Tech Support Forum banner
Status
Not open for further replies.

Computer won't power on, Motherboard or CPU?

15K views 30 replies 5 participants last post by  Jim@HiTek 
#1 ·
Hey, so a little while back I woke up one morning to find that my Dell Inspiron 546 wouldn't power on. When I plug it in, there's an Amber light on the motherboard but, there's no activity at all when I hit the power button. No fans or anything. Of course I started by running through the basics and I checked cables, different outlets, the voltage switch, checked to make sure it wasn't in sleep/hibernate etc. None of the basics were the problem. I've tested the power supply with a power supply tester and it works just fine. I've tried manually turning on the motherboard by shorting the two power pins and still nothing. Also tried disconnecting all the cables from the motherboard except the board power cables to see if I could boot into the BIOS but, still nothing. So at this point I know that I have either a problem with the motherboard or the CPU. Is there any way to tell which I need to replace or will I have to try one and then the other? Note, about a month before the computer wouldn't power on, the on-board NIC stopped working so I had to result to buying a USB Ethernet adapter. This makes me lean towards the motherboard being the problem but, I have no way to tell. Any feedback would be great.
 
#2 ·
First disconnect power from CDs, DVDs, USB devices, and HDDs. Then remove the RAM and power up. Do you get a beep? If yes, replace the RAM one stick at a time installing the last stick into the first socket. If no, it's probably the mobo.

But before I'd condemn the mobo, I always remove the CPU and check the socket for dust/dirt and clean as necessary. Then reinstall and operate the capture latch several times to clean the contacts. Then add new paste and the heat sink.
 
#3 ·
Just got the chance to do that. No beep and still no activity other than the greed light on the power supply and the Amber light on the mobo.

---

I'll go to CompUSA tomorrow and pick up some thermal paste to try this. If there's still no activity after this, I found a New replacement board for a nice price online.

Just wondering, unless there's any bent pins or other damage to a CPU, is the CPU ever really the problem or not really? Can a CPU go bad?
 
#4 ·
A CPU can go bad, but it's rare.

Depending on how clean the area the computer is in, how often the owner cleans the interior of the case, and the relative humidity and ambient temperatures, what happens a lot is connectors can develop a patina that slowly grows over time. Causing an intermittent open circuit. This can play havoc with a CPU as it needs good contact with every pin.

Before you run out and buy thermal paste (but you should have some anyway), just use what's available from the old paste and operate the latch several times as I suggested above. Running it for several minutes with inadequate thermal paste won't bother it in the short run. That way you can make your decision quicker.

BTW, I seldom use a PSU tester, preferring instead to swap the PSU with a different one. So, I can't really give a qualified answer if you asked me if PSU testers do a good job. If you have an extra PSU, you might try a swap first.
 
#6 ·
Ah, okay. Thanks a bunch for the information.

Will do this today without buying thermal paste.

Unfortunately I do not have an extra power supply. Reason I don't really have any of these is because I'm a high school student who has only ever worked on old school computers up until my home computer stopped working, haha.
I'm not sure if the power supplies available over at the school would work with my computer here since those are about 10 years old and I'm not exactly positive that all the connectors and such are the same. I will check though.

If I can't find a known working power supply to replace and test on the machine, would you try buying a replacement power supply or motherboard first?
That is unless I could find a local shop who would do a quick power supply swap to test for cheap...
 
#7 · (Edited)
Power supply failure is far more likely with the 546 as that model is shipped with a 300W PSU.

I'd be interested in knowing what yours is rated for...look for a paper tag on the case of the PSU.

Then, are you running a plug-in GPU or are you using the on-board VGA?

You can get the specs for your Dell by going to Dell's web site and entering your service tag number.
 
#9 ·
Mine is the 300W PSU that it was originally shipped with.

I'm using an ATI Radeon HD 4350 plug-in GPU. I also have a plug-in sound card. Although the computer has worked just fine since 2009 other than the on-board NIC having stopped working a couple months ago.
 
#13 ·
A 550W good quality PSU is the minimum we suggest for any PCI-E GPU.

Will try both testing with a different PSU and Jim's suggestion above when I get the chance.

