1. Forum
  2. Usenet
  3. COMP.SYS.SGI.HARDWARE

  • Is there a way to bench test Octane power supply?

    From Steve Allen@21:1/5 to All on Mon Dec 9 15:57:32 2019
    Hi. Doesn't look like these groups get used much any more, but I thought I'd try...

    I have an Octane (original) that worked when I put it out in my shop years ago. Sitting on a concrete floor doesn't seem to have been good for it -- it's a bit discolored on the metal surfaces. When I plug it it, I hear a relay click and the whine of a
    switching regulator, but pressing the power button has no result.

    Before I start getting parts to swap things out, I'd like to know, is there a way to bench test the power supply? I have disassembled the power supply and see nothing wrong; the fuses are all good. If I can trigger the power supply on, I can at least
    eliminate that as the problem and start looking elsewhere.

    Thanks,
    Steve

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Dave McGuire@21:1/5 to Steve Allen on Fri Dec 13 08:16:48 2019
    On 12/9/19 6:57 PM, Steve Allen wrote:
    Hi. Doesn't look like these groups get used much any more, but I thought I'd try...

    Nonsense. This particular one is very quiet, but Usenet in general is
    bigger than I've ever seen it. (speaking as a former news admin at a
    1990s ISP)

    I have an Octane (original) that worked when I put it out in my shop years ago. Sitting on a concrete floor doesn't seem to have been good for it -- it's a bit discolored on the metal surfaces. When I plug it it, I hear a relay click and the whine of
    a switching regulator, but pressing the power button has no result.

    Before I start getting parts to swap things out, I'd like to know, is there a way to bench test the power supply? I have disassembled the power supply and see nothing wrong; the fuses are all good. If I can trigger the power supply on, I can at least
    eliminate that as the problem and start looking elsewhere.

    Concrete is porous, and moisture is a problem.

    I believe the Octane power supplies are soft-controlled, so you'd need
    to fake the power switch, at least, but you sound like the type of
    person who can handle that easily. Some early switching regulators
    don't like running without a load, but the Octane is well after that
    era, so that should be ok.

    The first thing I'd look for is mold or other gunk on the PCBs, then
    possibly swollen capacitors. If you strike out with those, you're in
    "anything goes" territory.

    -Dave

    --
    Dave McGuire, AK4HZ
    New Kensington, PA

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Steve Allen@21:1/5 to Dave McGuire on Sat Dec 14 17:44:39 2019
    On Friday, December 13, 2019 at 6:05:02 AM UTC-8, Dave McGuire wrote:
    On 12/9/19 6:57 PM, Steve Allen wrote:
    Hi. Doesn't look like these groups get used much any more, but I thought I'd try...

    Nonsense. This particular one is very quiet, but Usenet in general is bigger than I've ever seen it. (speaking as a former news admin at a
    1990s ISP)

    Yeah, I was talking about the SGi groups in particular. Most of them haven't had
    posts in months, except for spam.

    I have an Octane (original) that worked when I put it out in my shop years ago. Sitting on a concrete floor doesn't seem to have been good for it -- it's a bit discolored on the metal surfaces. When I plug it it, > Concrete is porous, and
    moisture is a problem.

    I didn't really think of that when putting it in the shop. I have lots of shelving,
    but the few bigger computers were put on the floor. Thinking about it now, I need to get my Mac G5 off the floor, too.

    I believe the Octane power supplies are soft-controlled, so you'd need
    to fake the power switch, at least, but you sound like the type of
    person who can handle that easily. Some early switching regulators
    don't like running without a load, but the Octane is well after that
    era, so that should be ok.

    I guess I'll find out. ;-)

    The first thing I'd look for is mold or other gunk on the PCBs, then possibly swollen capacitors. If you strike out with those, you're in "anything goes" territory.

    That's what I've done. The PCBs look clean, and the caps all look OK. I was just hoping somebody had some pinouts for the power supply connectors so
    I could see what I need to trigger. I have traced one side of the power switch to the power supply, but can't find what the other side connects to. So I don't
    know if I need to hit the PS with a ground on that pin, or a voltage of some sort. (I would expect the supplied voltage to be derived from the switching regulator that I hear running, somehow.)

    Thanks,
    Steve

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)