Onyx Problems, seemingly power related

SGI hardware problems, solutions, tips, hacks, etc.
Forum rules
Any posts concerning pirated software or offering to buy/sell/trade commercial software are subject to removal.
User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Onyx Problems, seemingly power related

Unread postby Dodoid » Tue Aug 22, 2017 2:01 pm

So, my new Onyx has ran a total of 1 time so far. The first time I booted it, it said "-12VDC OVER VOLTAGE" and shut down, so I had to run it again. That time, it worked. I used it for 15 minutes or so, and then shut it down. Now, a few days later, I'm trying to run it again, but every time I turn it on, it gets part way through the boot process on the System Controller screen (how far this is varies) and then says either "-12VDC OVER VOLTAGE" or "-5.2VDC OVER VOLTAGE" and stops booting. Interestingly, it seems to always be one of those two negative voltages that have problems, never any of the other voltages. Then I have to turn it off with the key.

I thought maybe it was using too much power from the wall and shutting down, but according to my plug-in watt meter thing when it shut down it was using 1130 watts or so, and the power from the wall was still at 116V AC (119V before the Onyx ran), so I don't think that's the issue. I have also not seen any POK(something) errors, which can apparently also happen but do not appear to be an issue on my machine.

What's the problem? I have heard to look for red LEDs on the boards, but they all look normal (and are not red) until the system shuts down, when all the board LEDs turn off (presumably, all the boards stop getting power).

Thanks!
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
HurricaneJames
Posts: 85
Joined: Thu Jul 06, 2017 10:08 pm
Location: Santa Clara, CA

Re: Onyx Problems, seemingly power related

Unread postby HurricaneJames » Tue Aug 22, 2017 2:42 pm

I'm guessing you have not replaced and/or reformed the capacitors in the power supply? I just did that with the power supply in my Octane. It was not great fun, but it made the power supply come back to life.
:320: :1600SW: :1600SW: :Octane: :Octane: Image

User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Tue Aug 22, 2017 2:50 pm

No, I have not changed any capacitors. For the Onyx, would the ones in question be in the big power supply thing that slides out of the back with three screws, or on one of the little power daughter boards behind the rear panel?
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
jan-jaap
Donor
Donor
Posts: 4867
Joined: Thu Jun 17, 2004 11:35 am
Location: Wijchen, The Netherlands
Contact:

Re: Onyx Problems, seemingly power related

Unread postby jan-jaap » Tue Aug 22, 2017 3:00 pm

The Onyx has a 110/220VAC to 48VDC offline switcher (OLS). Most boards in the Onyx are fed with 48VDC and transform it locally to the voltages they need.

The OLS is probably fine. Don't mess with it.

Dodoid wrote:either "-12VDC OVER VOLTAGE" or "-5.2VDC OVER VOLTAGE"

I'm pretty sure the only regulators for those voltages are on the IO4 VCAM.

There's a Diagnostic Roadmap for the Onyx systems which should have the definitive answer to that.
:PI: :Indigo: :Indigo: :Indy: :Indy: :Indy: :Indigo2: :Indigo2: :Indigo2IMP: :Octane: :Octane2: :O2: :O2+: Image :Fuel: :Tezro: :4D70G: :Skywriter: :PWRSeries: :Crimson: :ChallengeL: :Onyx: :O200: :Onyx2: :O3x02L:
To accentuate the special identity of the IRIS 4D/70, Silicon Graphics' designers selected a new color palette. The machine's coating blends dark grey, raspberry and beige colors into a pleasing harmony. (IRIS 4D/70 Superworkstation Technical Report)

User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Tue Aug 22, 2017 3:53 pm

Took the board out, nothing looked wrong, so I put it back together. Now I get "-12VDC UNDER VOLTAGE" and it happens much faster. So, opposite problem, I guess? Or maybe it's not connected right?
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Tue Aug 22, 2017 4:31 pm

Disregard previous post, the board wasn't in right. So now we have the "-5.2VDC OVER VOLTAGE" and "-12VDC OVER VOLTAGE" again. No blown caps visible. Should I check the power bricks on the rear of the backplane?
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

mazzinia
Posts: 96
Joined: Thu Nov 18, 2010 1:14 pm
Location: Italy

Re: Onyx Problems, seemingly power related

Unread postby mazzinia » Tue Aug 22, 2017 5:09 pm

page 2-27 of the pdf that was posted (think fits your case):

2.5.1.1 Power Not Operating Within Acceptable Voltage Levels
Note: Before starting to troubleshoot the board, verify that both the 3.3V and 5.0V bricks
are working within acceptable levels. If the 5.0V brick fails while the 3.3V brick
continues to work, the ASICs will overheat.
1. Check the power fault LEDs on the board (see Figure 2-3).
2. If the fault LEDs are off, but the bricks are still suspect, check the voltages at the red
and black test points at the back of the board.


and

3.2.2 System and Power Boards onward


....


I think you should try to disable the VCAM, and test if by chance it's that one that needs being changed
:O2: :Fuel: HP c8000 Digital PWS 500au (x2) , IBM 9402-P03

User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Tue Aug 22, 2017 6:42 pm

I read those sections. The real time voltage display doesn't show anything too out of spec, until it runs for a while (again, this varies). Then, I get the message I described before.

