New build - initial boot problems

[Kytrid]

My build:
ASRock Fatal1ty Z170 Gaming-ITX/ac Mini ITX LGA1151 Motherboard
Intel Core i5-6600K 3.5GHz Quad-Core Processor
Corsair H80i GT 70.7 CFM Liquid CPU Cooler
Crucial Ballistix Elite 16GB (2 x 8GB) DDR4-2666 Memory
Mushkin ECO2 512GB 2.5" Solid State Drive x 2
EVGA GeForce GTX 970 4GB FTW ACX 2.0 Video Card
Silverstone Sugo SG13B Mini ITX Tower Case
Silverstone Strider Gold 450W 80+ Gold Certified Fully-Modular SFX Power Supply
Noctua NT-H1 thermal paste

This is a new build. It has never successfully booted.

My actual build went extremely well, no problems at all. Until initial boot, that is.

When the power switch is depressed, all fans (cpu cooler and GPU fans) spin up as expected. Then the EVGA video card fans stop spinning after 15-20 seconds (My assumption is that there is no load, therefore no cooling is needed). The cpu cooler fan runs continuously as long as there is current running through the system. The video card becomes warm to the touch, as though under a light load. The LEDs on the cooler and in the case also light up. Nothing else happens. Any monitor that is attached fails to sense a data stream through the cable. I have a chassis speaker hooked up to the chassis speaker out, but have never managed to get a sound from it - this motherboard does not have a digital post code display. No matter what configuration I have tried, the fans have spun up in exactly the same way, and the lights light up. Otherwise, nothing.

When I first put the computer together, it was with video card, both sticks of RAM, and SSDs. This was my first failed boot configuration.

My diagnostic process:

I stripped the system to bare bones – CPU, mobo, CPU cooler, PSU, one stick of ram. The video is running through one of the mobo’s native HDMI outs. No change. No post codes, no monitor activity.

All psu connections have been checked, rechecked, re-re-checked, etc. Video card and RAM are always well-seated.

I have switched out the one ram stick with the loose one. No change.

Switched video cable to the second HDMI slot on the back of the mobo. Nothing.

I switched out the new monitor with a known monitor, including all cabling. No change.

I put the EVGA video card back in, hooked it up via HDMI. No change.

Then I replaced the HDMI cable with the normal computer DVI cable on the EVGA card. Tried it in both plugs on the back of the video card. No change.

Finally, I pulled a PSU that is known to be fully functional from another build, and hooked it into the system. Once again, the exact same thing occurs.

I have also reset the CMOS via the button on the back of the motherboard.

Also, pulled the build out of the case and tried it, just to make sure I wasn’t having some sort of short to the case. No change.

NOW – here’s the zinger – I couldn’t come up with any further diagnostic tests, so I decided to replace the motherboard. Upon putting the new mobo (same model) in the system, I get EXACTLY the same results. I can see one of two possibilities at this point: Either I have a bad CPU, or both sticks of RAM are bad, or I’ve somehow missed something.  I’ve never managed to get a post code out of either board.

This mobo model has no mini-PCI-e slot, or I’d hook in a diagnostic board to try to get a digital post code. It only has one full PCI-e slot and two slots for RAM. No others. Well, not counting the M.2 on the bottom.

Any ideas? And thanks in advance for any help you can provide.

[Lisa_maree]

When you had the motherboard running out of the case were you still using the same connections to the front panel?
I would try the motherboard, cpu and memory out of the case and use something metal to touch the power switch pins and see if you have a display.
I have found people in the past have put one of the Leds across the reset pins which can keep the motherboard in a reset state. That the reset switch is stuck doing the same thing.

[SOAP]

I had something similar happen on my build. It turned out i had attached 2 wires on my mother board in reverse positions. Try another stick of ram if you have one. Could just be a bad cpu... So hard to tell in this case.

