25 Jan
2018
25 Jan
'18
3:06 p.m.
Hi all,
I mentioned this in the thread where I'd asked about basic Microvax II
info, but it may have got lost in traffic...
The machine's H7260 PSU is somewhat unwell - one of the internal +5/12V
supplies appears to be healthy, but the other has outputs which are sitting
at around 2.5V (both on the 5V and 12V rails) under a test load.
Initial questions...
1) Are schematics are available online? I couldn't find them (either under
the DEC p/n or the Astec AA13010 one), but perhaps they're buried in
schematics for a specific DEC machine somewhere.
2) Upon initial glance, the 'first' board of the three in the PSU module
appears to be a pair of control relays, bridge rectifiers and capacitors,
supplying +/- DC voltages to the two individual PSU boards. Does anyone
know if I can disconnect these from the PSU boards safely* and measure
their outputs, and if so what voltages I should expect to see? That might
be a good initial test before pointing a finger at the PSU board associated
with the low outputs.
* I mean without component damage - I expect they might be sitting at a
significant DC voltage, so there's an obvious personal safety aspect too ;-)
cheers
Jules
25 Jan
25 Jan
3:47 p.m.
-----Original Message-----
From: cctalk [mailto:cctalk-bounces at classiccmp.org] On Behalf Of Jules
Richardson via cctalk
Sent: 25 January 2018 23:07
To: xx Classiccmp mailing list <cctalk at classiccmp.org>
Subject: DEC H7260 PSU fault
Hi all,
I mentioned this in the thread where I'd asked about basic Microvax II
info,
but
it may have got lost in traffic...
The machine's H7260 PSU is somewhat unwell - one of the internal +5/12V
supplies appears to be healthy, but the other has outputs which are
sitting at
around 2.5V (both on the 5V and 12V rails) under a
test load.
Initial questions...
1) Are schematics are available online? I couldn't find them (either under
the
DEC p/n or the Astec AA13010 one), but perhaps
they're buried in
schematics
for a specific DEC machine somewhere.
I doubt it. For the H7864 from the BA23 MicroVax II Enclosure I couldn't
find anything and in the end made my own schematic.
2) Upon initial glance, the 'first' board of
the three in the PSU module
appears
to be a pair of control relays, bridge rectifiers and
capacitors,
supplying +/- DC
voltages to the two individual PSU boards. Does anyone
know if I can
disconnect
these from the PSU boards safely* and measure their
outputs, and if so
what
voltages I should expect to see? That might be a good
initial test before
pointing a finger at the PSU board associated with the low outputs.
* I mean without component damage - I expect they might be sitting at a
significant DC voltage, so there's an obvious personal safety aspect too
;-)
cheers
Jules
26 Jan
26 Jan
12:30 p.m.
On 25/01/18 23:47, Rob Jarratt via cctalk wrote:
I doubt it. For the H7864 from the BA23 MicroVax II Enclosure I couldn't
find anything and in the end made my own schematic.
If it was externally sourced then DEC probably didn't have a schematic.
That was the case with the DEMSA (iirc) and also with various PSUs from
another company I worked for.
So doing rolling your own (if you indeed do need one) may be the only way.
Antonio
--
Antonio Carlini
arcarlini at iee.org
25 Jan
25 Jan
5:16 p.m.
On Thu, Jan 25, 2018 at 3:06 PM, Jules Richardson via cctalk
<cctalk at classiccmp.org> wrote:
Hi all,
I mentioned this in the thread where I'd asked about basic Microvax II info,
but it may have got lost in traffic...
The machine's H7260 PSU is somewhat unwell - one of the internal +5/12V
supplies appears to be healthy, but the other has outputs which are sitting
at around 2.5V (both on the 5V and 12V rails) under a test load.
Previously you also said:
OK, the top PSU of the pair in this machine appears to
be healthy. The
bottom one, however, sits at around 2.5V on both the +5V and +12V outputs.
That's without any cards, and the card-cage fan / temperature sensor
unplugged, i.e. just with my dummy loads in place, so it looks like that PSU
has issues (oh, and I tried decoupling the PSU from the backplane and
loading it via the drive connector, so it doesn't appear to be a backplane
issue).
