010 000030 000020 0 P 27
027606 10 060002 01000002 000010 000030 000020 0 P 27
027606 10 060004 01000004 000010 000030 000020 0 P 27
027606 10 060006 01000006 000010 000030 000020 0 P 27
027606 10 060010 01000010 000010 000030 000020 0 P 27
<<
SNIP >>
Well clearly it is only affecting certain address bits - or the
diagnostic would not run at all - note that it is starting at 010000000,
so that points to the memory, rather than the processor or bus, at
least as a first approximation. No guarantees, but I'd sure start with
that as a working theory.
Another sign: this is right at the boundary between two rows.
If you can't find a schematic, you can use the address to identify the
address lines on the bus (See Table 3, page 1-5), and trace them on the
board to find the relevant row of chips. Then use the bits the same way
to identify the specific chips.
If the chips are in sockets, you could always pull them one at a time to
find the relevant place in the array, as well.
...
Are you seeing the parity error light when this occurs?
Anyway, once the relevant chip(s) are identified, if they are in sockets
you can swap them with other bits or the same bits in other rows to
confirm. Otherwise you get to unsolder the suspects, and put in new ones.
JRJ
3082
days inactive
3083
days old
3 comments
4 participants
participants (4)
-
cube1＠charter.net -
glen.slick＠gmail.com -
jacob.ritorto＠gmail.com -
jrr＠flippers.com