Aurora-R4, won't turn on

I've reseated all the power cables and tried turning the PC on without certain things; the graphics card, ram and water cooler. The result is the same.

This is the orange light above the power cable to the master I/o board

The MI/O board orange LED 'on' means the board has power. It should turn off when the system turns off. It should turn on for as long as the power supply is on. If the LED is always on, (on, even though power supply is 'off'), that would not be a good sign. So does it simply 'turn on for a moment after pressing power button, then turns off when power supply inexplicably shuts off'

I've never had to test a system using the power supply rear self-test button, I'm sure the procedure can be looked up online, a uTube vid exists actually (see uTube > No Power? Learn about Dell's Built In Self Test for power)

So, you swapped in a new CPU cooler, powered up, it worked, system ran fine, you powered down, next time you went to turn it on, the PC won't start, upon pressing the power button the fans spin for about a second and the whole thing powers Down again, pressing the power button again yields the same results over and over

Test the motherboard's (round coin-cell) 3volt battery for ~2.65v - 3volts, or simply replace it. A low battery can render a motherboard 'dead'

Google search of "power supply shuts off"

Having a spare power supply can help, example, eBay search = alienware 875. J556T W299G (w/NRHJ9 harness) ~$40 shipped

I suppose if a power supply has 'short circuit protection', it would either fail to turn on or immediately turn off. There comes a point where we might remove motherboard, and bench test it on a table to rule that out ...

An even tighter review of the wires / cables / connectors would be to make sure none of the pin terminals have backed out of any connectors (ie connector fully seated but no connection present on a crucial pin)

The 'paperclip test' can force the power supply on: if it turns on, fan spins, it can either indicate a good power supply, or, a power supply willing to turn on but that would fail under any kind of load. Paper clip test, psu turns on is a good sign but not a final verdict it's still useable

Make sure the 66pin black bulk connector is fully seated into PSU

I suppose a bad CPU might cause instant system turn off, but they rarely go bad. Another scenario is CPU was removed / replaced and one or more socket pins were bent

Motherboard may be faulty, not due to cooler swap, just bad timing, or maybe you bumped something during the swap without noticing

Likely scenario: you keep fiddling, remove / replace items, connect / disconnect, system turns on as normal, all is well & we never know the why or the what

Options:

a > swap in new PSU

b > remove motherboard / psu, do a bench test

c > take system to someone nearby for diagnosis / repair

--> Link: Alienware Desktop - General Hardware Troubleshooting

photo of my Area-51 mthrbrd bench test: https://i.postimg.cc/KYF6ZTVm/xdj4c_bench.jpg

It occurs to me, that I've come across a pair of bad 'power on' assemblies (pcb / power on button) for the Area-51, Aurora may be similar 

Symptom: turn pc on > pc shuts off after ~5-10 seconds later

Cure: replace top power on button PCB board

Test: Normally, we might by-pass the top power on PCB board as a test, by:

A) With system fully connected

B) disconnect motherboard front panel (5x5 / Dupont) connector (photo) B) use flat blade tool (small screw driver)

C) short / touch pins 6 & 8 together ('jump start PSU on')

D) if system turns on & stays on?

E) Monitor CPU temps & shut down if idle temp exceeds 50c, if yes then shutdown PC, replace top power on pcb board, then re-test

search eBay item 'MP-00004839-000 Alienware ALX Aurora R2 Front I/O Panel Bracket Cable Assembly' --> this $13 part should have the power on PCB board *I can not say if this one or any other will arrive working or not*

The top power on pcb board becoming faulty is probably rare, but the steps above (by-pass it) may diagnose a bad board

see uTube: 'jump start motherboard on' (procedure is identical to bench testing motherboard)

_______________________________

below, remove JFP1, momentarily 'short' / touch pins 6/8 together until motherboard / psu turn on (then back tool away); observe system (stays on / turns off)

if system turns off right away, problem may be elsewhere (top pcb may still be good)

https://i.postimg.cc/CLFHgTjB/PTDC0015.jpg

Hi thanks for the advice! 

I've fully completed a bench test of all the components and even went to the trouble of using a volt meter to test the coin cell battery (it's fine giving out 2.5-3v) and the power cables all of which are working fine. Every component put together works. 

I saw your comments about a faulty power button and whilst doing the bench test actually plugged in the front panel cables (ones that control the power on button) it's a nine pin connector and of everything it was the only thing that didn't work when the power was connected. Now I've just connected the motherboard and the master I/o board with all relevant cables attached to soley test the front panel, everything comes on as it should but only for about a second then it all powers down. Using the black button on the back of the power supply turns it on which leads me to believe the fault lies with the front panel cables or general set up. 

