PDP-11/23+ M8189: finally some result

[fjkraan]

After a very long time trying, I finally managed to get my PDP-11/23+ (M8189 plus RAM) working.
With working I mean entering programs with ODT and having them work as expected. Usually they just stopped somewhere.
The cause appears to be configuring J11 to be grounded and always applying 50Hz to BEVNT. With either removed, programs can now run for eternity.

But interrupts are useful or even required for some OS's, how to get it working? Other than adding a switch between the 50Hz and BEVNT...

BTW, I can now boot "RT-11SJ V05.03" from tu58em, but not "XXDP-XM" (first toggle the bootstrap).

Greetings,

Fred Jan

 

[Radix]

Hi Fred, the 11/23+ has onboard bootstrap that supports the TU58, you should be able to enter "DD" or "DD0" at the boot prompt, to save some typing..

RT11SJ only needs 64K (OK 56K) RAM to start, but XXDP-XM might need more RAM, but can't remember specifically how much it needs offhand, sorry...

What memory and other boards do you have? What backplane are you using?

Robin

 

[AK6DN]

After a very long time trying, I finally managed to get my PDP-11/23+ (M8189 plus RAM) working.
With working I mean entering programs with ODT and having them work as expected. Usually they just stopped somewhere.
The cause appears to be configuring J11 to be grounded and always applying 50Hz to BEVNT. With either removed, programs can now run for eternity.

But interrupts are useful or even required for some OS's, how to get it working? Other than adding a switch between the 50Hz and BEVNT...

BTW, I can now boot "RT-11SJ V05.03" from tu58em, but not "XXDP-XM" (first toggle the bootstrap).

Greetings,

Fred Jan

Have not heard of XXDP-XM. There are TU58 bootable XXDPv2 images here with 11/23 diagnostics: https://ak6dn.github.io/PDP-11/TU58/
Try 1123_1.DSK and 1123_2.DSK. Or 11XXDP.DSK which is an XXDP only disk with utilities.

 

[fjkraan]

Hi all, thanks for the suggestions.

@Radix: The onboard bootstrap doesn't work (yet). The 23-380E4/23-381E4 are there and visible, but don't work. RAM shouldn't be the problem as I have tested it with both an M8067 MSV11-PL and an M7551 MSV11-QB. The MMU might still be the problem above 64k.

@AK6DN: XXDP-XM just happened to be the name of the .TAP file. There is no other I/O other than the console (SLU1) at 0o77654 and SLU2 at 0o776500. Boot/Diag ROMs are at 0o773000 but not working.

The system as it came is very a-typical, just CPU, RAM, two DEQNA M7504 and a "LSI-11 SYSTEM ID REGISTER BOARD". Probably the system was intended for net-booting (complete with custom boot-ROMs). A special board, maybe a "DRV-11-WA"? is missing.

My plan was to just learn the PDP-11 instruction set, not to create a standard PDP-11/23Plus booting all the standard media. Some way of native storage might be convenient. I am considering buying a Qbone, but creating OFT-feed script would do too. Will try that first, it is more work, but cheaper.

A description of the system (prepared that for the 'problem-post to the forum' case I couldn't get it working): https://electrickery.nl/comp/pdp11_23plus/SystemConfig.txt

Greetings,
Fred Jan

Addendum: The "XXDP-XM.tap" is probably a faulty copy of "XXDPD2D.TAP". Other copies report "BOOTING UP XXDP-XM EXTENDED MONITOR" before dropping to the ODT-prompt (000014).

 

[fjkraan]

Having fixed the first memory error on the MSV-11P, the startup of the Boot/Diag ROM (KDF11B-BJ ROM V1.1) reports no error and says:

512KB MEMORY KDF11B-BJ ROM V1.1 CLOCK DISABLED

The diagnostics still report errors:

TESTING MEMORY

ERR 3 MEMORY FAULT
ADDRESS EXPECTED ACTUAL
01000000 000401 000000 P
01000002 000401 000000 P
01000004 000401 000000 P

01000000 octal is 40000 in hex, which is 256 k, but are these bytes or words?

DEC usually reports in bytes, but this seem to be words. Is this an elaborate way of saying there is no memory beyond 07777776?

At least XXDP actually boots now. Managed to run CVMJAB0 (from 1123_1.DSK):

CVMJAB0 ECC/PARITY MEMORY DIAGNOSTIC
NO CACHE AVAILABLE
SWR = 000000 NEW =


CSR MAP

CSR 0 1 2 3 4 5 6 7 8 9 A B C D E F
MEMTYPE P


256K OF Q-BUS PARITY MEMORY
256K WORDS OF MEMORY TOTAL

MEMORY CONFIGURATION MAP
16K WORD BANKS
1 2 3 4 5 6 7
012345670123456701234567012345670123456701234567012345670123
ERRORS XXXX
MEMTYPE PPPPPPPPPPPPPPPP
CSR 0000000000000000
PROTECT PP
MEMORY DATA ERROR
PC BANK VADD PADD GOOD BAD XOR CSR MTYP INT PAT
025150 10 002414 01122414 000000 060002 060002 0 P 06
025150 10 002414 01122414 000000 060002 060002 0 P 06
025150 10 002414 01122414 000000 060002 060002 0 P 06
....
025234 10 002412 01122412 000000 060000 060000 0 P 06
025250 10 002414 01122414 000000 060002 060002 0 P 06
MEMORY DATA ERROR
PC BANK VADD PADD GOOD BAD XOR CSR MTYP INT PAT
025770 10 157776 01077776 000377 01077776 01077776 0 P 17
025770 10 157774 01077774 000377 01077774 01077774 0 P 17
025770 10 157772 01077772 000377 01077772 01077772 0 P 17
....
025770 10 060002 01000002 000377 01000002 01000002 0 P 17
025770 10 060000 01000000 000377 01000000 01000000 0 P 17
025770 10 157776 01077776 177400 01077776 01077776 0 P 17
025770 10 157774 01077774 177400 01077774 01077774 0 P 17
025770 10 157772 01077772 177400 01077772 01077772 0 P 17

Is this an elaborate way to report the parity bit is broken?

Fred Jan

 

[daver2]

I had a similar issue with a faulty address line from the CPU.

I installed a much smaller (memory size wise) memory card and realised that the memory was 'shadowing' with the same CSR address at two locations within the memory map. I tracked the problem down to a broken address line on the CPU card to the backplane.

Dave

 

[fjkraan]

Hi Dave,

Thanks for your response. Could you boot RT-11 of XXDP in this fault situation?

I ran the Boot/Diag ROM map routine, and could account for all the used addresses, based on User Guide information..

MEMORY
Starting Ending
Address Address
00000000 01777776

I/O
Starting Ending
Address Address
17772100 17772100 MSV-11P CSR
17772300 17772316 PDR No. 3 ?
17772340 17772356 PAR No. 3 ?
17772516 17772516 Status Register 3
17773000 17773776 Boot/Diag ROM page
17776500 17776506 SLU2 (TU58 port)
17777520 17777524 Boot/Diag ROM CSR
17777546 17777546 Line Clock Status Register

Greetings,
Fred Jan

 

[daver2]

I had another fault (isn't it always the case) whereby the CPU hung when I tried to perform a boot from any disk controller type! I never got to the bottom of that... I swapped the CPU for another one... This was after I fixed my address line fault and it was correctly (and repeatedly) running the self diagnostics fine.

The CPU correctly sized my 4MB memory card - it is just when it came to testing. The results from part of the test were being written elsewhere - so the diagnostics were (correctly) reporting an error...

I just couldn't see the shadowing with the 4MB memory card installed.

Dave