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
praetor242
Posts: 246
Joined: Thu Feb 23, 2017 5:41 pm
Location: Arlington, TX

Re: Onyx Problems, seemingly power related

Unread postby praetor242 » Thu Aug 24, 2017 1:47 pm

Little is funny, cuz it's bigger than my mini-fridge. You'll get it working, bud!
:O2: - trajan - R5000 180Mhz - 256 MB RAM - 300GB HDD
:O2: - hadrian - R5000 180Mhz - 128 MB RAM - 147 GB HDD
:Octane2: - augustus - COMING SOON :D :D

User avatar
Zerolapse
Posts: 417
Joined: Mon Sep 12, 2005 8:32 pm
Location: New York, USA

Re: Onyx Problems, seemingly power related

Unread postby Zerolapse » Thu Aug 24, 2017 10:32 pm

Probably unrelated, but for *other* reasons do make sure that your power supply isn't one of the recalled ones...

I think the only safe part number is 060-0120-001....anything else is a serious fire hazard.

-Jesse
:O3x06R: :ChallengeL: :Octane: :Octane2: :O2: :Indy: (4x Challenge S)
The current fleet.

User avatar
Dodoid
Posts: 634
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 11:13 pm

So my dad (who understands power circuitry of this sort far better than I do) and I fiddled with the resistors on the VCAM for a while. The official SGI documentation is wrong about the positions of the regulators, the 5.2V one is closer to the side and the 12V is further inwards on the board. We added a 22k resistor on the bottom of the board, between a pin related to the voltage divider used to create the 2.2V reference voltage for the regulator and a via, and managed to make the regulator output -10.1V instead of -13.1 as it was before. Still not correct, but with a different resistor we should be able to get it to -12. The -5.2 also appears to have drifted to -5.4, which according to some parts of the docs is actually correct but according to the labeling on the board is incorrect. Either way, it should be within spec, so seeing an over voltage shutdown related to -5.2 being -5.4 was weird. Quite possibly has to do with the abnormally low -12V caused by the resistor tests. Will try to get the values displayed by the system controller (and the multimeter connected to the VCAM via some long wires) as close to the nominal as possible, and then if it's still complaining, then something else is up. Perhaps the regulators themselves would need replacement.
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
Dodoid
Posts: 634
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Sat Aug 26, 2017 10:47 pm

Alright, update time.

So, it seems like the voltages the VCAM is making are very reasonable since I added the resistors to adjust it. Both the System Controller's voltage screen and external measurement using a multimeter confirm that they are well within spec and not much to worry about.

I probed around with a multimeter and mapped out the power monitoring NAND gate latches on the VCAM. After connecting them with some probes to an oscilloscope, it appears that the NAND gates on the VCAM intended to latch on power failure are, in fact, never triggered. This would imply that the LM393 chips on the VCAM are not reporting any issues, which makes sense as the voltages are well within spec, especially with the adjustments made to the regulators.

As the outputs of the NAND latches are directly broken out to the IO4 board via the connectors on the bottom of the VCAM (4th, 6th, and 8th pin from the left on the bottom row of the left connector, looking at the top of the board with VME connector facing away from you) this also shows why the LEDs on the IO4 intended to indicate power failures never turn on. They simply never recieve any signal indicating that there has been a power fault.

As such, I believe that the primary issue, the one which is causing the system to shut down, has to do with incorrect -12V readings by faulty monitoring circuitry not located on the VCAM. Where else in the Onyx is power monitoring circuitry present? Is there something on the backplane that might be important?
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
thegoldbug
Posts: 384
Joined: Sat Jul 12, 2003 8:36 pm
Location: AZ, hot like Crematoria
Contact:

Re: Onyx Problems, seemingly power related

Unread postby thegoldbug » Sun Aug 27, 2017 12:45 pm

Is this for the Deskside Onyx or the Tower Onyx?

I have spare boards, power modules, etc if you want to go that route.

Let me know. Maybe a trade for some of those 3D printed Indigos.

thegoldbug
thegoldbug
:ChallengeXL: :ChallengeXL: :ChallengeXL: :OnyxR: :ChallengeL: :ChallengeL: :ChallengeL: :ChallengeL: :Onyx: :O200: :Fuel: :Fuel: :Octane2: :Octane: :Octane: :Octane: :Octane: :Indigo2IMP: :Indigo2IMP: :Indigo2IMP: :Indigo2IMP: :O2: :Indigo2: :Indigo2: :Indigo2: :Indigo2: :320: :320: :320: :320: :Indy: x5 :Indy: x2 :1600SW: x5 :Indigo: x7
Vault L, 6-1200s, Fibre Hub, Dials, Boards, spares, 3 Sun Blade 2000s, Sun Fire V480, 2-Hero 2000 Robots, 5-Hero 1 Robots

User avatar
guardian452
Donor
Donor
Posts: 3447
Joined: Tue Aug 21, 2007 10:12 pm
Location: United States
Contact:

Re: Onyx Problems, seemingly power related

Unread postby guardian452 » Sun Aug 27, 2017 3:02 pm

IIRC on the back where the 505/303/etc vrms go, one of the boards holds the system controller; the screen on the front panel is just that, a screen, and the actual system controller is on this board. I thought it had a couple voltage rails of it's own.

