Fuels: a writeup on the design flaws
#11
RE: Fuels: a writeup on the design flaws
That isn't really unique to SGI systems though. PSUs by their nature get quite warm and have big, electrolytic caps in them which dry out over time. Many PSUs will struggle after 15+ years of hammer and SGI supplies aren't unique here. The main issue with the Fuel PSU is that SGI could have used a standard PC-type PSU to make replacement simple, but they just had to fiddle with it to make it proprietary. Thankfully Kuba's adaptor allows a stock PC PSU to be used, assuming it can deliver the necessary juice on the right rails.

Indigo2 R10000/IMPACT  R10000 195MHz, 384MB RAM, MaxIMPACT (1MB), 36GB 15k & 300GB 10k drives, new/quiet fans, 100Mb G160 NIC, IRIX 6.5.22&
[Image: Fuelb.png] R14000 600MHz, 4GB RAM, V10 Graphics, 72GB 15k & 300GB 10k drives, new/quiet fans, 1Gb NIC, IRIX 6.5.30
O2  in storage...
Trippynet
Indigo2 IMPACT

Trade Count: (0)
Posts: 304
Threads: 7
Joined: Dec 2017
Find Reply
06-01-2020, 12:23 PM
#12
RE: Fuels: a writeup on the design flaws
(06-01-2020, 11:39 AM)jan-jaap Wrote:  IIRC there was simply a "bad year" in the production of the DS1780 chips. There have been quite a few people who replaced DS1780 chips, I never heard anyone come back with an "OMG they burned out again and this time it took out the entire system" or similar.

The problem with the Dallas chips doesn't just affect the Fuel, the early O3000 bricks had the same problems. There just aren't many people operating an O3K from their basement.

It's always the systems with the over-complicated power systems that give you the biggest headache. Ask anyone with a Challenge or Onyx.

That's also possible. It was simply a theory that we came up with. I wouldn't have a problem with Fuels if they didn't share any parts commonality with any other SGI. If their parts go scarce, so what for me, but if they burn up other supplies that other machines without the flaw can use, well it's a problem for a whole lot more than just Fuel users. I don't think we'll ever be able to replicate VPro cards, sadly, so we don't have a whole lot of options.

I'm the system admin of this site. Private security technician, licensed locksmith, hack of a c developer and vintage computer enthusiast. 

https://contrib.irixnet.org/raion/ -- contributions and pieces that I'm working on currently. 

https://codeberg.org/SolusRaion -- Code repos I control

Technical problems should be sent my way.
Raion
Chief IRIX Officer

Trade Count: (9)
Posts: 4,240
Threads: 533
Joined: Nov 2017
Location: Eastern Virginia
Website Find Reply
06-01-2020, 10:19 PM
#13
RE: Fuels: a writeup on the design flaws
I do hope there are still some "dead" VPro examples in people's closets. Some day we might get access to better troubleshooting equipment (my fingers are crossed on a contact I made last week for equipment that might become available for sale in December), and when the call goes out for dead samples, I hope there are some to get. Only comparison between known good and dead boards it going to yield any clue on stopping (reversing?) this occurrence. In other words, what fails and if it's not the main ICs, is it actually fixable?

So I guess I'm reminding people that dead parts have value too, please don't throw them away, advertise them on the marketplace to find them another home until they can be studied.
weblacky
I play an SGI Doctor, on daytime TV.

Trade Count: (10)
Posts: 1,716
Threads: 88
Joined: Jan 2019
Location: Seattle, WA
Find Reply
06-02-2020, 06:50 AM
#14
RE: Fuels: a writeup on the design flaws
Absolutely. That's the point for sure. The big fear I have is that Fuel users who ignore instructions just continue destroying parts because "muh Fuel". I've seen at least one case where someone bought two VPro cards after their new card died, and I told him to stop what he was doing and calm down, repair his motherboard's broken chip. He ignored my instructions.

I'm the system admin of this site. Private security technician, licensed locksmith, hack of a c developer and vintage computer enthusiast. 

https://contrib.irixnet.org/raion/ -- contributions and pieces that I'm working on currently. 

https://codeberg.org/SolusRaion -- Code repos I control

Technical problems should be sent my way.
Raion
Chief IRIX Officer

Trade Count: (9)
Posts: 4,240
Threads: 533
Joined: Nov 2017
Location: Eastern Virginia
Website Find Reply
06-02-2020, 07:14 AM
#15
RE: Fuels: a writeup on the design flaws
Hey guys,

I'm glad there are still discussions about Fuels, very close to my heart Wink

The case is based on Palo-Alto PA 810 (https://manualzz.com/doc/31533214/pa-810...•-supports), you can clearly see the resemblance, and it seems that the only inside modification is the swing-out drive bay.

As for failures, I'll add a couple more on top of PSU and DS1780 chips (I'm actually in process of replacing chips on one of the boards).

*) I've had Ethernet MAC EEPROM literally fall off - the chip was in the socket, but the socket somehow didn't stay in the motherboard

*) D-SUB9 socket on the motherboard (near SCSI) is not attached properly, it's only held up by the 9 pins, but nothing more even though the plastic is rather large and there's more than enough space to affix it to the motherboard. Any vertical movement or a heavy RS-232 cable and you risk damage (one of my fuels) literally ripped it out so I now have to use external USB-based L1.

*) Flashing PROM/L1, I bricked one Fuel because of unstable V12 that I had. Somehow I decided to flash PROM (hoping to fix the V12 issues) and it died on me in the middle of flashing.

*) PIMM issues, I haven't personally had any, but I've heard of many broken PIMMS, I would risk saying those may also be aftermath of dying PSU's. The PIMM has a few voltage regulators onboard, not surprising.


Cheers
kubatyszko
Tezro/Fuel/Indy/Octane2/Indigo[12]/O2

Trade Count: (0)
Posts: 72
Threads: 16
Joined: Oct 2018
Location: Los Angeles, CA
Website Find Reply
06-03-2020, 08:32 PM
#16
RE: Fuels: a writeup on the design flaws
I've always wanted to like the fuel I just have had my heartbroken by so many of them breaking on me.

I had a top-of-the-line V12 DCD 900MHz Fuel that became unusable. I paid over $1,000 for it and I was not very pleased. I ended up parting it out. The V12 was fine but the board was toast.

I'm the system admin of this site. Private security technician, licensed locksmith, hack of a c developer and vintage computer enthusiast. 

https://contrib.irixnet.org/raion/ -- contributions and pieces that I'm working on currently. 

https://codeberg.org/SolusRaion -- Code repos I control

Technical problems should be sent my way.
Raion
Chief IRIX Officer

Trade Count: (9)
Posts: 4,240
Threads: 533
Joined: Nov 2017
Location: Eastern Virginia
Website Find Reply
06-03-2020, 08:42 PM


Forum Jump:


Users browsing this thread: 1 Guest(s)