KDJ11-B PDP 11/73 getting stuck in Exit standalone mode diag #56

[Mark G Thomas]

Hi Bill,

> ________________________________________
> From: cctalk [cctalk-bounces at classiccmp.org] on behalf of Mark G Thomas via cctalk [cctalk at classiccmp.org]
> Subject: KDJ11-B PDP 11/73 getting stuck in Exit standalone mode diag #56
> 
> Hi,
> 
> I picked up an empty 4x8 Q-bus chassis at VCF-East this year. It's
> labelled "USDC CSS-823 Processor System" on the outside. It's not a DEC
> backplane, but the chassis is light aluminum, smaller and in some ways
> nicer than my pile of BA23s. It includes a switching power supply,
> LTC and q-bus termination, and I confirmed it's wired for Q22. I was
> hoping to transplant an 11/83 into it, since it will fit more comfortably
> in my work area. It has lit DCON/RUN/LTC/HALT buttons on the front pane.
> The etching on the backplane says something like "MDB Systems, Inc.
> East Orange, CA 92665"
> 
> Unfortunately, I discovered slots are all wired Q22/Q22, no top ABCD
> (Q22/CD) PMI slots like a BA23. So, not good for an 11/83 CPU wired
> for PMI on slots C-D.
> 
> So, I was hoping I could get my KDJ11-B (quad-wide 11/73) running in it.
> Upon power up or reset I get "Testing in progress. Please wait" on
> the serial console, and the LEDs get stuck on test 56. CTRL-C doesn't do
> anything yet, and my only options seem to be to hit reset, or if I hit the
> run/stop toggle, I can reset to the ODT prompt. This is with either just
> the KDJ11-B in the top slot, or with the addition of a suitably configured
> MSV11-QA in the next slot. If I turn off the LTC on the front panel,
> I get the expected Error 61 M8190 clock error, PC and register values,
> and options to re-run once or loop on test, so I know the LTC is working.
> 
> I confirmed the KDJ11-B works fine in a BA23, getting past test 56. I also
> tried an 11/53 CPU here in the ABAB backplane and that gets past the self
> tests into dialog mode, but I really would like to get the KDJ11-B working
> in this chassis. I do see that a couple traces to the KDJ11-B C-D slots
> besides the Grant Continuity connections. I assume since the KDJ11-B
> has those Grant Continuity connections on slots C-D, it can run in a
> qbus/qbus Q22/Q22 quad Q-bus slot.
> 
> I noticed the FP "RUN" lamp never comes on (and is not burned out), but maybe
> that's normal -- monitoring something on the bus that doesn't happen during
> power on self test.
> 
> Does anyone here have suggestions about next steps in diagnosing this?
> 
> https://archive.org/stream/bitsavers_decpdp1111aintNov86_5833755/EK-1184A-MG-001_1184maint_Nov86_djvu.txt
> -----------------------------------------------------------------------------
> 4.10.18 Exit Standalone Mode (Test 56)
> 
> In the 22-bit mode, the exit standalone mode is checked by using the
> guaranteed timeout address of 17 760 000 to verify that the timeout
> logic works without hanging up the CPU.
> -----------------------------------------------------------------------------

On Wed, Apr 05, 2017 at 03:02:55PM +0000, Bill Gunshannon via cctalk wrote:
> OK, I have the book on that.  You need to determine which backplane you have.
> MLSI-BPA84-T    Serpentine A-B-A-B
> MLSI-BPA84-A-T   Serpentine except for slots 6 and 7.
> 
> We'll ignore the MLSI-BPA82 as that's just a dual wide A-B.
> 
> I will see if I can scan the book and send it to you but no promises how
> soon.  It's pretty good sized and my scanning capabilities are primitive.
> 
> bill

I didn't realize the M8190 (11/73) was PMI, and really the same board
as my 11/83. Fortunately, there was definitely no smoke, and the 11/73 
still works fine in my BA23. I think the magic smoke problem might be
more likely if someone plugs a Q22 board into PMI slots, due to the 
unfortunate wiring of those grant continuity connections.

I cannot find anything like BPA84-anything on this backplane.
There is a tiny whale emblem on the trace side of the board saying 
something like "MO__ Quick" under it, though faint, and bold easy 
to read "MLSI-CHASSIS #40040328" close to the end with the power terminals.

I confirmed with a meter that all slots are A-B-A-B, including 6 and 7.

I'm not sure I need the manual, unless it also includes schematics for
the power supply, RTC, and front panel control electronics wiring.

I'm now up and running with a dual-wide M8192 from a batch from a
scrap board guy on ebay, and an M7195 borrowed from another system
for testing.

Next challenge is figuring out the jumpers on an M8043 for SLU console
and seeing if I have or can find a working SCSI controller with a boot 
ROM on it, since I don't have another M7195 and they seem to go for
high prices these days on e-bay when they do show up.

Mark

-- 
Mark G. Thomas (Mark at Misty.com), KC3DRE