Looking here at page 107 of the maintenance print set, BA123-A Basic
Enclosure Power Harness Wiring Diagram
http://www.bitsavers.org/pdf/dec/vax/630/MP02071_630QB_Mar85.pdf
Did you disconnect the wiring harness from J10 completely? If you
disconnected the temperature sensor module or the card-cage door
interlock switch I would expect that the power supply might not
operate normally.
I forget what the card-cage door interlock switch does, whether it
prevents the power supply from fully powering up, or if the power
supply does fully power up but also turns the fans on full blast if
the card-cage door is open which disrupts the normal cooling airflow.
I would check what my BA123 does, but that would require too much
moving of stuff about at the moment to get working space around it.
7:29 p.m.
On 01/25/2018 07:16 PM, Glen Slick via cctalk wrote:
On Thu, Jan 25, 2018 at 3:06 PM, Jules Richardson via
cctalk
I believe that I tried it both connected and disconnected, with no change
in behavior, but I'll double-check (not until tomorrow evening now, though).
I need to look at how the signals on J10 interact with the entire PSU
internals - it certainly seems to get a -15V feed from the lower PSU of the
pair (the one that's causing me issues), but many of the wires run to the
fourth, small PCB that's in the entire PSU assembly; I assume that one
handles fan control as it also connects the the PSU fan and to J7 (which
supplies the fan in the mass storage area)
It seems strange to me that one PSU of the pair would sit in an
under-voltage condition due to a sensor issue while the other one would
continue to operate normally, though; at this stage I think it's more
likely that I've got a PSU fault (or a fault with the rectifier/relay board
which supplies it - I'm most of the way through tracing a schematic for
that board, so I can test some voltages sensibly tomorrow)
Will let you know how it goes...
Jules
OK, the top PSU of the pair in this machine
appears to be healthy. The
bottom one, however, sits at around 2.5V on both the +5V and +12V outputs.
That's without any cards, and the card-cage fan / temperature sensor
unplugged, i.e. just with my dummy loads in place, so it looks like that PSU
has issues (oh, and I tried decoupling the PSU from the backplane and
loading it via the drive connector, so it doesn't appear to be a backplane
issue).
Looking here at page 107 of the maintenance print set, BA123-A Basic
Enclosure Power Harness Wiring Diagram
http://www.bitsavers.org/pdf/dec/vax/630/MP02071_630QB_Mar85.pdf
Did you disconnect the wiring harness from J10 completely? If you
disconnected the temperature sensor module or the card-cage door
interlock switch I would expect that the power supply might not
operate normally.
26 Jan
26 Jan
5:15 p.m.
On 01/25/2018 09:29 PM, Jules Richardson via cctalk wrote:
On 01/25/2018 07:16 PM, Glen Slick via cctalk wrote:
OK, I just verified that - it doesn't matter whether the J10 harness is
connected or not, same behavior.
The outputs of the main relay/regulator board which feeds both supplies
appear to be the same, so I don't think that's at fault. I think my next
step is to trace out the wiring for the temp sensor, along with that fourth
"fan control" PCB within the PSU module so that I can better understand
what it's trying to do.
After that... well, I need to work out how to get the entire supply
assembly to start up while it's open, so that I can work on it outside the
machine - then I can at least start looking for differences between the two
supplies (boy, are those things complex, but maybe I'll get lucky and it's
a bad solder joint or cap somewhere)
cheers
Jules
On Thu, Jan 25, 2018 at 3:06 PM, Jules Richardson
via cctalk
I believe that I tried it both connected and disconnected, with no change
in behavior, but I'll double-check (not until tomorrow evening now, though). OK, the top PSU of the pair in this machine
appears to be healthy. The
bottom one, however, sits at around 2.5V on both the +5V and +12V outputs.
That's without any cards, and the card-cage fan / temperature sensor
unplugged, i.e. just with my dummy loads in place, so it looks like that
PSU
has issues (oh, and I tried decoupling the PSU from the backplane and
loading it via the drive connector, so it doesn't appear to be a backplane
issue).
Looking here at page 107 of the maintenance print set, BA123-A Basic
Enclosure Power Harness Wiring Diagram
http://www.bitsavers.org/pdf/dec/vax/630/MP02071_630QB_Mar85.pdf
Did you disconnect the wiring harness from J10 completely? If you
disconnected the temperature sensor module or the card-cage door
interlock switch I would expect that the power supply might not
operate normally.
