RL02 Spinup fails
aaron at aaronsplace.co.uk
Sat Nov 11 17:04:47 CST 2017
> Could elaborate a bit on this part of your earlier message.
>> I managed to boot into XXDP today using tu58em and ran the RLV12
>> controller test. It seems though that the version of XXDP I loaded does
>> not support the 22bit address (17774400) and defaults to the 16bit
>> address (174400) so it just shows errors for everything.
> VRLBCO expects 16 bit addresses for hardware setup. Most of the
> XXDP diagnostic programs understand the 16 bit address for IO page
> type devices.
> Separately, if this diagnostic itself is not able to use 22 bit memory
> on an 11/73 with 1MB ram, then you may have other issues. It
> should not be tossing errors. The fact that XXDP cannot boot
> into extended addressing (XXDPXM) is not a good sign either.
I can try - it's not throwing errors as such but the tests are
failing. For example, here is output for one of the tests.
] ROUTINE TRACE SEQ:
] INIT STATE TEST
] OPER: GET STAT
] RESULT: CNTLR HUNG
] BUS ADD=174400 DRV=0
] RLCS RLDA RLBA RLMP CYL HD
] OP INIT = 000104 000003 000000 00000
] OP DONE = 000000 000000 000000 00000 000 1
] CZRL1 HDR ERR 10002 ON UNIT 00 TST 009 SUB 000 PC: 020074
It's saying the controller has hung, but address 174400 is just RAM so I
don't expect it to get anything interesting back.
Maybe telling a bit about the system will help. Here are the cards in
the order that I have them. There were originally more but I have
removed them as they are not necessary for getting RL02 drives working.
- M8192 CPU (PDP-11/73)
- M7185 Multifunction (boot, serial, 128K RAM)
- M8067-LA (512K RAM)
- M8067-LA (512K RAM)
- M8061 (RLV12 controller)
The M7185 has its own RAM. I'm not sure where this appears on the
system, there may even be some kind of conflict with the proceeding RAM
cards? There is not much documentation online for this card. I have no
idea how to start the boot loader. A friend dumped the ROMs for me and I
compared them to someone elses dumps and they seemed okay. I've read
about people making a minor modification to this card to disable the
The RLV12 has its jumpers in the default configuration, so 22bit
addressing. I can query the controller's registers at
17774400..17774410. The CSR usually has something interesting in which I
can decode and it always makes sense, and matches the current drive
state, for example whether the pack is loaded or not on the least
significant bit. I'm not sure what is going on with the MPR. I perform
a get status as described in Appendix C3 of the RL02 technical
manual. If I am decoding the contents correctly it says things like the
cover is open and that the drive type is an RL01. For example, at one
point the contents was 167:
w h w s s w v d d h c h b s s s
d c l k p g c s t s o o h t t t
e e t e e e c b a
0 0 0 0 0 0 0 0 0 1 1 1 0 1 1 1
- ST A,B,C are status bits, being 1,1,1 means "Spin down"
- BH is brushes home, but I don't think this drive has brushes
- HO is heads out
- CO is cover open... the cover is closed unless it's referring to the
- HS is head select. 1 indicates the lower head
- DT is drive type, 0 indicates RL01, but it is an RL02.
- DSE is drive select error
- VC is volume check
- WGE is write gate error
- SPE is spin error, either too slow or too fast
- SKT is seek time out
- WL is write lock, the status of the write lock button on the front
- HCE is head current error
- WDE write data error
So there seems to be something wrong somewhere but I have no idea where.
More information about the cctalk