11/45 angst

Ok, got a bit of time yesterday and today to clean up the work area so I could work on the /45, and to actually go through some of the tests Tony had suggested. First, let me describe the current setup, as I changed things around a bit in order to simplify the setup. Here is how the machine is configured right now, and how it was configured during the following tests. I did not leave out any cards in my list here, so if you don't see it listed, there isn't anything in that slot. Current configuration System Unit 1 (A-B) M930C 2-5 empty 6 (A-F) M8100 7 (A-F) M8101 8 (A-F) M8102 9 (A-F) M8103 10 (A-F) M8104 11 (A-F) M8105 12 (A-F) M8106 13 empty 14 (A-F) M8116 15 (C-F) M8109 16-25 empty 26 (A-B) M9200 side 1 (C-F) M7856 {note, the LTC option on this board is enabled, there is nothing in slot 1C} 27 (A-B) M9200 side 2 (C-D) Dual Grant 28 (A-B) BC11A-0F side 1 (C-D) Dual Grant MF11-L in system unit 1 (A-B) BC11A-0F side 2 (C-F) H214 2 (A-F) G231 3 (A-F) G110 4 (A-F) G231 5 (A-F) G110 6 (C-F) H214 7-8 empty 9 (A-B) BC11A-15 side 1 BA11-KU with DD11-DK 1 (A-B) BC11A-15 (C-D) Dual Grant 2 (A-F) M7762 3-7 (C-D) Dual Grant 8 (D) single grant 9 (A-B) M9312 (D) single grant Connected to the M7762 is an RL02, and connected to the M7856 is an ADDS vt100 clone. Before I go further, I should point out that the system DOES have some basic functionality. It CAN boot up xxdp, and does pass a few front panel tests. The basic problem is I suspect that interrupts aren't working right. I suspect this because when trying to boot a known good RT11 pack, it hits the disk for at least 30 seconds before halting, and it's definitely progressing and not just retrying one read over and over. Many diagnostics fail, but they seem to fail whenever something with interrupts is being tested. Let me also say I'm new to the /45, but learning, so I'm not entirely 100% sure I'm operating the front panel right. In other words, it is very possible this is all due to some stupidity on my part! Initial setup - I power on the system with the front panel knobs set to cons phy and data paths. The console emulator "@" shows up on the system console. The status lights lit are Pause, Master, and Kernel. I type "DL" on the console, the lights start counting up steadily - because I forgot to hit load on the rl02. I hit load and when it goes ready the disk chugs for a minute then prints the XXDP loaded message on the console and prompts for date & time. I can load a few diags and it finds them at least. There is basic sanity here. I happened to find a program on the diag pack called "SIZER". Below is it's output from powerup, which may be helpful to someone "in the know". It reads... --------- p 031757 077317 030217 034317 @DL CHMDLE0 XXDP+ DL MONITOR BOOTED VIA UNIT 0 16K UNIBUS SYSTEM ENTER DATE (DD-MMM-YY): RESTART ADDR: 072010 THIS IS XXDP+. TYPE "H" OR "H/L" FOR HELP. .R SIZER SIZER .BIN PDP 11 SYSTEM SIZER V05.0 DIGITAL STANDARD CONFIGURATIONS REFER TO THE PERIPHERALS HANDBOOK (1975, 1976) CENTRAL PROCESSOR PDP 11/45,50,55 (KB11 SERIES) CPU INTERNAL OPTIONS STACK LIMIT REGISTER EXTENDED INSTRUCTION SET (ASH,ASHC,MUL,DIV) MEMORY SIZE AND ADDRESS MAP 16K WORDS OF CONTIGUOUS MEMORY , 000000, 20000, 40000, 60000, 100000, 120000, 140000, 160000 -------+-------+-------+-------+-------+-------+-------+-------+-------+ 000000 I 4K I 8K I 12K I 16K I *NXM* I *NXM* I *NXM* I *DEV* I -------+-------+-------+-------+-------+-------+-------+-------+-------+ ' 17776' 37776' 57776' 77776' 117776' 137776' 157776' 177776 PERIPHERAL DEVICES OPTION - ADDRESS - DESCRIPTION KY11 777570 PROGRAMMERS CONSOLE (SWITCH REGISTER) CONSOLE 777560 MAIN TERMINAL. (LA30, LA36, LT33, VT50, VT05, ETC) KW11-L 777546 LINE FREQUENCY CLOCK DP11 774400 SYNCHRONOUS INTERFACE BOOT 773000 READ ONLY MEMORY BOOTSTRAP BOOT 773100 READ ONLY MEMORY BOOTSTRAP BOOT 773200 READ ONLY MEMORY BOOTSTRAP BOOT 773300 READ ONLY MEMORY BOOTSTRAP BOOT 773400 READ ONLY MEMORY BOOTSTRAP BOOT 773500 READ ONLY MEMORY BOOTSTRAP BOOT 773600 READ ONLY MEMORY BOOTSTRAP BOOT 773700 READ ONLY MEMORY BOOTSTRAP M9301 765000 DIAGNOSTIC BOOT USER AND FLOATING PERIPHERAL DEVICES NONE PRESENT ----------------------------- At this point, nothing else is displayed. The above output makes perfect sense to me, with one exception. What on earth does it think when it shows a DP11 sync interface at 774400? Um, that address is what my RL11 is set to. Not sure why it shows as a DP11, maybe that's a bug in this sizer program. In any case... First, Tony suggested I store sucessive bits and read them back, ie. start at loc 0 and store 0,1,2,4,8 etc. I did that. All stored and read back correctly. Then Tony suggested that I run instruction diagnostics. I did that, and below is the output. ----------------------------- 022700 000012 001000 000020 @DL CHMDLE0 XXDP+ DL MONITOR BOOTED VIA UNIT 0 16K UNIBUS SYSTEM ENTER DATE (DD-MMM-YY): RESTART ADDR: 072010 THIS IS XXDP+. TYPE "H" OR "H/L" FOR HELP. .R CQKC?? CQKCG1.BIC DCQKC-G 11/40 & 11/45 INST EXER OPT.CP=045406 OPT.DEV.