27 Jan
27 Jan
9:04 a.m.
On 01/26/2018 07:15 PM, Jules Richardson via cctalk wrote:
After that... well, I need to work out how to get the
entire supply
assembly to start up while it's open, so that I can work on it outside the
machine - then I can at least start looking for differences between the two
supplies (boy, are those things complex, but maybe I'll get lucky and it's
a bad solder joint or cap somewhere)
OK, now I feel like an idiot :-)
I started looking into this, and found that neither PSU would do anything
without being plugged into the (currently unpopulated) backplane. My
initial assumption was that there was some interaction between the
backplane and the "other" wires between it and the PSUs (i.e. the ones that
don't carry +5V or +12V).
However, on a whim (and I really don't know why I did this) I plugged the
+5V PSU connectors up to the backplane, but left the others (carrying +12V
and the other signals) unplugged. Doing that, things behaved as before -
the top PSU started operating normally, and the lower PSU started putting
out around 2.5V.
It got me wondering though - maybe it was simply a load issue, and with my
dummy loads I was right at the (lower) limit for the PSU regulation to
work; perhaps I had just enough for the top PSU to operate normally (but
only with the backplane hooked up too) and the lower PSU (for whatever
reason) wanted just that little bit more than the top one.
Anyway, I tweaked my loads to draw a little more current - and bingo! I'm
now getting stable +5V and +12V outputs on both PSUs, a working card cage
fan, and the DC OK light comes on.
So, in summary... with no load at all, the PSUs do nothing, but with *some*
load they either work OK, or the regulation is goofy and the outputs are low.
Time to add CPU/memory in, wire up a console cable and see what happens, I
guess.
cheers
Jules
9:44 a.m.
On Sat, Jan 27, 2018 at 9:04 AM, Jules Richardson via cctalk
<cctalk at classiccmp.org> wrote:
Time to add CPU/memory in, wire up a console cable and see what happens, I
guess.
With the PSU working everything else should be easier to get going after that.
Did you figure out what the hard drive is? A working RD54 would be
nice and probably worth more than any of the boards.
If you start with just the M7606 - CPU and the M7608 - 4MB ram once
you get a console cable built it shouldn't be too hard to get some
basic signs of life. The M7606 KA630 CPU console firmware is fairly
limited compared to the subsequent M7620 KA650 and M7625 KA655. It
doesn't have a SHOW command to show memory and device details, and
doesn't even have a HELP command to list available commands.
12:43 p.m.
On 01/27/2018 11:44 AM, Glen Slick via cctalk wrote:
On Sat, Jan 27, 2018 at 9:04 AM, Jules Richardson via
cctalk
<cctalk at classiccmp.org> wrote:
Well, there's someone's hand-written sticker on the top saying it's a lowly
RD51 - however the actual p/n sticker says it's an RD53-A and the defect
table lists it as ST506, 85MB.
Of course it probably doesn't work anyway ;-)
Time to add CPU/memory in, wire up a console cable and see what happens, I
guess.
With the PSU working everything else should be easier to get going after that.
Did you figure out what the hard drive is? A working RD54 would be
nice and probably worth more than any of the boards. If you start with just the M7606 - CPU and the M7608 -
4MB ram once
you get a console cable built it shouldn't be too hard to get some
basic signs of life.
Yes, that's where I'm at right now...
Console default is 8 data bits, 1 stop bit and no parity (at whatever speed
the switch is set to), correct?
I'm getting an alternating F and 9 on the hex LED diag display on the
console panel, no output at all on my terminal, and the M7606 starts with 4
solid red LEDs, then a solid green, then the middle pair of the red LEDs
start a simultaneous slow flash on/off (while the outer two plus the green
remain steady)
This is just with the M7606 CPU and M7808 memory plugged in, nothing else
(apart from the half-height board at the end of the backplane which
distributes signals to the RD53) - oh and the LEDs were observed with the
card cage cover switch held down, so 'under normal operating conditions'.
For console cable wiring I'm using the following ref:
http://www.mcmanis.com/chuck/computers/vaxen/panels.htm
There's a ref here for the status codes:
https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-c01608433
.. however, I'm not sure what the alternating codes mean; 'F' seems to be
power-related, but I don't know why the CPU would proceed if it thought
there was a power fault, and '9' is related to terminal ID, but it seems to
suggest it should automatically continue after 6 seconds.
cheers
Jules
1:02 p.m.
