After the better part of 2 weeks I've finally gotten one of my
PDP-11/40's to execute some code. Both initially exhibited frustrating
similar problems.
Both were stuck in "reset". This was because the processor was seeing
AC-LO and DC-LO asserted. While the symptom was the same, the cause was
different. In one case the power supply was not deasserting the
signals. In the other, something in the RK-11D was pulling the signals
low. I've worked around both but fixed neither.
Once that problem was solved on the "first" 11/40, I discovered that
there was a stuck bit when trying to read or write memory. This was
temporarily fixed by swapping the data-paths module. At this point I
could read and write to memory.
Memory could be read and written through the front panel, but execution
would immediately trap to location 4 (6 showing on the front panel).
This was a "bus error". I unfortunately spent a lot of time swapping
memory around (this was not the problem). Replacing the data-paths
module with another suitably jumpered one solved the problem and the
11/40 happily runs a little test program. However, at this point I'm
hesitant to run it for very long since there is an acrid smell emitted
while the system is powered on. I don't want to burn anything up until
I track it down.
The second 11/40 still traps to location 4. I've verified that the
memory is good by swapping memory boards and backplane (I'm using
MF11-UP memory) with the first 11/40. I checked the supplies and they
seem a little low, but it got late tonight and I'll check and adjust
them tomorrow to see if that fixes the problem.
It's a bit hard to determine if any one card is at fault because the two
11/40 CPUs are configured very differently (which means there are almost
2 dozen jumpers between all the boards that need to be changed). Here's
the 2 different CPU configurations:
CPU #1 CPU #2
KD11-A KD11-A Basic CPU
KE11-E KE11-E EIS
KE11-F FIS
KJ11-A Stack limit option
KT11-D Memory management
I have another set of CPU boards (but no extra FIS) that if the voltage
checks don't solve the problem I'll try to see if that fixes it.
If anyone has any other suggestions, I'd appreciate hearing from you.
--
TTFN - Guy
Received on Fri Jan 09 2004 - 01:29:54 GMT