The only power supplies I have available are older ones that have the 4-pin molex connectors to give power to PATA drives. They do not have SATA connectors and such like the power supply my home computer has.
But, any power supply with a 20 pin and 4 pin connector to give power to the motherboard alone should suffice for testing, right?
 
#14 ·
I believe you'd need to borrow a PSU with a main power connector of 20+4, and which also has an additional CPU connector of +4.

And we're just trying to figure out whether or not it's the PSU or the mobo, so if the PSU connectors are SATA or PATA makes no difference at this point. For future reference, they do make PATA to SATA PSU adapters.

BTW, eBay shows a new DELL F896N complete for $90 + shipping with a 30 day warranty (case, mobo, PSU, case fan but no memory, CPU, or drives) so I'd not want to put more then $30 into a PSU unless you're planning on upgrading the computer in the near future and moving the PSU into the new 'puter. If, in case it turns out to be the PSU.
 
#15 ·
The ebay PC is an option but you're just using another low quality underpowered OEM PSU and a good quality PSU, with sufficient power, would cost less.
 
#16 ·
Sorry for the late reply. I borrowed a working power supply from a friend but, had to wait on him to give it to me. Well, turns out it is the power supply... With the working one plugged in, the computer worked perfect. Guess I learned to never trust a power supply tester.
If it weren't for me being told here that it only tests to see if the unit is getting power, I probably would have gotten another part that wasn't the problem. & In the future I should buy an extra power supply in case I need one like in this situation.
I hope to pick up a new PSU this weekend to get the computer running ASAP. Would this one fit in my case? Picked this one out because it had good reviews and they have a unit in my local CompUSA store so it would be an easy pick-up.
I'm not sure if I'm going to pick up the 550W, 650W or 750W yet but, it will be on of those three. CompUSA.com | LSP550 | Ultra LSP550 550-Watt Power Supply
 
#18 ·
Well, I just asked my parents who are buying it and it turns out they want something in the 400-450W range (to save some money).
So if I were go with XFX I would choose this CompUSA.com | P1-450S-X2B9 | XFX Core Edition 80 Plus Bronze 450W PSU

However, they're saying they would like something that they could pick up at the local store today and it doesn't look like they have the XFX supplies in stock...
It says they have limited stock of this one; If they have it in stock , is it just as good as the XFX supply? CompUSA.com | CP-9020046-US | Corsair CX430 V2 Series 430W Power Supply
 
#19 ·
Using a low quality underpowered PSU is only asking for a repeat of the problem and more possibly problems/damage.

The XFX 450W would suffice, and certainly worth the extra $2, but I would not consider the 430W Corsair. It's lower quality and well known for problems.
 
#20 · (Edited by Moderator)
Resurrecting an old thread because it is similar to my issue.

First, I've been in the computer hardware and software development industry for 25+ years. I've done a lot... but I'm retired, old, my technical skills are rusty, and can't see very well any more. :banghead:

I've had an Inspiron 546 running pretty reliably since I purchased new about 5 years ago. I recently had a USB 2.0-based external drive die and decided to upgrade to a USB 3.0 device (nightly backups of all the computers in the house was taking 4-5 hours on 2.0... I figured I'd benefit from the 3.0 bandwidth). So, for the FIRST TIME EVER IN THE COMPUTER'S ENTIRE LIFE, I popped open the case, vacuumed out the dust, and plugged in a USB 3.0 card in the PCI-e x16 slot. I purchased the card new from Newegg, and it arrived today. It's an offshore card, but in my experience, this isn't a particularly complex device, so...

After plugging in the card, I buttoned everything up without any pre-test and plugged the machine back in. It wouldn't turn on... nothing. I pulled it back out, pulled out the new card, and tried to fire up the computer on the desk. Still dead. I unplugged the 20-pin from the motherboard and jumpered across green and black and plugged in the power supply. It spun up and spun up the 2 drives (hard drive and DVD). I then buzzed the front panel switch with a continuity meter and it's working. I then popped out all 4 sticks of RAM and all the accessory power connectors to see if I could get the motherboard to come on and fail POST. Nothing. I get that a power supply can go bad, and I get that the motherboard can as well... but the great mystery is that it happened EXACTLY after a trivial upgrade... and it's the only time ever in its life to have even been opened up, much less fail.