On 01/27/2018 02:43 PM, Jules Richardson via cctalk wrote:
There's a ref here for the status codes:
https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-c01608433
OK, looking at the 630QB tech manual now, which seems to be where these
were lifted from... and of course it points out that the LEDs on the M7606
correspond to the hex codes on the panel (i.e. my four red LEDs with the
middle pair flashing on and off equate to the alternating F-9 I'm seeing on
the hex LED)
The green LED on the CPU board is 'DC OK' and I'm also getting a green 'DC
OK' on the front of the machine, and it seems odd those would be lit if
there were a DC fault... but who knows. No explanation of the flashing
codes, either; the manual seems to imply that it steps through in sequence
down from 'F' and will stop when there's a fault.
cheers
Jules
1:31 p.m.
On Sat, Jan 27, 2018 at 1:02 PM, Jules Richardson via cctalk
<cctalk at classiccmp.org> wrote:
OK, looking at the 630QB tech manual now, which seems to be where these were
lifted from... and of course it points out that the LEDs on the M7606
correspond to the hex codes on the panel (i.e. my four red LEDs with the
middle pair flashing on and off equate to the alternating F-9 I'm seeing on
the hex LED)
The green LED on the CPU board is 'DC OK' and I'm also getting a green
'DC
OK' on the front of the machine, and it seems odd those would be lit if
there were a DC fault... but who knows. No explanation of the flashing
codes, either; the manual seems to imply that it steps through in sequence
down from 'F' and will stop when there's a fault.
Do you have a scope that you can use to monitor the BDCOK line? If it
is periodically dropping low that would cause a loop of the processor
halting and resetting. That could be happening fast enough that the DC
OK LED appears to stay on continuously.
That would be the same as periodically pressing the Restart button on
the front panel, which works by pulling down the BDCOK line. (See the
BA123-A Front Panel sheet, page 115 of the maintenance print sheet PDF
MP02071_630QB_Mar85.pdf)
1:34 p.m.
On 01/27/2018 03:02 PM, Jules Richardson via cctalk wrote:
OK, looking at the 630QB tech manual now, which seems
to be where these
were lifted from... and of course it points out that the LEDs on the M7606
correspond to the hex codes on the panel (i.e. my four red LEDs with the
middle pair flashing on and off equate to the alternating F-9 I'm seeing on
the hex LED)
Success!
I'd left the rear panel console settings how they were when I got the
machine, I mean 'T' for terminal, right? Wrong... arrow for terminal, 'T'
for "sit there alternating between F and 9, and driving Jules insane" ;)
I'm getting console output now, and it's getting down to trying to load
system software before failing (understandably, because I don't have any
other boards in there right now)
Will throw boards back in there and see what happens...
J.
1:54 p.m.
On Sat, Jan 27, 2018 at 4:34 PM, Jules Richardson via cctalk
<cctalk at classiccmp.org> wrote:
Success!
Congratulations!
I'd left the rear panel console settings how they
were when I got the
machine, I mean 'T' for terminal, right? Wrong... arrow for terminal,
'T'
for "sit there alternating between F and 9, and driving Jules insane" ;)
'T' is Test...
I'm getting console output now, and it's
getting down to trying to load
system software before failing (understandably, because I don't have any
other boards in there right now)
Will throw boards back in there and see what happens...
J.
2:02 p.m.
On Sat, Jan 27, 2018 at 4:34 PM, Jules Richardson via cctalk
<cctalk at classiccmp.org> wrote:
Success!
Congratulations!
I'd left the rear panel console settings how they
were when I got the
machine, I mean 'T' for terminal, right? Wrong... arrow for terminal,
'T'
for "sit there alternating between F and 9, and driving Jules insane" ;)
'T' is 'Loop Test'...
http://www.mcmanis.com/chuck/computers/vaxen/panels.htm
http://www.vaxination.ca/vms/microvax/uVAXII_HWinfo_Sep87.pdf
I'm getting console output now, and it's
getting down to trying to load
system software
Will throw boards back in there and see what happens...
