16 Sep
2011
16 Sep
'11
8:44 a.m.
I'm convinced that my house was built over an Indian burial ground. Or
it was owned by someone heavily invested in everything but DEC...
I powered up my AlphaServer 1000A last night and it looks like it has
problems too. It powers up and the drives & fans spin up. The lcd
showed 'ec.' only once, every other time it is blank. Nothing out of
serial port 1 and the onboard VGA never initializes.
I tried setting the 'failsafe boot' jumper and I put a blank floppy in
the drive just to see if it would try to read the floppy and it didn't
even try :(
Any ideas at all?
So now it looks like both of my alphas are shot (the PWS500au never
initializes video, does anything with serial, and turns off all diag
lights) and my DECStation 5000/120 has a bad power supply.
On the plus side, all of my terminals still work and both of my vaxes
(MicroVAX 3100/90 and VAXStation 4000/90) boot to VMS without
problems.
It brings me to a question: are the alpha systems inherently fragile?
The dead DECStation is an obvious hardware issue due to (probably) a
leaked capacitor which is understandable due to its age, but the Alphas
are just hosed in a software kind of way.
Brian
16 Sep
16 Sep
9:06 a.m.
Some diagnosis required before any conclusions can be drawn.
divide the system into easily tested items and diagnose as you go,
so measure power supplies,,
is there a level/pin to enable the supply check it
reseat items, look for corroded contacts too
Dave Caroline
10:14 a.m.
At 8:44 AM -0400 9/16/11, Brian Wheeler wrote:
It brings me to a question: are the alpha systems
inherently fragile?
The dead DECStation is an obvious hardware issue due to (probably) a
leaked capacitor which is understandable due to its age, but the Alphas
are just hosed in a software kind of way.
I ran my PWS 433au for several years with a dead PCI slot. In fact
before I gave up on it in favor of an XP1000, I believe something
else had died. That actaully impressed me.
My XP1000/667's powersupply died, I had to move the CPU over the the
XP1000/500.
At the same time, don't get me started on VAXstation 4000's. I have
a 60, a 90, and a VLC. All three have issues, the 90 won't boot, the
60 is flakey, and the VLC has a bad powersupply.
Honestly I need to get things setup so I can powerup and check all my
DEC hardware, it's been months since I booted any of it. :-( I
simply don't have the time anymore, and I can't afford to keep a VAX
and Alpha running 24/7.
Zane
--
| Zane H. Healy | UNIX Systems Administrator |
| healyzh at aracnet.com | OpenVMS Enthusiast |
| | Photographer |
+----------------------------------+----------------------------+
| My flickr Photostream |
| http://www.flickr.com/photos/33848088 at N03/ |
11:15 a.m.
Hi Brian,
Sorry to hear about your continued bad luck.
If it's any consolation I have the opposite problem here - I keep killing
terminals although the computers work OK. Touch wood and all that.
Regards, Mark.
On Fri, 16 Sep 2011, Brian Wheeler wrote:
I'm convinced that my house was built over an
Indian burial ground. Or
it was owned by someone heavily invested in everything but DEC...
I powered up my AlphaServer 1000A last night and it looks like it has
problems too. It powers up and the drives & fans spin up. The lcd
showed 'ec.' only once, every other time it is blank. Nothing out of
serial port 1 and the onboard VGA never initializes.
I tried setting the 'failsafe boot' jumper and I put a blank floppy in
the drive just to see if it would try to read the floppy and it didn't
even try :(
Any ideas at all?
So now it looks like both of my alphas are shot (the PWS500au never
initializes video, does anything with serial, and turns off all diag
lights) and my DECStation 5000/120 has a bad power supply.
On the plus side, all of my terminals still work and both of my vaxes
(MicroVAX 3100/90 and VAXStation 4000/90) boot to VMS without
problems.
It brings me to a question: are the alpha systems inherently fragile?
The dead DECStation is an obvious hardware issue due to (probably) a
leaked capacitor which is understandable due to its age, but the Alphas
are just hosed in a software kind of way.
Brian
17 Sep
17 Sep
6:37 a.m.
I powered up my AlphaServer 1000A last night and it looks like it has
problems too. It powers up and the drives & fans spin up. The lcd
showed 'ec.' only once, every other time it is blank. Nothing out of
serial port 1 and the onboard VGA never initializes.
I tried setting the 'failsafe boot' jumper and I put a blank floppy in
the drive just to see if it would try to read the floppy and it didn't
even try :(
Any ideas at all?
Check the keyboard. I've had a AS100A that hung completely at test EC and
the problem turned out to be a book resting on the keyboard holding down a key.
Also check the mouse. If you disconnect both, it should beep and complain but
get past the problem.
I think test EC also includes onboard VGA but mine don't have that option.
Isn't there a jumper to disable onboard VGA? Maybe that might be worth trying
in case your onboard VGA is hosed?