[Kytrid]

Thanks for the replies. I tried your idea - first I disconnected all front panel connections, so the header was completely free. Then I manually initiated contact across the power switch pins. The result was exactly what I've had with both mobos - fans spin up, I can hear the cpu cooler pump humming (which I forgot to note in my OP), and cooler/front panel LEDs light up. Releasing the contact powers it all down (of course). When the header was/is wired to the front panel in the cases of both mobos, I could/can turn the system off every time by holding down the power switch until it powers down. Pushing it again causes the power to come back on - it seems the on/off switch is the one of the only things that is actually functioning correctly. The front panel leads are all connected in accordance with the mobo installation manual pin-out.

[Lisa_maree]

Well i have checked every part of your design and it is all compatible and should work. You seem to know a lot about putting together systems so I haven't been suggesting  things like the 6 pin power to the motherboard being disconnected. Perhaps if you could take some pictures of the motherboard connected up and post them it might help.

 

[Kytrid]

Thanks for double checking the build, Lisa_maree. I should've just posted my pcpartpicker build - would've made it easier for you. The 24-pin and EPS 8-pin connectors are socketed flush with their plugs and clicked in tight. Besides those, there just aren't many more connections that could be causing the issue, especially since disconnecting the SSDs and the video card didn't change the outcome. At this point, the only front panel connection is the power switch lead, which works fine to turn the machine on and off.

I've just ordered some different RAM in hopes that there is something about the pair that I have that my mobo hates - I don't have any extra DDR4 RAM laying around for trying-out purposes. Typically when there's no post code output, it means that something has stopped the boot sequence even before the POST can get started - in my experience it's often been a problem with RAM. In this case, it's weird b/c switching out the RAM sticks didn't solve the problem. It's hard to imagine that I got two bad sticks in one package - it's a decent brand (though I'm a G.Skill fan, personally.) We'll see if replacement RAM fixes it; if not, I guess I'll replace the CPU. Or do something else magical that you guys can think up -- I'm all out of ideas, so I guess it's time to throw money at it.

My first build was a IBM PC XT (8088-based) clone way back yonder in about 1985 or so (IIRC). Usually builds don't snocker me like this one has, but everyone needs an eye-opener now and again, right?

I took a couple of pics, but they were most uninspiring -- here's one -- not much to see since it's stripped to bare bones.

Thanks again.

[attachment deleted by admin to conserve space]

[Lisa_maree]

I wonder if you could show a picture of the power supply label . Also with the  Video card out of the machine it appears the 12v pcie power is still connected. Could you try it with just power to the motherboard please.
Also check the back panel for anything shorting any of the ports , like an earth strip caught inside a USB port.     

[Kytrid]

Update here - After my last post I took some time to get away from the build, then reassembled the entire thing outside the case. Still no boot when the power button was depressed, just fans, pumps, etc. The new ASRock Fatal1ty Z170 Gaming-ITX/ac replacement I ordered acted exactly the same way as the first. So I ordered a new cpu. New cpu made no difference, same fans/pumps. I then bought an MSI Z170I Gaming Pro AC mobo and reassembled the build outside the case using new mobo and new cpu. All other components were the same as previous builds. IT BOOTED immediately.

I then disassembled and rebuilt the system using the new (good) CPU but with the replacement ASRock board that I still had in hand. Would not boot, and acted exactly as the other (first) ASRock board that I had received (fans/pumps only) --- This mobo is DOA. I then reassembled the system around the MSI board (all other same components), but dropped in the first  i5-6600K I had received. The MSI board has diagnostic LEDs, unlike the ASRock, and the bright red CPU LED came on and stayed on. The system refused to boot. CPU is DOA. Crazy, I know. Definitely a first for me - I've never had an Intel CPU do this. Burned up one once upon a time, purely PEBCAK, but that's another story.

So, my conclusions are that the second ASRock board and the first CPU were both definitely DOA. What are the chances of that? There's really no way to know about the first ASRock board, b/c I sent it back before I had a chance to test it with a second known-to-be-good CPU. This build has challenged me like no other I've ever taken on. As it stands, I finally have a working system that's built around the MSI board. Well, it's working outside the case, that is. We'll see what happens when it gets rebuilt in the case. I'll drop a note here to let you know how that works out.