Just be sure to follow the grant chain - the BA23 and BA123 have a
hybrid backplane where the first few slots are CD slots and the rest
are serpentine (3 CD slots for BA23, 5 CD slots for BA123).
http://web.frainresearch.org:8080/projects/pdp-11/chassis.php
http://vaxarchive.org/hardware/qbus1.html
-ethan
2:13 p.m.
On Sat, Jan 27, 2018 at 2:02 PM, Ethan Dicks via cctalk
<cctalk at classiccmp.org> wrote:
On Sat, Jan 27, 2018 at 4:34 PM, Jules Richardson via
cctalk
<cctalk at classiccmp.org> wrote:
http://www.bitsavers.org/pdf/dec/vax/630/
AX-GLBAA-MN_MicroVAX_II_630QB_Technical_Manual_Nov85.pdf
Page 37 of the PDF, page 2-6, Figure 2-2, CPU Patch Panel Insert.
Might be helpful to print out a copy of that figure and tape it to the
back of the BA123. I can never remember what the rotary and toggle
switch positions mean.
Success!
Congratulations!
I'd left the rear panel console settings how
they were when I got the
machine, I mean 'T' for terminal, right? Wrong... arrow for terminal,
'T'
for "sit there alternating between F and 9, and driving Jules insane" ;)
'T' is 'Loop Test'... Just be sure to follow the grant chain - the BA23 and
BA123 have a
hybrid backplane where the first few slots are CD slots and the rest
are serpentine (3 CD slots for BA23, 5 CD slots for BA123).
First 4 slots of a BA123 are Q22/CD, not the first 5. Page 22 of the
PDF, page 1-11, Figure 1-9 Backplane Grant Continuity of the same
manual referenced above.
2:39 p.m.
On Sat, Jan 27, 2018 at 5:13 PM, Glen Slick <glen.slick at gmail.com> wrote:
Good catch. I looked it up because I remembered that it was different
but I misread one of the links that mentioned slot 5... slot 5 is the
_first_ Q22 slot. 1-4 are Q22/CD.
Thanks for spotting this.
-ethan
Just be sure
to follow the grant chain - the BA23 and BA123 have a
hybrid backplane where the first few slots are CD slots and the rest
are serpentine (3 CD slots for BA23, 5 CD slots for BA123).
First 4 slots of a BA123 are Q22/CD, not the first 5. Page 22 of the
PDF, page 1-11, Figure 1-9 Backplane Grant Continuity of the same
manual referenced above.
3:38 p.m.
On 01/27/2018 04:02 PM, Ethan Dicks wrote:
On Sat, Jan 27, 2018 at 4:34 PM, Jules Richardson via
cctalk
<cctalk at classiccmp.org> wrote:
ish... :-)
It seems that my DHV11 board is bad - the system stalls on the PSU tests
with it plugged in.
The TK50 drive sits there with a rapidly-flashing light (I saw someone
saying that's likely the tape feed band or just dirty heads, so it might be
fixable).
The RD53 drive... hmm. I tried powering it without it plugged in, just to
see if it at least sounded healthy; it spun up to a "sensible-sounding"
speed for maybe 15 seconds, then spun down again. I tried again, and it
remained spun-up, but the initial spin-down was hardly encouraging.
With the drive actually hooked up to the controller, I'm getting "?54
RETRY" messages when it's trying to boot - however, I'm not entirely sure
what device it's trying to boot from! Maybe it's attempting the TK50, or
via Ethernet. I still need to read up on that and work out how to force it
to attempt a disk boot.
One other issue is that the RD53 was in the second bay (and hooked up to
the associated cables) - it makes me wonder if it was just an auxiliary
disk, and the system originally had another drive in the first bay
(containing actual OS) which has been removed.
Success!
Congratulations! Just be sure to follow the grant chain
Yup, I *think* I've got my head around how that works! (along with the
special slots for CPU/RAM)
cheers
Jules
4:01 p.m.
-----Original Message-----
From: cctalk [mailto:cctalk-bounces at classiccmp.org] On Behalf Of Jules
Richardson via cctalk
Sent: 27 January 2018 23:39
To: General Discussion: On-Topic and Off-Topic Posts
<cctalk at
classiccmp.org>
Subject: Re: DEC H7260 PSU fault
On 01/27/2018 04:02 PM, Ethan Dicks wrote:
with it
On Sat, Jan 27, 2018 at 4:34 PM, Jules Richardson
via cctalk
<cctalk at classiccmp.org> wrote:
ish... :-)
It seems that my DHV11 board is bad - the system stalls on the PSU tests Success!
