Wouldn't boot this morning. Doesn't reach the POST BEEP most times but occasionally it does, and don't see any BIOS messages (video or system). No signal to monitor most times.
I just took it to the bench to start testing some voltages but I'm in the dark a bit.
All internal components less than a year old and reasonably good quality.
Fans are turning but I just noticed the processor fan doesn't look to me to be turning very quickly - seems to be slowing them speeding up again, with a bit of a click, but that could be an illusion.
I am almost sure CPU overheating is not the issue though, you'd expect that to run normally for a bit then shut down.
What do we reckon, PSU? I found a webpage on things to test but not sure that it'll be decisive. www.sevenforums.com/tutorials/301799-psu-test-dc-output-voltage.html
EDIT: If I take out the memory I get the POST beeps saying no memory.
Last edited by: smokie on Thu 7 Jan 16 at 18:13
|
|
possibly motherboard have you tried to reset the cmos,does the display come on witohut checking looks to me a hardware error,
|
Yep tried the CMOS reset. Once in many attempts the display worked, but I am watching the signal indicator (which turns blue from red) rather than the screen - and no signal )most times). Once in a while the signal light goes blue but then the PC powers down before there is anything on the display. (I hope I've got that right, it didn't happen very often at all, maybe just 3 times)
How can I test for faulty m/b? I have removed all internally and externally connected devices, cards and components and it still won't go.
|
Do you have any memory at all that you can borrow to test it with? It is vaguely reminiscent of an XPS fault I had which proved to be memory related.
Although to be honest it sounds like the MB.
|
Might have some old stuff knocking around, even so I can try one stick at a time and see what happens.
All the components are under warranty but it's a matter of knowing which one is at fault... all came from Amazon except the PSU so I oughtn't have any problems swapping them.
|
...having had similar issues in the past, my suspicion (though it could be m/b, lack of post info implies components are less likely) would be the PSU, but they aren't the easiest to test.
If you can't switch it out for another , or fail to diagnose anything else, you might consider a cheap psu tester off Amazon or eBay for around a fiver.
|
|
The fan will speed up and slow down during the boot process. Remove any USB devices first. thenn test boot, Remove the hard drive and the cd drive, and test boot up. Got a separate Video card? take it out or swap it for another.
|
My hunch was the PSU. I have my test stuff ready to follow through that link I posted earlier, tomorrow. I did one test on a molex and the paperwork said it should be 12.4v but mine was 12.04, which seems quite a difference.
I realised the fan is variable speed and I've already tried taking everything out, including all USB devices and the video card ( - it has onboard).
Following Mark's suggestion earlier on memory, I played around taking one out then the other. At the bench I have no monitor (will get one there tomorrow) but about 50% of he time I am getting the single beep of successful POST. It doesn't matter which memory stick is out ( - they are a matching pair).
I have a much lower rated PSU in another PC and tomorrow I'll see if I can get it out to try, and e;ave all the peripherals off.
|
>> My hunch was the PSU. I have my test stuff ready to follow through that
>> link I posted earlier, tomorrow. I did one test on a molex and the paperwork
>> said it should be 12.4v but mine was 12.04, which seems quite a difference.
Its not - thats usual. .
|
>> Its not - thats usual. .
Agreed, spec is +12V +/- 5% IIRC, then there's the inaccuracy of the meter.
Last edited by: VxFan on Fri 8 Jan 16 at 10:22
|
...I'd say it was "in range", but not particularly "usual" for a decent quality power supply (and if measured correctly might be a symptom that issues exist on other "rails").
The prime suspects for the symptoms are the PSU, a fried or shorted board, or a fried or badly seated CPU. (patently, there are other possibilities, but POST would normally help identify most of these).
The easiest to test is the PSU, so it makes sense to do that.
Testing the CPU and board are not going to be at all easy (unless you can "borrow" a compatible CPU), so it then makes sense to test the removal of, and individual replacement of, all the separate components (which you've done/are doing).
Frankly, if you then have no luck diagnosing, you'll have to choose whether to "return" the CPU or board as faulty (and I would choose the board, but that decision might be affected by which supplier it came from).
|
Thanks.
