18 Jan
2002
18 Jan
'02
6:27 p.m.
On Jan 17, 0:36, Tom Leffingwell wrote:
I finally managed to obtain an ADV11-C without parting with a
large sum of money, but unfortunately, it didn't solve my problem. The
system still hangs at the same point it did when I used a DZ to "fake"
the
ADV11 by setting its CSR to 770400. If I remove the
ADV11, the program
runs, but it is forced into its simulate mode because the missing
hardware. I set the vector to the default on the DRV11, DRV11-B, and
ADV11-C. I didn't check the serial module though. Could this be a
vector
setting problem, or do I have the wrong symptoms? Is
there anything I
can
learn by halting the system when its hung and looking
at addresses in
memory?
Probably, but I'm not sure what :-) You could try halting it and see if i
always halts in the same place(s). If so, it would be worth looking to see
if you can figure out what instructions it executing at that point. It
might be a WAIT or a polling loop. Also look at the vectors in low memory
(from locations 0 to about 20), and the vectors for the devices you have in
the system. They should point to somewhere in the program.
I did try doing a "770400/" and it returned
000000, with the
module out it gives a ?, so that seems good. 770402, the data buffer
returns 007777.
Sounds like you have it set up correctly, at least as far as addressing is
concerned.
Does anyone know what to try next? I'm running
out of ideas.
The only difference between my system and the one I'm trying to replace
is
that the backplane is 4 slot instead of 9, I don't
have a BDV11-A like
the
original had, and I have a DSD systems generic RX02,
instead of the real
DEC M8029. Also, do the limit of backplane space, my modules are in a
slightly different order.
The order of the modules shouldn't matter, though it's best to have
CPU-memory-serial followed by whatever device needs highest priority
response. I expect the DSD RX02 controller works just like a DEC one as
far as addresses and interrupts are concerned, so that's unlikely to
matter. The BDV11 is another story, though. It includes a line-time
clock, and your system may need that. IIRC, it's controlled by a
combination of a switch on the BDV11, a bit in a register on the BDV11, and
the LTC switch on the front panel. All of those have to be enabled, and
the BDV11 must be present, for the clock to run. I have seen systems that
boot but then hang if the software needs a clock but it's not present or is
disabled.
--
Pete Peter Turnbull
Network Manager
University of York
20 Jan
20 Jan
1:52 p.m.
Thanks. I tried all that, and then I moved the system home. (It was
stored about 20 miles away before, and I was getting tired of driving back
and forth, but that finally overcame my laziness of loading it all into my
truck). Anyway, when I got home I noticed it got just slightly further
than before. I took out all the modules and checked for the fourth time
that they were set correctly. I found that one of the zero ohm
resistor-style jumpers on the DRV11 had been cut even with the board, so
it looked very much like it was still connected, and I managed to miss
this before. I soldered a jumper around it, but still the same problem.
Remembering that it got further than it did before, I decided to hook back
up the external hardware that talks to the DRV11 and DRV11-B. (I hadn't
attached it since I got home). Sure enough, it worked. I assume the
vector address on the DRV11 had something to do with it, but why it got
that little bit futher with nothing but hauling the equipment around still
bothers me. Anyway, hopefully it will stay that way.
Thanks very much again for all your help, as well as everyone elses.
-Tom
On Sat, 19 Jan 2002, Pete Turnbull wrote:
Probably, but I'm not sure what :-) You could
try halting it and see if i
always halts in the same place(s). If so, it would be worth looking to see
if you can figure out what instructions it executing at that point. It
might be a WAIT or a polling loop. Also look at the vectors in low memory
(from locations 0 to about 20), and the vectors for the devices you have in
the system. They should point to somewhere in the program.
Sounds like you have it set up correctly, at least as far as addressing is
concerned.
The order of the modules shouldn't matter, though it's best to have
CPU-memory-serial followed by whatever device needs highest priority
response. I expect the DSD RX02 controller works just like a DEC one as
far as addresses and interrupts are concerned, so that's unlikely to
matter. The BDV11 is another story, though. It includes a line-time
clock, and your system may need that. IIRC, it's controlled by a
combination of a switch on the BDV11, a bit in a register on the BDV11, and
the LTC switch on the front panel. All of those have to be enabled, and
the BDV11 must be present, for the clock to run. I have seen systems that
boot but then hang if the software needs a clock but it's not present or is
disabled.
--
Pete Peter Turnbull
Network Manager
University of York
8386
days inactive
8388
days old
1 comments
2 participants
participants (2)
-
pete＠dunnington.u-net.com -
tom＠sba.miami.edu