Congratulations! plugged in.
The TK50 drive sits there with a rapidly-flashing light (I saw someone
saying
that's likely the tape feed band or just dirty
heads, so it might be
fixable).
The RD53 drive... hmm. I tried powering it without it plugged in, just to
see if
it
at least sounded healthy; it spun up to a
"sensible-sounding"
speed for maybe 15 seconds, then spun down again. I tried again, and it
remained spun-up, but the initial spin-down was hardly encouraging.
Did you hear the heads move? Spinning up and down again likely means it
couldn't find the first track. Very likely to be stiction from the rubber
bumper inside which has turned to goo.
With the drive actually hooked up to the controller,
I'm getting "?54
RETRY"
messages when it's trying to boot - however,
I'm not entirely sure what
device
it's trying to boot from! Maybe it's
attempting the TK50, or via Ethernet.
I still
need to read up on that and work out how to force it
to attempt a disk
boot.
One other issue is that the RD53 was in the second bay (and hooked up to
the
associated cables) - it makes me wonder if it was just
an auxiliary disk,
and the
system originally had another drive in the first bay
(containing actual
OS) which
has been removed.
Can't remember exactly, but not necessarily. You need to try commands like
BOOT DUA0/1/2/3, but if the heads are sticking you won't get far. I would
suggest net booting the machine as a satellite and then seeing if the local
disk is visible to the OS.
Just be sure to follow the grant chain
Yup, I *think* I've got my head around how that works! (along with the slots for CPU/RAM)
cheers
Jules
4:36 p.m.
With the drive actually hooked up to the controller, I'm getting "?54
RETRY" messages when it's trying to boot - however, I'm not entirely sure
what device it's trying to boot from! Maybe it's attempting the TK50, or
via Ethernet. I still need to read up on that and work out how to force it
to attempt a disk boot.
On a VAXStation/MicroVAX 2000 (which could behave the same as or completely
differently to an MV-II), "?54 RETRY" is what you get when attempting to boot
from the network when nothing is responding (also, if I
recall correctly, so
take this with a giant pinch of salt...)
Regards,
Peter Coghlan.
28 Jan
28 Jan
10:53 a.m.
On 27/01/18 23:38, Jules Richardson via cctalk wrote:
With the drive actually hooked up to the controller, I'm getting "?54
RETRY" messages when it's trying to boot - however, I'm not entirely
sure what device it's trying to boot from! Maybe it's attempting the
TK50, or via Ethernet. I still need to read up on that and work out
how to force it to attempt a disk boot.
B DUA0
should do it (although you may have yours set up for DUA1 or whatever
depending on the drive jumpers).
One other issue is that the RD53 was in the second bay
(and hooked up
to the associated cables) - it makes me wonder if it was just an
auxiliary disk, and the system originally had another drive in the
first bay (containing actual OS) which has been removed.
I think you get a different message if the disk is there but it has no
valid boot block.
If you had a later CPU (say the KA650) and anything other than the very
earliest ROM (which may never have made it out to the field) then you
get a bunch of additional commands (like SHOW DEVICE) which would help
diagnose what's what. You would be able to tell whether the M7555
was visible at the correct address, for example.
The suggestion of booting it as a satellite from a VMS cluster node is a
good idea (although if you've never set this up before then it may take you
some time to get it set up properly). SIMH makes this quite easy I think.
Antonio
--
Antonio Carlini
arcarlini at iee.org
27 Jan
27 Jan
5:11 p.m.
On 01/27/2018 02:43 PM, Jules Richardson via cctalk wrote:
MM, that doesn't sound good at all. The single LED on the
console panel is supposed to count down from F as it runs
through the on-board diags. It is supposed to send a 2-line
(I think) description of the board rev level and ROM
firmware level, and then do something like this:
F.....E.....D.....C....B.....A.....9.....8....7 (this is
the memory test, and can take a LONG time) and then continue,
finally leaving you with a console prompt or a boot or boot
failure message.