I doubt anything is fried as in it got too hot as all the fans are working. I don't do much that stresses it particularly and it is a quite well specced machine.
Seeing as I have that web page I will carry on with the voltage testing. All components except the PSU came from Amazon so I can't send the lot back, but I could send back the PSU and either board/processor/memory. It's a matter of making the best judgement on the available evidence.
|
...I have to say that when I re-read my previous post (on the matter of the 12.04V), I thought it needed a bit of clarification.
The spec is indeed +12V +/-5%, but this should be met across the load range (and is a fundamental reason why such a range is tolerated).
As the load rises, for most PSUs the voltage will fall, but should stay in spec.
In order to meets this, PSUs will generally "over-deliver" at low loads, and may be specified (as yours and mine is) at say 12.4V (which will be the design maximum, and at that low load).
Measured through BIOS, my-moderately configured machine (two HD, two optical, and a few PCI(-e) add-ons) generally reads 12.43V, and doesn't drop much under any load.
The better specified PSUs will generally be set with less "headroom", but tests/reviews show that many/most still deliver 12.2V at low loads (and few drop as low as 12.04V at full load).
I'm assuming your test (given the non-functioning) was carried out at virtually no load?
Far from conclusive, and I'm sure it isn't helping your diagnosis, but I thought I'd add my clarification.
|
Ta!
Testing is simply that in the link I used above, i.e. turn on the PSU and test each of the pins.
Which I've just completed.
Would I expect the PSU fan to turn when the m/b connector is not connected to the m/b? Cos it wasn't, which may mean the results are invalid! (It was def turning when connected).
Anyway the results are in. There are only two pins which register a voltage - the purple one registers 5.01 (5.01 in the link) and the green one registers 1.99 (0.06 in the link).
So I am missing loads of voltages.
Do we think that's sufficiently conclusive to replace the PSU? Or can someone advise how I can test differently, or whether I have done the tests correctly.
I did look at swiping he PSU from the other machine (this one!) but I'd sooner not if I don't have to... I guess that's the only conclusive test though.
The PSU btw is a Corsair CX430, driving an I5 processor with 8Gb, 3 x SSD, 1 x HDD and 1 x DVD, plus a half decent NVidia card and additional Ethernet card. I only have 1 USB disk which doesn't require external power.
The problem I have is that the details of where I bought it from are in Outlook on the dead computer... Might be able to use the backup pst on the wife's laptop to load the folders though...
|
...it may be worth checking, but I don't think you can test a PSU without it being "on-load" (or at least with some connections) .
(I'm working from memory of when mine went **ts-up a few years ago).
The page you linked to shows testing from the back-end of the connectors (in-situ) for the items it covers, which rather bears this out.
I do know that the PSU fan won't turn with it unplugged, as one basic test of a "dead" PSU is to short two connections with a paperclip to see if the fan fires up (research may show that this may be the main switching point and will also cause the other outlets to be powered).
www.cryptobase.co.uk/portfolio-item/how-to-start-a-corsair-power-supply-without-a-motherboard/
The above is also consistent with the existence of PSU testers, which, I think, have the appropriate internal connections and loads to be able to test "off-board".
|
|
Thanks, that's helped a lot. With pins shorted as suggested in your link, all the pins showed a voltage not far from those suggested in my link except the blue one, which should have been 121.34 but was actually showing -11.77 (it was the only one in the minuses, and I hadn't reversed the probes...). Is that my problem?
|
AFAIK, that pinout is -12V (reference), so you reading looks OK.
Looking more like a m/b problem.
(I'll do a quick cross-check between the voltages shown in your link with another source, just to see if there are any other inconsistencies)
|
Your PSU is ATX 2.3 compliant, and you would be better referring to something like this:
pinouts.ru/Power/atx_v2_pinout.shtml
...and I suspect that you've just checked it out as OK....
|
Thank you.
So, how to test the mb...?