Sorry it's been literally years since I've even seen one of these things :( you might try shooting Dr. Dave a PM, see if he is still active.

User avatar
Dodoid
Posts: 634
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Sun Aug 27, 2017 3:58 pm

I am looking at the system controller board as we speak. It appears to make its own power from the 48V of the OLS, but also monitors other power rails (I think).

I have recently been in touch with an Ottawan SGI owner named Dave via email, which lines up pretty well with Dr. Dave's info. He gave me an Indigo2, which was quite good as my other one had some issues, as well as an SGI monitor which I now use with my Octane and a SUN machine I haven't gotten around to setting up yet. Perhaps he would know about these Onyx problems, his Nekochan signature says he owned some larger SGIs (though, correct me if I'm wrong as I've only been here for a little over a year, it doesn't look like he ever owned an original Onyx).

thegoldbug, it's the deskside. I'm not sure which part is at fault yet, though I'm thinking maybe system controller since the VCAM doesn't seem to be having any problems and the NAND latches on it never trigger. I will probe around on the one I have for a while before I consider replacing the entire board, but if it seems like a replacement is the best option, I would certainly be interested.
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
pentium
Posts: 4742
Joined: Mon Aug 28, 2006 6:29 pm
Location: Kamloops, BC

Re: Onyx Problems, seemingly power related

Unread postby pentium » Sun Aug 27, 2017 8:21 pm

Is it not possible that the flat dallas chip has caused the system controller to lose its mind? Have you strapped in a new battery yet?
:Crimson: :Onyx: :O2000: :O200: :O200: :PI: :PI: :Indigo: :Indigo: :Indigo: :Octane: :O2: :1600SW: :Indigo2: :Indigo2: :Indigo2IMP: :Indigo2IMP: :Indy: :Indy: :Indy: :Cube:

Image <-------- A very happy forum member.

User avatar
Dodoid
Posts: 634
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Sun Aug 27, 2017 9:05 pm

I wouldn't put it past the Dallas chip, though it would be a very strange thing to get messed up by the chip. One thing I have noticed is that as the system runs for longer (presumably due to heat) the shutdowns happen sooner and sooner after powering on. On the first boot in a while, I can get to IRIX and even use it for a few minutes, whereas after a few attempts it will only last 5-10 seconds. I wonder what components on the board would be affected by heat? Maybe the Dallas chip would be as it contains a battery? I will order a new Dallas chip and see if it fixes it, but I'm skeptical.
: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: 4923
Joined: Thu Jun 17, 2004 11:35 am
Location: Wijchen, The Netherlands
Contact:

Re: Onyx Problems, seemingly power related

Unread postby jan-jaap » Mon Aug 28, 2017 12:15 am

A flat Dallas doesn't cause power errors (mine didn't at least). If the battery on the IO4 is empty also it'll loose the system serial#, and probably complain about changed system config all the time.

If the voltages on the IO4 are fine, my next suspect would be the system controller. Which is indeed on the backplane with the 512 and 505 power modules.
: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: 634
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Wed Aug 30, 2017 5:20 pm

So, I've been looking at the system controller, looking up chips, and probing around with a multimeter for a few days, and I must say, I'm stumped. Where does the actual voltage measurement take place? Not POKA (done by the voltage comparators and NAND latches on the VCAM), but the actual measurement (my error is -12VDC OVER VOLTAGE, not POKA FAIL). Have I missed a chip on the system controller? Is it somehow done by the Motorola MCU directly? I thought maybe it was on the backplane, but the only ICs on there seem to be related to clock generation, not power monitoring, and I can't find anything of the sort on the 512 and 303 boards either.

The System Controller seems to input the voltage somehow, as there is a resistor connected to the -12V pin on the VME P1 connector, and a test pad on the other side of the board connected to the resistor, but I have yet to work out where it goes from there. None of the pins on the chips are directly connected to the test pad (using continuity mode) but maybe it goes through some more passives first? Not sure.
:Onyx: :O2000: :Fuel: :Octane: :Octane: :Octane: :O2: :O2: :Indigo2: :Indigo2: :Indy: :Indy:
and a small army of Image

User avatar
Dodoid
Posts: 634
Joined: Mon Jul 04, 2016 1:36 pm
Location: Ottawa, Canada
Contact:

Re: Onyx Problems, seemingly power related

Unread postby Dodoid » Fri Sep 01, 2017 6:33 pm

So I got the new Dallas chip from Digi-Key and swapped it in. As expected, no change to the problem, but good to have. Does anyone know where on the System Controller voltage monitoring takes place? I could multimeter probe it and see what it's doing if I knew. Thanks!
: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 1 guest