At 01:27 AM 2/9/2013, Charles wrote:
I tried this program tonight, after making up a new
ribbon cable that
had some mechanical damage which I was *hoping* would fix things :P
A very interesting result - when I hit run, BOTH drives started
seeking and eventually Drive 1 faults. If I unload one of the drives,
the other one seeks when the test program is run. Doesn't matter if
unit 0 or 1 is the active drive!
Time to start chasing a hardware (drive select) problem I guess...
sigh. Nothing like 30-40 year old hardware...
That's what I was worried about. You're lucky that it hasn't scrambled
the OS on your drives. :)
-Charles
ps I think your "assembly listing" has a bug, at 0221 shouldn't that
read JMP 0204?
Yup. The original for this (from the RL01/02 pocket service guide)
didn't have anything to do a drive select so I added it in. That
shifted everything, leading to all the offsets shifting. I missed
fixing the comment there, but jmp to 0204 is correct.
-Rick
I'm pulling my hair out... I disconnected the 2nd drive and now just
have one RL02 attached to my RL8A, with a terminator.
A bit of background - I could not find a cable that goes from the 40
pin header on the controller card to the first drive, so I made one
out of ribbon cable and IDC headers, removing the first external
connector and just using headers as it does inside the drive. Now I'm
beginning to wonder if the original cable has a "twist" somewhere in
the drive selects like the old PC floppy cables, because...
Although I can set and reset the two drive select bits on the RL8A,
and the 75113 line drivers also change state appropriately, those
signals are not reaching the correct place on the logic board inside
the RL02! At the 75107 line receivers (E57), those two bits are always
logic 00, i.e. drive 0. So the drive won't select at all unless I put
the Unit 0 plug in, at which point it effectively ignores the drive
select bits and always selects regardless of the state of those two
bits at the controller. Confirming that the drive select lines are dead.
The only thing between the line drivers and receivers is 6' of ribbon
cable. It leaves the RL8A on J1 header pins TT, SS, RR, PP and is
supposed to enter the RL02 logic board on J12 header pins C, D, E, F
which accounts for the cable being reversed. Just in case I even tried
flipping the header connector around and of course the drive won't
load or work at all and the Fault light stays on, so I know I have it
facing the right way. Unless it's possible to make a cable upside down
and backwards, or some such?
This setup DID previously run OS/8, ADVENT, FOCAL, whatever was on the
pack, but only using one of the RL02's (I don't think I was ever able
to access the other drive).
I can't think of anything that could "lose" the drive select lines
except a defective cable (will continuity-check tomorrow, even though
I made a new one by copying the old one that used to work) or else the
correct cable (BC80J-20) is not a straight-through pin for pin
connection, in which case I have to have one.
Just to further confuse things, I have the identical two-drive setup
with homemade controller-to-drive cable on my 11/23+ and RLV11 card,
and it works perfectly. So now I'm just going in circles... that's
enough for tonight!
EDIT: Additional note: When the two-drive system was running OS/8, I
had the ribbon cable going to the bottom drive in the rack which had
Unit 1 plug, then the proper BC20J cable from there to the top drive
with Unit 0 plug. Makes me wonder anew if there is something non-
obvious about the connections of the drive select lines within the DEC
cables...