Actaulyl I am going to re-assemble the lot tomorrow and see what happens. When I put the plug back into the m/b and [powered it up, and put both memory sticks in, I started it twice, and twice it reached the single POST beep. I don't doubt there could be an underlying problem lurking but maybe there was a badly made connection.
I only have a few weeks before the warranty is up, so I guess even if it's working I will swap something. Which would we go for? I would say m/b in preference to processor.
|
...I'm going to be no better at guessing than you.
Putting everything back together carefully, and ensuring cards, etc. are properly bedded is a good idea, but the law of sod will dictate that it will make no difference.
Prime candidate if the above doesn't make any difference is now the m/b (especially if you are getting some POST bleeps), but it still could be the processor.
Problem is that testing these (for the average punter like you and me) is not very easy.
(BTW, the voltages you've been referring to on the page you linked to are those that were measured in that testing, not the reference voltages, which are, of course, +12V, +5V, +3.5V, -12V, etc for the various pins).
|
Sooo...
Moved on a little this morning. I changed the memory and there is no difference with the new memory. So it doesn't look like a memory issue.
I am now consistently able to get to the single beep which I thought was unattainable. I genuinely think there was an issue with it but I have to accept that some of the time I simply was being insufficiently patient, as sometimes it takes quite some time to reach it (not least as sometimes, after a config changer, it shuts down and recycles). (Definitely was an issue thinking about it).
So when I hook it back up to my screen I thought it would be fixed. No. :-(
What happens is that it single beeps, then the BIOS splash screen is shown, on which I can see and move the mouse cursor, then after a few seconds the machine recycles. I've tried using the keyboard to get into the BIOS but no luck. I have also reset the BIOS a number of times now...
Are we reasonably confident ruling out the PSU? I do have another one (in the machine I'm currently using) but I'm loath to spend the time, effort and risk of taking it apart on a likely wild goose chase.
For me it's looking like mobo but I'd welcome other opinions or ideas.
EDIT: These are the parts:
Mobo www.amazon.co.uk/gp/product/B00K9R1OCW?psc=1&redirect=true&ref_=oh_aui_detailpage_o08_s01
CPU www.amazon.co.uk/gp/product/B00JST2QEW?psc=1&redirect=true&ref_=oh_aui_detailpage_o08_s01
Memory www.amazon.co.uk/gp/product/B00J8E91T0?psc=1&redirect=true&ref_=oh_aui_detailpage_o08_s01
The CPU is obviously the priciest bit.
Last edited by: smokie on Sat 9 Jan 16 at 10:43
|
Can you eliminate the HDD?
What you're describing sounds similar to my problems with The Lad's PB laptop related here:
www.car4play.com/forum/post/index.htm?f=6&t=10938
|
...if you've measured the PSU voltages as you say, that looks considerably less likely to be the problem.
Beyond that we're scratching a bit.
One further (elimination) suggestion would be to remove the graphics card, connect to the cpu/mb onboard graphics port (which I think your combination supports, though HDMI only, so might be a bit awkward, and to which it should hopefully fall back with no card in), and see if it will boot properly then.
If it works, graphics card is suspect, if not, then it's simply another component eliminated.
(you might be able to do a quick test by removing the graphics card and not connecting HDMI, as I would think it should work (i.e. not recycle) - if it's going to - with no monitor/tv attached, but it will be less conclusive).
|
...and, as per Bromp, it is always worth seeing whether you can boot from another device, though if it locks up/recycles whilst actually in BIOS settings (as opposed to in/after the boot process itself ), I'm not convinced it's HDD.
I have a number of boot/diagnostic disks around. This is quite useful:
www.ultimatebootcd.com/download.html
...but requires a m/c to download and burn a CD ISO (or alternatively, create a bootable USB stick - instructions are in the text).
Again, not over-good if it's falling over in BIOS settings.
|
Graphics card is already out. I can easily disconnect all drives, had already done a bit of that but if it looks promising I have numerous flavours of OS on CD which I can try to boot from, also bootable USB sticks which might be preferable.
Will be later or tomorrow now as I seem to have got roped in to household chores :-(
|
No luck there then. Disconnected all the data cables from the drivers (3 x SSD, 1 x HDD and 1 x DVD). I'm pretty sure from memory that it was set to boot USB first, but anyway it didn't.