I vaguely remember putting a graphics card in an Alphaserver 800 that does
have onboard VGA but I think doesn't have a jumper to disable it. AFAIR
nothing worked until I changed a jumper on the graphics card to disable VGA
there.
It brings me to a question: are the alpha systems inherently fragile?
The dead DECStation is an obvious hardware issue due to (probably) a
leaked capacitor which is understandable due to its age, but the Alphas
are just hosed in a software kind of way.
I've had a fair amount of grief with DEC 3000/600s and Alphaserver 1000As
as well as my VAX 4000/100A.
Regards,
Peter Coghlan.
11:43 a.m.
On Sat, 2011-09-17 at 11:37 +0100, Peter Coghlan wrote:
I've tried several things to no avail:
* I disabled onboard vga
* there's nothing in the PCI slots now
* I reseated the processor board
* I unplugged the scsi bus and floppy cable
* I even went as far as removing all of the memory
and still there's no change: nothing from serial port 1, no beeps, no
OCP messages. The memory pull should have at least issued a set of
1-3-3 beeps, right? I verified and the speaker is working.
I wasn't able to find able to find a voltage pinout for the power supply
to test it.
Brian
I powered up my AlphaServer 1000A last night and it looks like it has
problems too. It powers up and the drives & fans spin up. The lcd
showed 'ec.' only once, every other time it is blank. Nothing out of
serial port 1 and the onboard VGA never initializes.
I tried setting the 'failsafe boot' jumper and I put a blank floppy in
the drive just to see if it would try to read the floppy and it didn't
even try :(
Any ideas at all?
Check the keyboard. I've had a AS100A that hung completely at test EC and
the problem turned out to be a book resting on the keyboard holding down a key.
Also check the mouse. If you disconnect both, it should beep and complain but
get past the problem.
I think test EC also includes onboard VGA but mine don't have that option.
Isn't there a jumper to disable onboard VGA? Maybe that might be worth trying
in case your onboard VGA is hosed?
I vaguely remember putting a graphics card in an Alphaserver 800 that does
have onboard VGA but I think doesn't have a jumper to disable it. AFAIR
nothing worked until I changed a jumper on the graphics card to disable VGA
there.
It brings me to a question: are the alpha systems inherently fragile?
The dead DECStation is an obvious hardware issue due to (probably) a
leaked capacitor which is understandable due to its age, but the Alphas
are just hosed in a software kind of way.
I've had a fair amount of grief with DEC 3000/600s and Alphaserver 1000As
as well as my VAX 4000/100A.
Regards,
Peter Coghlan.
1:03 p.m.
I've tried several things to no avail:
* I disabled onboard vga
* there's nothing in the PCI slots now
* I reseated the processor board
That would have been my next advice. I have found the position of the
processor board to be very critical at times.
* I unplugged the scsi bus and floppy cable
* I even went as far as removing all of the memory
and still there's no change: nothing from serial port 1, no beeps, no
OCP messages. The memory pull should have at least issued a set of
1-3-3 beeps, right? I verified and the speaker is working.
Yes. I would expect it to beep when there is no memory.
It seems that the B-Cache on the processor board can be prone to failure.
I've had this but I can't quite recall what the symptoms were. I will
check to see if I have any notes. The B-Cache can be disabled by
changing a jumper on the processor board and the machine will run without
it at reduced performance. The location of the jumper varies depending
on what processor you have.
I wasn't able to find able to find a voltage pinout for the power supply
to test it.
I think I took note of the voltages on the power supply connectors when I
was diagnosing a recent problem on mine. I will see if I can find my
notes.
(The power supply is uses significant amounts of power even when the machine
is switched off, so unplug when not in use.)
Regards,
Peter Coghlan.
5:07 p.m.
I wasn't able to find able to find a voltage pinout for the power supply
to test it.
Here are the notes I made on the power supply connectors in my Alphaserver 1000A
which has a single power supply in the lower bay which I called PSU0.
The readings were taken with a cheap and cheerful analogue multimeter which may
be reading a little on the high side.
The location of pin 1 is marked on the board but I couldn't tell whether the
numbers run up and down or across. I assumed up and down, like an IDC connector.
J21 power connector on main board:
PSU Main board
-------- ----------
1 (purple) GND GND
2 (black) Output? Input? -7.7V when running
3 (black) GND GND
4 (black) GND at PSU0 Connected PSU detect?
5 (blue) Output? Input? +12V when running
6 (black) PSU1 only Connected +4.2V when running PSU detect?
7 (yellow) Output? Input? -12V when running
8 (red) Output Input +5V at all times
9 (red) Output? Input? +5.1V when running
10 (grey) Output? Input? -5V when running
11 (black) Connected GND
12 (orange) Output? Input? +3.5V when running
13 (white) Input? Output? Normally 0V. Pulse high to trip PSU.