I'll get a new one shipped as it's generally inexpensive so even if it isn't the problem I won't waste that much money. I'm also aware that the Aurora comes with a safety feature that doesn't let you boot the computer with the side panel missing but I've disconnected it so that wouldn't be the issue. 

I'll try everything you've further suggested and get back to you. Thanks for your help so far I'll keep this thread updated until I find the solution.

Joe

EDIT: I've opened up the front panel and retrieved the cables to take a closer look, there's a cable that requires a 20 pin socket either side and I can't find the end it's meant to connect to, I assume this is a power cable for the whole thing but I can't find where it's meant to go. Also, the end that connects to the small board that controls the USB ports and power button, the port for that has a pin missing and one bent out of shape I assume causing my current issues , would you agree ?

My R4, an old pic, taken when powered up w/side panel off, done it many times, but that isn't something I wish to affirm or de-bunk. Try panel on & off, see what happens

https://i.postimg.cc/Njc7vJrP/PTDC0002.jpg

A bench test involves removing 'all' components + the case harness, staging them etc, I'm impressed you were able to accomplish that. Desperate times / measures etc. I take it you 'jumped' front panel header pins 6/8 together momentarily, but system still turned on then off just as it was when PC was 'whole', earlier in the day. If so, yeah, doesn't sound good ... I was hoping it was the coin-cell batt

I've an x79 AsRock motherboard here with its own 3820 CPU, so were it me I'd yank the 4820 out of my Alienboard, pop in the 3820 (transfer 4820 to my other x79 board), for testing purposes, to see if CPU + motherboard work. Test x79 boiard here with the PSU of course. Spare parts. People who do not have spare parts either seek out pro help, or, invest in a new PSU, then new mthrbrd 1st (+ hope CPU is ok). If new PSU, then mthrbrd fail, invest in new CPU. It's an ugly picture, in other words, but also tough to pinpoint the exact failure from afar. I don't know the exact part you need or where to go from here, short of calling in spare parts or seeking pro help, ok?

No need to answer this, but if your CPU was removed for whatever reason recently, now's a good time to check if any pins were bent on re-install

I suppose you can check the computer case, see if anything fell off the mthrbrd (plastic shunts, capacitors, anything identifiable)

--> Is it wise to invest in an NRHJ9 case harness? Dunno ... if all else fails, then yes

I'd like to be wrong, maybe I am, but after a bench test, likely remaining culprits sound like PSU, mthrbrd, CPU, in that order. Swapping in spare parts one-by-one to get at the fix gets expensive, I know that. Parts that were purchased but un-needed can be re-sold or stored for rainy days. A 3820 fetches $50, the 7JNH0 or FPV4P mthrbrd can fetch $100-200 or more. My remote diagnosis here represents my best guess. Chances are low, but still a chance it may be a ridiculously easy fix no one thought of, unfortunately. Let us know what you figured out 

In the news this week (Aurora-R4, retail LGA 1151 motherboard, i7-8700k CPU), someone's swapped in CovfefeLake + EVGA mthrbrd ... new Z390 + 9700k/9900k debuts in October, also --> wait for other member input as well, & Good luck

Your edit: I've opened up the front panel and retrieved the cables to take a closer look, there's a cable that requires a 20 pin socket either side and I can't find the end it's meant to connect to, I assume this is a power cable for the whole thing but I can't find where it's meant to go. Also, the end that connects to the small board that controls the USB ports and power button, the port for that has a pin missing and one bent out of shape I assume causing my current issues , would you agree ?

I have a spare 5W6MX, what we tend to call the front Top I/O assembly, which houses the 'front panel' parts (on/off, LEDs etc). These are the connections, labeled, is this what you opened up?

There are three Dupont (5x5) connectors, the 20pin (10x10) is USB3.0, small white JST is to MI/O

I use postimages.org for image hosting here, you're welcome to use my pic or load up your own to discuss this. All four connector types above plug into their motherboard headers (#5 to MI/O) 

If you removed the motherboard from the case, 'bench tested it', jump-started the motherboard on by shorting front panel header pins 6/8 w/a flat blade (any metal) tool, & it still turned on then off > you effectively by-passed the Top I/O, therefore anything wrong with it (missing / bent pins etc) I wouldn't, or can't ascribe as the culprit for shutdown issue. Is the top I/O the basis for your edit? If you isolated motherboard from the case, which is what you wanted to do, then you'd rule out the top I/O if the issue persisted on your bench / table ... (bench-testing outside of the case also isolates the MI/O board & its related wiring, so we can rule those out also)

Let me know where your thoughts are at

Hi!