I can pause it successfully on the BIOS splash screen, then whatever I select it just moves on, the splash screen disappears, a large (movable) mouse icon appears for seconds then it recycles. So I can't get into the BIOS to check the boot order.
I dunno if the BIOS can get screwed up, and if it can how to fix it. When I next get an opportunity I also want to check the grey wire which provides power to the CPU. (Maybe I'll bite the bullet and get hold of the other PSU).
I've now posted on Toms Hardware for thoughts. However I really appreciate everyones help here so far, and would be delighted to follow up any more (polite!!) suggestions...
|
...does the motherboard support a boot device menu display? (invoked at same point in starting as BIOS entry).
From a quick browse, I think it's likely it does.
It may be F12, and should show you the list of bootable devices present (so USB stick needs to be inserted), rather than entering full BIOS.
If you have, it may give you the option to boot in both UEFI and non-UEFI mode, and I'd try both.
|
Yes it does, and it is F12, but I am pretty confident it will be the same result i.e. black screen with large mouse pointer then recycle itself. Will try that tomorrow though :-)
A suggestion from Toms Hardware thread is remove the CMOS battery completely for a minute. Will also be trying that.
|
|
If you happen to be passing through Woodley, I've got a couple of PSUs that stepson left behind when he moved out recently. IIRC they were just surplus to requirements and barely used (if at all), although I can't guarantee they work. But you're welcome to take them if you think it might help - one's 450W (brand unknown but quite heavy) and the other is 500 (CIT 500CB).
|
Thanks Focusless, I might just take you up on that but unlikely to be today.
Today I took out the last remaining non-essential item (network card) and left the mouse and keyboard off, and took out the CMOS battery (as suggested on Toms Hardware). I tried it without the CMOS battery then with the CMOS battery.
Other than not seeing the BIOS screen which I was seeing yesterday it's still not working in the exact same way. I managed to use a paper clip to test the 5v to the CPU (grey wire) and that is fine.
|
....the board is DUALBIOS specified, which means that it should fall back to a second copy of the BIOS if the main one fails (and warn you about it in the process). Views appear to differ as to how well this works in practice, however. (And the usual way a BIOS gets corrupted is if an attempt to "flash" it fails). Corrupt BIOS thus looks like a low probability.
Having tested/measured the power supply, I think you're probably postponing the inevitable.
You could flip a coin on the m/b or cpu being borked, but all the indications are that it is one of them.
Unfortunately (I don't think you can get to POST with no cpu present, otherwise it might be worth removing the cpu and seeing if there were any bleeps) so I'm lost on how to determine which without substitute parts.
|
Yeah I read that it won't go anywhere without CPU too.
Shame my 2nd PC has an AMD processor. I have a mate who's Intel PC I maintain but I really think it'd be a bit risky to use his as my testing ground... I also don't think his is the right spec board. Also my neighbour may have a suitable PC and is IT literate enough to understand the issue but I think his newer devices have all been handhelds.
Thanks for your support on this, it's always useful to have someone to chew over this kind of thing with. I'll update later with my plan...
|
I completed testing with another PSU, as expected it failed in the same way.
So either mobo or CPU is going back.
Which would you reckon is most likely?
(I'm thinking mobo is more likely but that's only 1/3 the value of the CPU if I get it wrong - I suppose I only have one chance to get it right, given that they both came from Amazon in the same order).
|
Before you do, pop the CPU out and back in again, and try it.
The Motherboard is more likely to go than the cpu.
|
Not read this thread, only just seen it but in all the years of me repairing PCs, I have rarely ever seen a faulty processor. On the other hand I've come across lots of faulty motherboards including brand new ones. I built a PC for a mate last year he bought all the stuff and bought an AS Rock board, of course the voltage regulator on it was faulty.
Edit just read the first post about the fan, I would think it was because the motherboard isn't getting the correct power in the correct places possibly due to the board itself rather than the PSU (it seems you have already swapped that).
Have you also checked the RAM?