=000001 PASS #0000 VPC=072452 PSW=000344 PASS #0000 VPC=072452 PSW=010344 PASS #0000 VPC=072452 PSW=010344 PASS #0000 VPC=072452 PSW=010344 PASS #0000 VPC= ---------------------------- On the above diagnostic output, the system prints up to and including "OPT.DEV.=000001" very quickly, then runs for a long time (maybe 2 minutes or more), and then prints out the 5 lines of "PASS..." all very quickly. The cursor just sits to the right of VPC= on the last line, forever and the processor is halted. I'm betting money this is a bad result from this diagnostic. No clue what it means. But.... read on.... There is a website for "hints on testing a dead pdp11" that I used the line time clock test on. Here is what the website says to do.... ---------------------------- Line Time Clock Interrupt Test Most of the older PDP11's had a line time clock that would interrupt at mains frequency (50 or 60Hz). On early LSI-11's, it was not programmable but enabled by a switch on the front panel. Load the following program (which is just a trap catcher and interrupt enable), and start. If an early LSI-11's, substitute 000240 for the lines marked #, run the program for a second or two, then enable the clock. The program should halt at address 104 (the interrupt vector address is 100, and the vector is loaded with 102, which then executes a halt at 102, the PC will then be 104). Location Contents Opcode Comment 001000 012706 mov #770,sp set the stack pointer 001002 000770 001004 005000 clr r0 memory pointer 001006 012701 mov #2,r1 trap value 001010 000002 001012 010120 loop: mov r1,(r0)+ set trap vector value 001014 005020 clr (r0)+ set halt instruction in trap 001016 062701 add #4,r1 update r1 for next trap 001020 000004 001022 020600 cmp sp,r0 load up to stack pointer 001024 001372 bne loop # 001026 012737 mov #100,@#ltc enable clock interrupt # 001030 000100 # 001032 177546 001034 000777 br . loop for ever If the processor doesn't halt, then your clock isn't running. If it halts at 6, then the clock register is missing. -------------------------------- Now, I entered the above program. When I run it, it halts with PC = 4, NOT 104 and the data display shows 000770. Yes, I'm sure there are no burned out front panel lights. To my limited knowledge, this sounds like the program got the interrupt but instead of jumping to 100 for the vector, it jumped somewhere around 0 or 2, since the PC said 4. The plot thickens though, read on.... I took the above program and modified it to set bit six of the receive CSW to 1 (receive character interrupt enable) on the SLU, I believe I did it correctly by just changing location 1032 to 177560. When I run this program, the system is looping at 1034 just like it should... until I hit a key on the console. The moment I hit a key on the console, the processor halts with PC=4, and the data display shows 000770. So it seems that whether the interrupt comes from the LTC or the SLU, no matter what it halts at loc 4 with data 770. Now, I suspect I don't have firm enough ground to walk on to proceed with the following, but just in case it might mean something/confirm things to someone.... I decided to try to run the RL11 diagnostic. Below is the output from that.... ---------------------------------- 000000 003260 000000 000000 @DL CHMDLE0 XXDP+ DL MONITOR BOOTED VIA UNIT 0 16K UNIBUS SYSTEM ENTER DATE (DD-MMM-YY): RESTART ADDR: 072010 THIS IS XXDP+. TYPE "H" OR "H/L" FOR HELP. .R ZRLG?? ZRLGD0.BIC DRS-E0 CZRLG-D-0 CZRLG TESTS CONTROLLER FUNCTIONS, INTERFACE LOGIC, REGISTER OPERATION UNIT IS RL01,RL02 RSTRT ADR 067642 DR>START CHANGE HW (L) ? Y # UNITS (D) ? 1 UNIT 0 RL11=1, RLV11=2, RLV12=3 (O) 1 ? 1 BUS ADDRESS (O) 174400 ? VECTOR (O) 160 ? DRIVE (O) 0 ? DRIVE TYPE = RL01 (L) Y ? N BR LEVEL (O) 5 ? CHANGE SW (L) ? N NXT TST MAY ZERO LD UNIT. DOIT ANYWAY?Y ILL INTER 000 PC 013634 PS 000004 ---------------------------------- If I'm understanding what I'm seeing, it seems the processor is ignoring all vectors as Tony suggested. Of course, I am not skilled enough to know where to go from here. I can say that my processor cards (M8101 through M8106) were borrowed from another collector and tested before I got them, but I guess we can't rule out a problem with those. All the other cards, including terminators and grants, are mine. The BC11A-15 was replaced recently with a brand new one. I've rung out the bus grant lines from slot 26 in the system unit to the terminator card in the BA11 and those seem ok. I've checked the NPR jumper is everywhere, or a grant card is present - in all unibus slots, with the exception of slot 2 in the BA11 where the RL11 is. I've tried two or three different M7762's. Hopefully all this paints a clear picture of the problem, and hopefully someone can suggest a course of action that a DEC neophyte like myself can take :) Thanks for any input! Jay West