14 (black) Connected GND
15 (white) PSU0 only +4.5V after running. Power good?
16 (white) Input Output Ground to enable PSU power on
17 (white) PSU1 only 0V at all times? Power good?
J19 power connector on main board:
All red leads +5.1V
All black leads 0V
J20 power connector on main board:
All orange leads +3.5V
All black leads 0V
Regards,
Peter Coghlan.
19 Sep
19 Sep
9:01 a.m.
I ran my meter across all of them and I'm getting the same voltages you
did.
Any other ideas on what could be this beastie's problem? Its looking
more an more like something I'm not going to be able to work around.
Brian
On Sat, 2011-09-17 at 22:07 +0100, Peter Coghlan wrote:
I wasn't able to find able to find a voltage pinout for the power supply
to test it.
Here are the notes I made on the power supply connectors in my Alphaserver 1000A
which has a single power supply in the lower bay which I called PSU0.
The readings were taken with a cheap and cheerful analogue multimeter which may
be reading a little on the high side.
The location of pin 1 is marked on the board but I couldn't tell whether the
numbers run up and down or across. I assumed up and down, like an IDC connector.
J21 power connector on main board:
PSU Main board
-------- ----------
1 (purple) GND GND
2 (black) Output? Input? -7.7V when running
3 (black) GND GND
4 (black) GND at PSU0 Connected PSU detect?
5 (blue) Output? Input? +12V when running
6 (black) PSU1 only Connected +4.2V when running PSU detect?
7 (yellow) Output? Input? -12V when running
8 (red) Output Input +5V at all times
9 (red) Output? Input? +5.1V when running
10 (grey) Output? Input? -5V when running
11 (black) Connected GND
12 (orange) Output? Input? +3.5V when running
13 (white) Input? Output? Normally 0V. Pulse high to trip PSU.
14 (black) Connected GND
15 (white) PSU0 only +4.5V after running. Power good?
16 (white) Input Output Ground to enable PSU power on
17 (white) PSU1 only 0V at all times? Power good?
J19 power connector on main board:
All red leads +5.1V
All black leads 0V
J20 power connector on main board:
All orange leads +3.5V
All black leads 0V
Regards,
Peter Coghlan.
20 Sep
20 Sep
11:37 a.m.
I ran my meter across all of them and I'm getting the same voltages you
did.
Any other ideas on what could be this beastie's problem? Its looking
more an more like something I'm not going to be able to work around.
As I mentioned before, I had a lot of trouble with the edge connector
that the CPU slots into. I think the machine has been in a damp atmosphere
and suffered from corrosion. I ended up cleaning it and reseating the card
several times before it worked right. There were all sorts of varied symptoms
including doing nothing.
I also had a B-Cache failure but as far as I recall, that resulted in
large numbers of error messages during the self tests. As the B-Cache can
be disabled (or tested I think) by just moving jumpers, it may be worth
trying that, just in case.
I'd be happy to take measurements from mine for comparison purposes but
without schematics, it is difficult to know what would be useful to measure
other than the power supply lines which we've already covered.
Regards,
Peter Coghlan.
17 Sep
17 Sep
9:28 p.m.
Hi Brian,
I have a several 5000/125,133, 200,etc if you need a complete unit or
a parts unit. I also have several 3100's, but have a friend looking
for a complete 31/90 if you want to get rid of yours.
I'm next door in Illinois.
Thanks, Paul
On Fri, Sep 16, 2011 at 7:44 AM, Brian Wheeler <bdwheele at indiana.edu> wrote:
I'm convinced that my house was built over an
Indian burial ground. ?Or
it was owned by someone heavily invested in everything but DEC...
I powered up my AlphaServer 1000A last night and it looks like it has
problems too. ?It powers up and the drives & fans spin up. ?The lcd
showed 'ec.' only once, every other time it is blank. ?Nothing out of
serial port 1 and the onboard VGA never initializes.
I tried setting the 'failsafe boot' jumper and I put a blank floppy in
the drive just to see if it would try to read the floppy and it didn't
even try :(
Any ideas at all?
So now it looks like both of my alphas are shot (the PWS500au never
initializes video, does anything with serial, and turns off all diag
lights) and my DECStation 5000/120 has a bad power supply.
On the plus side, all of my terminals still work and both of my vaxes
(MicroVAX 3100/90 and VAXStation 4000/90) boot to VMS without
problems.
It brings me to a question: ?are the alpha systems inherently fragile?
The dead DECStation is an obvious hardware issue due to (probably) a
leaked capacitor which is understandable due to its age, but the Alphas
are just hosed in a software kind of way.
Brian
4694
days inactive
4698
days old
10 comments
6 participants
participants (6)
-
bdwheele＠indiana.edu -
CCTECH＠beyondthepale.ie -
dave.thearchivist＠gmail.com -
healyzh＠aracnet.com -
mark＠wickensonline.co.uk -
useddec＠gmail.com