Edit also if it was the processor, you could get a consistent beep code each time tried to POST, the fact it doesn't always get that far suggests it is the motherboard.
A quick test is to remove the board out of the case, remove the CPU and RAM (at different times) , connect a speaker, you should always get the correct relevant beep code.
Last edited by: RattleandSmoke on Mon 11 Jan 16 at 12:09
|
I've used alternative PSU and RAM and the fault remains so they are OK. Also removed all non-essential attached devices (internal and external).
I was consistently getting to the single beep now. Yesterday I was getting the BIOS splash screen just before the power recycle but today (I can see the monitor going active (receiving signal and blackness changing) but I am getting no splash screen before it recycles. Still no response to keyboard (del or F12).
Re-seated the CPU as suggested, no change . Power up without CPU - it just restarts regularly - doesn't get as far as any beeps.
So later today I'll be contacting Amazon for a refund/replacement. Any recommendations that I should buy a different mobo (same budget!!), if I have the opportunity? Tjis one is Gigabyte Z97P-D3 Intel LGA1150 Z97 ATX Motherboard (4x DDR3, 4x USB3.0, 10x USB2.0, GBE, LAN, HDMI). I need the USB3 and (at least) 6 SATA ports. This is it
www.amazon.co.uk/gp/product/B00K9R1OCW?psc=1&redirect=true&ref_=oh_aui_detailpage_o08_s01
|
|
I always go for an ASUS board, Must have built about 40 pcs in my time, only had one give me build issues, and then was a pita for 18 months before the bios updates finally sorted it, (it was an A7V) and never had one fail.
|
I've never had any issues with Gigabyte or Asus boards, looks like a decent board.
It was an AS Rock Z97 gaming board I had all the problems with.
|
Amazon, "as an exception", are sending a replacement tomorrow. Pretty impressed with that, and the free returns label, but they said that normally I should go to the manufacturer with warranty claim, even in warranty period. Isn't that a bit deceptive? I though responsibility was with the retailer (in this case Amazon SARL)
Anyway - I'll update once fitted.
I suppose it opens another avenue - if I still have a fault then it MUST be the processor, so I go direct to Intel...
Last edited by: smokie on Mon 11 Jan 16 at 17:54
|
|
Yep it is the retailers responsibility. Pretty sure it is not the processor because I would imagine you would get a beep code every time if it was. I would be extremely surprised if the processor was faulty, just never ever seen it.
Last edited by: RattleandSmoke on Mon 11 Jan 16 at 18:02
|
A gold star to everyone who said Motherboard. The new one arrived today and is fitted and working.
And a virtual pint to all who provided help and assistance!!
|
" The new one arrived today and is fitted and working. "
Which one did you get? Was it the Gigabyte Z97P-D3 one in the Amazon link?
|
|
Yeah, a direct replacement. Interestingly £6 or so more than I paid a year ago but it as an exchange so that's fine. I noticed the memory is a lot cheaper now and the processor was about the same.
|
Glad you managed to get there.
As far as Direct Warranty with the manufacturer is concerned, whilst any further problem and follow-up after such action might affect your rights under the Sale of Goods Act, I don't think I'd be over-concerned with a major manufacturer - many are set up for direct warranty.
It isn't always convenient, however. I recently returned an MSI graphics card just within 2 years (but still in warranty). To confirm warranty status, I contacted MSI by email, and they did confirm it, and were entirely happy to act on it as long as I mailed to the Netherlands (where their EU base is). At their suggestion, I took the cheaper route of mailing to the UK supplier (slower, as it ended up in NL) but considerably cheaper.
|
Oo'er. Came back to the PC earlier to find it having rebooted, and "Preparing and automatic repair". I left it on that screen for a short lifetime but it didn't change. Rebooting got me to various stages very similar to the problem described in this thread. Ended up taking out one memory stick and it booted, went through the repair thing properly then restarted OK.
So I have backups going on at the moment, "just in case", and will play around tomorrow with the moment to see if I can pin the problem down.
Or maybe it's a flaky Windows build...
|
My PC is broken... - sajid
