1

I'm hoping that someone has real-world experience starting up an Intel Skulltrail D5400XS system, with less than the Intel-recommended 1,000 watt power supply. And/or, real-world experience with multiple blown power supplies. (Because with 30 years of PC building experience, I'm not lacking for opinions, theories, or reading manuals ☺️. But I've only had one PSU blow up before - and catch fire...the system still worked with a new PSU.)

There's usually no limit to opinions on what one "should" do relative to manufacturer's recommendations. I'm not asking what one "should" do, just trying to better estimate the risk/reward of buying a new, expensive 1kw PSU. Because if it seems the odds are high that the board is fried, that's just throwing good money after bad, and I have no other use for a 1kw PSU. But if the odds look more favorable, at some point the gamble becomes worth it. This is not my primary working machine, and nothing valuable is at risk other than 12 year-old hardware. However there's some emotional attachment, as I've fiddled with it endlessly, including upgrades to 32gb RAM, the fastest supported CPUs, GPUs, etc.

At full synthetic load to all CPU cores, and GPU - and with 20 HDDs spinning - this system tops out at 450 watt draw from the wall. (Much less now without the HDDs.)

Here's the quick backstory on what happened: I've had this system for 12 years. It ran like a champ in a 4U rackmount server chassis. I recently downsized the system to a smaller desktop EATX case and just a single boot SSD. Unbeknownst to me at the time, sometime during the last 12 years, the fan in the 1,200 watt PSU died. The only reason the system ever worked at all, is because one of the massive CPU cooler fans was doing the job for the PSU fan, an inch away and directly over the PSU fan intake. (Yes, a terrible case layout for this particular board.)

(That the PSU fan hadn't been working for some time, only became apparent during postmortem teardown.)

The new case locates the PSU on the opposite end of the board (at the bottom of a tower), and isolated in terms of airflow. While burning the system in to make sure the transplant was successful, the PSU (apparently) overheated. A very loud "POP!" was heard, and the system was down.

Like I mentioned, I've had something similar happened before with a different system, but the system was fine with a new PSU. So, hoping this was also the case now, I tried resurrecting the system with a new 750 watt PSU I have on hand.

The "power standby" indicator light comes on, and the LAN LEDs blink briefly, but that's as far as it gets. The POST LED readout doesn't turn on, no fans, nothing else happens. Although, one time - just once - the southbridge fan and other LEDs briefly, for maybe half a second, turned on. I could never get that to repeat. But gives me some hope.

So I'm trying to figure out if this board is fried, or if it just needs more juice. Like I said before, the maximum I've ever seen this system pull from the wall is 450 sustained watts.

Also, if it matters - this PSU only came with one CPU power cable, but does have a modular connecting port specifically labeled for a second one. While installing the PSU I also installed a second CPU power cable from Amazon.

I tried booting with all peripherals removed (GPU and SATA cards), to reduce power draw. Didn't help. (It will normally power up without those, just won't get past POST.)

There is no visible damage around the top of the board.

Thanks for any experienced thoughts!

Improve this question
2
  • 2
    This is either a problem due to the PSU CPU cables and the fact you only have one, or the motherboard is simply dead, but I would expect if the motherboard wasn’t dead that the machine would at least attempt to POST. My speculation would make a horrible answer
    – Ramhound
    Feb 7, 2020 at 2:19
  • Thanks for the thoughts. It actually has both CPU power cables, one of which I had to order separately. I'll update the question to make that clear.
    – Jim
    Feb 7, 2020 at 3:36

1 Answer 1

Reset to default
0

Solution: The motherboard was dead. (And also the PSU.) A new PSU failed to revive it. Not a 750-watt (below mfg recommendation), and not 1200 (above).

This happened before to a different machine, many years ago - it actually caught fire. But everything except for the PSU was fine, and a new PSU revived it. That's why I held out hope the same might be true this time.

This time, however, no such luck. What is unknown is whether the CPUs (fastest this mobo will run) and/or RAM (32gb - beyond 16gb spec) are damaged. The only way to find out is to buy a new motherboard.

The only problem is, it was $600 new 12 years ago, and they are currently nonexistent on ebay. I've found some workstation and server mobos with similar chipsets and specs, but even used they are still very pricey.

I could literally build a low-end Ryzen system (with existing case, fans, and PSU) that would be faster - than just the cost of a replacement mobo.

Which is a shame, because this computer rocked for 12 years, and even today comes shockingly close to modern desktops in terms of benchmark performance. I had really grown fond of it.

But no sense in throwing good money after bad.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .