16 Aug
2007
16 Aug
'07
5:47 p.m.
Am I right in thinking that code 010 on the LED display is "hung at disk
boot"? If so, is there any way of getting a bit more detail as to the actual
error at this stage?
This machine has a 5.25" ST-506/412 drive in it. I don't think it's the one
with the dreaded sticking rubber problem; at any rate I've had it spinning up
on the bench - there seems to be a healthy clunk as heads unpark, and it
doesn't spin down, suggesting that side of things is healthy.
Other question: at the point when it's doing whatever checks yield the 010
code, should the display have initialised and be displaying anything? I'm
getting nothing out of this system - but it's entirely possible that the
display init is further down the boot process, and so it's never getting that
far... (Anyone know which manual contains the pinouts for the video output and
landscape display?)
ta
Jules
17 Aug
17 Aug
3:10 p.m.
Am I right in thinking that code 010 on the LED display is "hung at disk
boot"? If so, is there any way of getting a bit more detail as to the actual
error at this stage?
IIRC, 010 means the microcode ROM self-tests have completed, and the
thing is waiting for a boot device.
Now, any PERQ will try to boot from the PERQlink (if you have an OIO card
installed), then the floppy drive, then the hard drive. But certainly a
T2 (or T4) will only boot from the floppy drive if the hard drive has
gone ready.
This machine has a 5.25" ST-506/412 drive in it. I don't think it's the one
What drive is it, and has the PERQ ever worked with this drive? Have you
set the drive select jumpers correctly?
with the dreaded sticking rubber problem; at any rate
I've had it spinning up
on the bench - there seems to be a healthy clunk as heads unpark, and it
doesn't spin down, suggesting that side of things is healthy.
Other question: at the point when it's doing whatever checks yield the 010
code, should the display have initialised and be displaying anything? I'm
No. The video system is quite complciated to set up (rememebr there's no
hardware text mode), it's not something that can be done by the 256-word
microdoe ROM. The first screen output (assuming a normal OS) is the
obvious patterns from the memory test routints in VFY (microcode loaded from
the boot drive),
getting nothing out of this system - but it's
entirely possible that the
display init is further down the boot process, and so it's never getting that
far... (Anyone know which manual contains the pinouts for the video output and
landscape display?)
It's possible to decuce them from the motherboard ('MOM') schematics in
the schematics book. The pinout is given in the hardware FAQ too.
What I would do is first check that the hard disk is going ready -- that
is with a logic probe on the control connector with the drive cabled up
to the PERQ.
If it was my machine (as I have the facilities), I'd trace the microcode.
The low 12 microcode address lines (all you need for this) are brought
out the bottom connector (26 pin) on the front of the CPU board, along
wit hte microcode clock. A logic analyser connected there can trace the
boot microcode. Then look at the source listings to see just what the CPU
is doing.
If you don't have the boot ROM sources, ask me nicely for them :-)
-tony
19 Aug
19 Aug
3:07 p.m.
Tony Duell wrote:
What I would do is first check that the hard disk is
going ready -- that
is with a logic probe on the control connector with the drive cabled up
to the PERQ.
Presumably I can do that on the bench - I think it's just a case of applying
power and twiddling the appropriate drive select line so that the drive
enables the output control signals... (I'll check via bitsavers - I think the
drive selects are active-low)
Worst-case if the drive's hosed, we've got various install disks - just a
question of whether there's a formatter (low or high level) in there somewhere!
cheers
Jules
20 Aug
20 Aug
2:37 p.m.
Tony Duell wrote:
You can, yes. Althought personally, I'd do it in the PERQ (a 50 way cabel
with an edge connector on one end and a 2-row header socket on the other
is very useful for cabling the DIB to the EIO in this sort of situation
;-)). Reason being that if the PERQ isn't asserting drive select or
something, you'll spot that oo.
What I would do is first check that the hard disk
is going ready -- that
is with a logic probe on the control connector with the drive cabled up
to the PERQ.
Presumably I can do that on the bench - I think it's just a case of applying power and twiddling the appropriate drive select line
so that the drive
enables the output control signals... (I'll check via bitsavers - I think the
drive selects are active-low)
YEs, just about all the signals on the ST412 control connector (and for
that matter a floppy drive connector) are active low. The drive outputs
may well be (should be!) open-collector signals, requring an external
pull-up resisotr.
A couple more things. The 3 wire power connector on the DIB is +5V,
Ground, -5V. Not +12V anywhere. And the DIB breaks the rules by using
totem-pole drivers to drive the disk signals. if the power cable falls
off, these darn things can assert the appropraite signals and corrupt the
disk. Don't ask how I found that out!
Worst-case if the drive's hosed, we've got
various install disks - just a
question of whether there's a formatter (low or high level) in there somewhere!
The formatter wasn't included with any normal PERQ OS distribution. But
you should have the engieer's disk somewhere that includes it.
-tony
6334
days inactive
6337
days old
3 comments
2 participants
participants (2)
-
ard＠p850ug1.demon.co.uk -
julesrichardsonuk＠yahoo.co.uk