Thanks for you suggestions along the way. I hope my process as posted here may help someone here in the future.


 

[patio]

I'd spend more...and stay away from both ASROCK...and MSI personally...

[Kytrid]

I'm a long-time fan of both Gigabyte and Asus, but ASRock and MSI have both put a great deal of blood, sweat, and R&D into their enthusiast/gaming boards over the past 5-6 years-- I think that, while your reservations may be spot-on for those companies in the past, their products have come a long way.

I appreciate a conservative viewpoint, and often tend towards one myself, but in this case, I think you're painting with a very broad brush.

As a matter of fact, the only motherboards that I've ever consistently had problems with through the years happen to have all been Gigabyte boards. Even so, I have three computers at the moment that are built on GB mobos, and still think highly of GB.

And, as a final point, I'd say that $189 isn't exactly a "low-priced" board, in the case of the MSI board that my system is now built on. The components that are soldered in it are as high quality as those found on any Asus board out there. If you don't believe me, dig into the spec sheets and you'll see that both companies are using decent components in their performance boards these days. This MSI board (and the ASRock board as well, as a matter of fact) is far more thoughtfully designed than the last Asus board that I used, which, among other things, had all the SATA ports pointing toward the front of the case instead of how a normal rational being would put them, namely facing the side of the case so you could plug something into them without having to remove the freaking board from the case....

I find it interesting that you didn't suggest staying away from Intel CPUs even though that was half of my problem with troubleshooting this build. =)

At any rate, no harm in having opinions. Everyone has them, along with sundry other undesirable and smelly bits.

[Lisa_maree]

Hi

In years of building 100's of PC's we have never had a DOA CPU or Motherboard. The automatic testing that is done at every stage of assemble insures that no product leaves the factory faulty whether Intel or AMD or MSI or Gigabyte or asus not sure about Asrock. They all test every board and every component on the board visually as well as electronically. So why do some people  get faulty boards and CPU's . Well they were not faulty when they left the factory so perhaps it is the way they are handled  when being installed.
Good that you now have a going setup. Feel sorry for the supplier when they get all the faulty parts back.

[Calum]

I haven't been following this thread but just wanted to comment on a few things...

In years of building 100's of PC's we have never had a DOA CPU or Motherboard.

You've been pretty lucky on the board front then...I've only ever had one dead Intel CPU, a couple of dead AMDs, but lots of DOA boards.

The automatic testing that is done at every stage of assemble insures that no product leaves the factory faulty whether Intel or AMD or MSI or Gigabyte or asus not sure about Asrock.

Not so sure about this.  Even if it has a QC pass sticker on it or a certificate of testing, that doesn't necessarily mean it's going to work out of the box.

far more thoughtfully designed than the last Asus board that I used, which, among other things, had all the SATA ports pointing toward the front of the case instead of how a normal rational being would put them, namely facing the side of the case so you could plug something into them without having to remove the freaking board from the case....

Not sure why you'd want the SATA ports at 90 degrees to the board (if I'm reading this right) - there's potential for them to block expansion cards and it makes cable management much more difficult.  I guess it depends on your case to a certain extent but I've never felt the need to remove a board from a case to insert SATA cables into a board with "front-facing" or right angled ports, whatever it's easiest to call them.

[patio]

I completely dis-agree on the testing statements made...i don't build hundreds...but at least 3 to 6 a month...
And no way is your failure rate quoted accurate....
No matter which manuf.

[Lisa_maree]

I have tried to find internet sites listing failure rates of parts . All seem to be component suppliers as well as system builders and list RMA percentages, they don't seem to give details of Dead on arrival parts which is different. Most RMA parts have been installed and used. Some are damaged by the assembler and some by the customer. From the beginning we had a policy of not selling components for DIY install so perhaps that is why our DOA rate is 0%.

[patio]

Fascinating...

Did you think a manuf. would post their DOA rates on a website for public consumption ? ?

OK.

BTW...who do you work for ?