Jon
Well, there's someone's hand-written sticker on the top
saying it's a lowly RD51 - however the actual p/n sticker
says it's an RD53-A and the defect table lists it as
ST506, 85MB.
Of course it probably doesn't work anyway ;-)
The likelihood of a 30+ year old MFM hard drive working is
pretty small. If it DOES work, get the data off it as soon
as possible, the grease in the bearings are likely to have
turned into gunk and it won't last.
If you start
with just the M7606 - CPU and the M7608 -
4MB ram once
you get a console cable built it shouldn't be too hard to
get some
basic signs of life.
Yes, that's where I'm at right now...
Console default is 8 data bits, 1 stop bit and no parity
(at whatever speed the switch is set to), correct?
I'm getting an alternating F and 9 on the hex LED diag
display on the console panel, no output at all on my
terminal, and the M7606 starts with 4 solid red LEDs, then
a solid green, then the middle pair of the red LEDs start
a simultaneous slow flash on/off (while the outer two plus
the green remain steady)
10:51 a.m.
On 27 Jan 2018, at 17:04, Jules Richardson via cctalk
<cctalk at classiccmp.org> wrote:
On 01/26/2018 07:15 PM, Jules Richardson via cctalk wrote:
We?ve all been there including me earlier! When you have two PSUs in a machine but only
one is connected to the mains make sure you don?t connect a flying lead from the OTHER one
and spend 2 hours wondering why the bloody thing doesn?t work.
After that... well, I need to work out how to get
the entire supply
assembly to start up while it's open, so that I can work on it outside the
machine - then I can at least start looking for differences between the two
supplies (boy, are those things complex, but maybe I'll get lucky and it's
a bad solder joint or cap somewhere)
OK, now I feel like an idiot :-) Anyway, I tweaked my loads to draw a little more
current - and bingo! I'm now getting stable +5V and +12V outputs on both PSUs, a
working card cage fan, and the DC OK light comes on.
This is good, particularly since my own BA123 has been sat in 3 garages then a utility
room since 1998, you?re making me think it MIGHT just power up without any issues.
And THEN I can tackle the RL02 that?s sat on top of it? being able to read my own source
code from the 80s would be a special thing indeed.
?
Adrian/Witchy
Binary Dinosaurs - Celebrating Computing History from 1972 onwards
w: binarydinosaurs.co.uk t: @binarydinosaurs
f: facebook.com/binarydinosaurs
1:06 p.m.
On 01/27/2018 12:51 PM, Adrian Graham via cctalk wrote:
Hey, my rifa caps haven't exploded yet, either! :-D
See reply to Glen, though - I'm getting an alternating F-9 on the hex
display / CPU diag LEDs, which the docs don't mention as being a 'thing';
they seem to imply that startup codes should decrease from 'F' and stop if
there's a fault, not go into a condition where they alternate.
Anyway, I
tweaked my loads to draw a little more current - and bingo!
I'm now getting stable +5V and +12V outputs on both PSUs, a working
card cage fan, and the DC OK light comes on.
This is good, particularly since my own BA123 has been sat in 3 garages
then a utility room since 1998, you?re making me think it MIGHT just
power up without any issues. 
8:29 p.m.
On 01/27/2018 12:04 PM, Jules Richardson via cctalk wrote:
On 01/26/2018 07:15 PM, Jules Richardson via cctalk
wrote:
The PSUs used are a switching type and do not behave well with out about
a 5-10% load as a minimum.? They must have a load.
NOTE: the door switch does only one thing and that is to run the fans at
100% when the covers are off.? When they are
on then the switch allows the fan speed to be controlled by temperature.
FYI all this is covered in the BA123/microVAC manuals.? Seriously, if
your not familiar with Qbus and microVAX boxes
RTFM and as many as you can find.
Allison
After that... well, I need to work out how to get
the entire supply
assembly to start up while it's open, so that I can work on it
outside the
machine - then I can at least start looking for differences between
the two
supplies (boy, are those things complex, but maybe I'll get lucky and
it's
a bad solder joint or cap somewhere)
OK, now I feel like an idiot :-)
I started looking into this, and found that neither PSU would do
anything without being plugged into the (currently unpopulated)
backplane. My initial assumption was that there was some interaction
between the backplane and the "other" wires between it and the PSUs
(i.e. the ones that don't carry +5V or +12V).