So this is where I'm at right now , I've taken every part out and bench tested it, by every part I mean everything even the front panel assembly (after closer inspection of the wire I referenced one end was capped off leading me to believe the cable wasn't in use, my bad) 

There seems to be an issue somewhere between pressing the power button and the PC starting, either the no post issue or a no boot issue which causes the whole computer to just turn off. The fact all Individual parts work when using the black button on the back of the PSU leads me to believe it's one of the two issues I mentioned, both of which are to do with my mobo arnt they? As that button would supply power but not boot the operating system in any way. Please correct me if I'm wrong.

I told my friend every issue with it and what I'd done so far and he recommended jump starting it he walked me through how to and nothing happened but I was doing it while it was being bench tested so I'm guessing I'd have to reassemble the whole thing and try jump start it again for it to boot? Again correct me if I'm wrong. He said the mobo was dead and that I should replace it so I'll summarise everything I've done and their results so far so you can give me a second and third opinion. 

1) computer turns on for 1-2 seconds spins all the fans and turns off

2) checked all power cables were correctly connected (they were) 

3) bench tested individual parts (everything works when using the black button on the back of the PSU) 

4) addressed a potential issue with the power button setup (my bad all was fine again) 

Current situation: still won't turn on potential no boot/ no post issue. I haven't had the time to take parts out and test them in other systems but im getting round to it, for now give me more suggestions / agree with my friend 

Big thank you for all your help !

Joe

"As that (black) button would supply power but not boot the operating system in any way." --> I've watched the Dell power supply self-test video a few times over the years, I no longer use Dell power supplies & have never run the self-test. From the looks of it, the rear button doesn't turn the system on, it passes a current through system, Green or white LED suggests system or PSU is ok, at least at a surface level (ie, PSU or system still fail after pressing power on button). I could envision a PSU that lights green on rear, but, a board component fails on power up; I could envision the mthrbrd spits out an unacceptable 'power ok' signal also (grey, pin 8), causing psu to shut down (mthrbrd fails P.O.S.T., power on self test, power not ok). In short, rear black button should not turn system on, unless yours does ...

"I'd have to reassemble the whole thing and try jump start it again for it to boot". No. Or, yes. You can put everything back in the case, disconnect the front panel connector, 'jump-start' mthrbrd on through pins 6/8. I posted a photo of my mthrbrd installed in the case, with photo of front panel header / connector for that reason, just back yours off & short pins 6/8, mthrbrd should turn on. Or, you can take it all out of the case & do the same, ie a bench-test. See uTube: 'turn motherboard on without power button', same deal, short pins 6/8. Jump-starting it on (pins 6/8) while inside the case is to rule out a bad power on button pcb, while jump-starting it on in a bench test is because the power on button isn't normally taken out of the case, hence the need for a metal tool to jump-start it on, lol

This intrigues me, #3

3) bench tested individual parts (everything works when using the black button on the back of the PSU) I posted a photo of my Area-51 mthrbrd + all necessary hardware being bench-tested, ie, removed from the case & placed on a table top (bench). My definition for bench-test, is all parts are out of the case & now on a table (this is done to isolate the case & case wiring as a source for the malfunction). I do not consider the rear button as a bench test, unless of course, all your parts are out of the case & on the table for the self-test & then for the jump-start test, make sense?

Self-test (psu), bench-test, jump-start on, they may mean different things to me than to u

https://i.postimg.cc/nhyYD6F4/bench.jpg

Here, w/a brand new mthrbd, I've assembled all hardware + monitor to 'bench-test' it 1st, before install. Shorted pins 6/8 to turn it on. As HDD booted, video on monitor, all is well, shut down then install into the case ... had any part of the system failed to work, I'd need to troubleshoot it, most likely with a troubleshooting guide &/or spare parts

My advice:

1) Take mthrbrd / cpu / cooler to a repair place

2) If not, buy PSU (or borrow a different one), test

3) If no joy, buy new mthrbrd, transfer CPU, test

4) If no joy, buy new test CPU, like a cheap 3820, test

These costly spare parts can of course be re-sold or held if they do not fix this

My last best advice:

Remove mthrbrd (+cpu +one stick of memory), cooler, GRFX card, power supply; stage on a table top, re-assemble, insert new 3volt coin-cell battery, plug in keyboard / monitor; jump-start on (pins 6/8). If shutdown, "take parts out and test them in other systems", & re-consider steps 1-4 above