Code: Select all

-5.2VDC OVER VOLTAGE
99:99:54  :0 ??? 99
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Tue Aug 22, 2017 9:59 pm

Dug through some of the (individually static bagged and bubble wrapped) spares I got with the Onyx and O2K. Besides learning that I have multiple Onyx CPU boards, a whole box of NUMAlink router boards, like 2 or 3 Origin2000/Onyx2 MSCs, a few more power supplies, a box of IO6s, some miscellaneous Octane stuff, and like an O2 and a half, I found another IO4 with VCAM installed. I will try switching the VCAM tomorrow.
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
jan-jaap
Donor
Donor
Posts: 4867
Joined: Thu Jun 17, 2004 11:35 am
Location: Wijchen, The Netherlands
Contact:

Re: Onyx Problems, seemingly power related

Unread postby jan-jaap » Wed Aug 23, 2017 1:20 am

Dodoid wrote:

Code: Select all

99:99:54  :0 ??? 99

Your system controller's Dallas chip has run out of juice :P

The Onyx has 505 and 512 power modules on the back of the backplane. They provide +5V and +12V to graphics (and I think the IO4, but I'm not 100% sure). The CPU and MC3 boards have their own converters for +3V3 and +5V, and others down the line.

The Power-OK error lines are OR-ed together, so a +5V fault can be a pain to locate. But like I said, there's only one -12V regulator and it's on the IO4 VCAM.
POKA.gif

I'm pretty sure the -12V regulator is at fault here. As you can see the -5V regulator is fed from the -12V line, so with that one out of whack the -5V can't be expected to work. NB: there seems to be an error in Figure 3.5, where it mentions a "-12V regulator (+12V to -5V)". I think Table 3.2 is correct.

The regulators themselves are not that special, but I've discovered in the past that soldering anything on an IO4 is near impossible because they're so thick. Maybe you're lucky and the VCAM board is thinner, but if you've got a working spare it's certainly easier. Beware that only the last revision of the VCAM works with IR graphics.
:PI: :Indigo: :Indigo: :Indy: :Indy: :Indy: :Indigo2: :Indigo2: :Indigo2IMP: :Octane: :Octane2: :O2: :O2+: Image :Fuel: :Tezro: :4D70G: :Skywriter: :PWRSeries: :Crimson: :ChallengeL: :Onyx: :O200: :Onyx2: :O3x02L:
To accentuate the special identity of the IRIS 4D/70, Silicon Graphics' designers selected a new color palette. The machine's coating blends dark grey, raspberry and beige colors into a pleasing harmony. (IRIS 4D/70 Superworkstation Technical Report)

User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Wed Aug 23, 2017 10:52 am

Alright, so I changed the VCAM board out for a spare, and it has the same problem. Perhaps both boards are faulty, or perhaps it's something else.
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
jan-jaap
Donor
Donor
Posts: 4867
Joined: Thu Jun 17, 2004 11:35 am
Location: Wijchen, The Netherlands
Contact:

Re: Onyx Problems, seemingly power related

Unread postby jan-jaap » Thu Aug 24, 2017 7:14 am

I had some POK_A errors with my Onyx RE2 and they were all related to the negative voltages and the VCAM. Maybe they fail more often. I went through several boards with that Onyx (it had the nasty habit of eating at least one board per year) but not enough to do reliable statistics :)

It should be easy to spot that regulator, look up the datasheet, and attach a voltmeter or oscilloscope to the inputs and outputs of the thing, or the various VCC and GND pins on the VCAM. If it produces stable (but wrong) voltage, one of the components around regulator controller may be out of spec. If the voltage is correct and stable, it may even be the monitoring circuit that's wrong. The LM393 chip(s) on the VCAM are probably used for that.
:PI: :Indigo: :Indigo: :Indy: :Indy: :Indy: :Indigo2: :Indigo2: :Indigo2IMP: :Octane: :Octane2: :O2: :O2+: Image :Fuel: :Tezro: :4D70G: :Skywriter: :PWRSeries: :Crimson: :ChallengeL: :Onyx: :O200: :Onyx2: :O3x02L:
To accentuate the special identity of the IRIS 4D/70, Silicon Graphics' designers selected a new color palette. The machine's coating blends dark grey, raspberry and beige colors into a pleasing harmony. (IRIS 4D/70 Superworkstation Technical Report)

User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Thu Aug 24, 2017 9:57 am

My multimeter connected to the test pins and the System Controller's "Voltage Status" screen say the same thing. The -12V line is actually more like -13.1V. So, my monitoring circuitry is fine, but the actual voltage is wrong. Now it's a matter of figuring out why. Would you suggest replacing the regulator itself, or is there some other components that are more likely to be the source of the problem?
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
praetor242
Posts: 202
Joined: Thu Feb 23, 2017 5:41 pm
Location: Arlington, TX

Re: Onyx Problems, seemingly power related

Unread postby praetor242 » Thu Aug 24, 2017 12:01 pm

Awws :( *hugs Dodoid's sick Onyx* It's okay little buddy! You're in a good place.
Image - trajan - R5000 180Mhz - 128 MB RAM - 147GB HDD
Image - hadrian - R5000 180Mhz - 128 MB RAM - 76 GB HDD

User avatar
Dodoid
Posts: 581
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Thu Aug 24, 2017 1:25 pm

praetor242 wrote:*hugs Dodoid's sick Onyx* It's okay little buddy!

I did get it to run for a little while, long enough to run a demo or two when I first got it, but it's definitely not stable yet. "little" is funny, seeing as it weighs more than I do :D .
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image


Return to “SGI: Hardware”

Who is online

Users browsing this forum: No registered users and 2 guests