However, on a whim (and I really don't know why I did this) I plugged
the +5V PSU connectors up to the backplane, but left the others
(carrying +12V and the other signals) unplugged. Doing that, things
behaved as before - the top PSU started operating normally, and the
lower PSU started putting out around 2.5V.
It got me wondering though - maybe it was simply a load issue, and
with my dummy loads I was right at the (lower) limit for the PSU
regulation to work; perhaps I had just enough for the top PSU to
operate normally (but only with the backplane hooked up too) and the
lower PSU (for whatever reason) wanted just that little bit more than
the top one.
Anyway, I tweaked my loads to draw a little more current - and bingo!
I'm now getting stable +5V and +12V outputs on both PSUs, a working
card cage fan, and the DC OK light comes on.
So, in summary... with no load at all, the PSUs do nothing, but with
*some* load they either work OK, or the regulation is goofy and the
outputs are low.
Time to add CPU/memory in, wire up a console cable and see what
happens, I guess.
cheers
Jules
28 Jan
28 Jan
6:35 p.m.
On 01/27/2018 10:29 PM, allison via cctalk wrote:
FYI all this is covered in the BA123/microVAC manuals.
Seriously, if
your not familiar with Qbus and microVAX boxes
RTFM and as many as you can find.
The 630QB tech manual seems to be pretty comprehensive, and I have digested
that one now, I think :-)
My RD53 is definitely unwell, though - sometimes the speed fluctuates, and
other times it sounds like it's stable (and I can hear the heads move) but
talking to it is a no-go. I've no idea what (if anything) might be on it
anyway, though - as set up when I got it, it was in the second drive bay
(and hooked up via some rather flaky non-DEC cables)
I'm not sure at the moment where I want to proceed from here; net-booting
might be an option, but I'm not sure if I have any transceivers on this
side of the Atlantic.
cheers
Jules
25 Jan
25 Jan
11:30 p.m.
-----Original Message-----
From: cctalk [mailto:cctalk-bounces at classiccmp.org] On Behalf Of Glen Slick
via cctalk
Sent: 26 January 2018 01:17
To: General Discussion: On-Topic and Off-Topic Posts <cctalk at classiccmp.org>
Subject: Re: DEC H7260 PSU fault
On Thu, Jan 25, 2018 at 3:06 PM, Jules Richardson via cctalk
<cctalk at classiccmp.org> wrote:
If the door is open the fans run full blast.
Hi all,
I mentioned this in the thread where I'd asked about basic Microvax II
info, but it may have got lost in traffic...
The machine's H7260 PSU is somewhat unwell - one of the internal
+5/12V supplies appears to be healthy, but the other has outputs which
are sitting at around 2.5V (both on the 5V and 12V rails) under a test load.
Previously you also said:
OK, the top PSU of the pair in this machine
appears to be healthy. The
bottom one, however, sits at around 2.5V on both the +5V and +12V outputs.
That's without any cards, and the card-cage fan / temperature sensor
unplugged, i.e. just with my dummy loads in place, so it looks like
that PSU has issues (oh, and I tried decoupling the PSU from the
backplane and loading it via the drive connector, so it doesn't appear
to be a backplane issue).
Looking here at page 107 of the maintenance print set, BA123-A Basic
Enclosure Power Harness Wiring Diagram
http://www.bitsavers.org/pdf/dec/vax/630/MP02071_630QB_Mar85.pdf
Did you disconnect the wiring harness from J10 completely? If you disconnected
the temperature sensor module or the card-cage door interlock switch I would
expect that the power supply might not operate normally.
I forget what the card-cage door interlock switch does, whether it prevents the
power supply from fully powering up, or if the power supply does fully power
up but also turns the fans on full blast if the card-cage door is open which
disrupts the normal cooling airflow. I would check what my BA123 does, but that would
require too much moving of
stuff about at the moment to get working space around it.
2300
days inactive
2304
days old
25 comments
9 participants
participants (9)
-
a.carlini＠ntlworld.com -
allisonportable＠gmail.com -
binarydinosaurs＠gmail.com -
cctalk＠beyondthepale.ie -
elson＠pico-systems.com -
ethan.dicks＠gmail.com -
glen.slick＠gmail.com -
jules.richardson99＠gmail.com -
robert.jarratt＠ntlworld.com