When I vacuumed, I wasn't heavy handed, and it wasn't plugged in... I just cleared out some dust. Perhaps the dust is what was holding everything together.

What I'm hoping is if anyone has some Dell-specific info on this, like a motherboard "reset" jumper (I used to work in the Dell Enterprise Systems Group in the early days... and our big boxes had reset buttons on the motherboards... unlabeled/undocumented). So, any advice? This is my primary computer, and it serves as sharepoint for all the other computers in the house... so this is devastating. I'm typing this on another machine that's cobbled together, and checking email via a webmail client into my offsite webserver since my POP client is sitting the harddrive of this dead machine.

:banghead::banghead::banghead:

I'm getting too old for this stuff.
 
#21 ·
Since you've been in the industry that long, you probably know how dangerous using a vacuum is to clean electronics...unless it's a low static device. My first question would be exactly that, did you use a low static vacuum? What's the RH where you are?

I'd suggest at this stage that you try a different PSU, first removing the new USB expansion card. After that, you should consider doing a bench test. You should probably get yourself some canned air, and electronic contact spray as well.

It's late, so I'll say goodnight.
 
#22 ·
My first question would be exactly that, did you use a low static vacuum? What's the RH where you are?

I'd suggest at this stage that you try a different PSU, first removing the new USB expansion card. After that, you should consider doing a bench test. You should probably get yourself some canned air, and electronic contact spray as well.
Thanks for the reply. I'm in Norcal. It's pretty humid here (and it's actually been raining all day). And yes, I was using a computer vac. It's old and crusty, so I can't say with certainty that it doesn't have ESD issues... but I have used it for many years without issue. I have ordered a power supply from an ebay seller... we'll see.

USB 3.0 card was removed right after first sign of problem. Also, it is now in another computer and working fine. As I suspected, there is nothing inherently wrong with the brand new (though offshore cheapo) card. I'd say I either zapped something with ESD or the power supply was "on the way out" and the extra card put it over the edge (though, again, a single 1x PCIe card isn't going to pull much current. One other possibility I thought of... if the PCIe slot had a loose pin, I could have pushed it into another pin when I inserted the card. I saw this problem years ago very early in the Intel Slot 1 ("slot 242") edge connectors, where pins that weren't precise could warp and come into contact with the next pin over when the CPU was inserted. However... I asked a family member with good eyes and a magnifying glass to look in the empty slot for bent pins, but they all seemed OK.

Thanks for the reply.
 
#23 ·
The new PSU idea is sound, as long as you're getting a good brand. I've also run into the crossed PCI slot pin problem back in the day. Glad we're not using them much anymore and that you're are OK.

You're probably way ahead of me but, with those other computers in the house, doesn't one of them have a working PSU you can test with?
 
#24 ·
All the other machines are laptops... except the brand new HP I got for my daughter. I decided not to scavenge that too hard, since it's only been out of the box for 1/2 day. I put the USB 3.0 card in the new HP and am setting up the machine from OOBE right now.

I did order both a power supply and mobo via ebay from a surplus place in Round Rock (parts are listed as new new surplus, so I'm guessing they clear out Dell's expired inventory after it retires them)... fingers are crossed that this will get me up and running again.

Once up, I need to make some adjustments to my DR plans. Most data has been on stand-alone devices, but I've had a string of bad luck with those in the last 9 months, first losing my primary share device, then after having to get _3_ replacements from Seagate because the first 2 were DOA, the backup device has failed as of last week... that was the reason for the USB 3.0 card that started my most recent problems... I was adding the card so I could move to a faster transfer for these external devices because the weight of the work is getting to be too much, and I suspect I'm just burning things up by running read/writes for several hours straight.

My current mechanism is that everyone's data is written to their personal share on the primary external drive. I then do a nightly robocopy style backup to the backup device. And then, 3x per week, I do a real "backup set" backup of that robocopy data. You'd think that would be enough redundancy, but none of that matters when I lose the machine that's managing all that :banghead: :banghead: :banghead:

I was contemplating abandoning all that for some kind of synchronous "cloud" solution, but then my wife would yell at me when "the Internet is broken." :hide:
 
#27 ·
The robocopy uses an incremental switch, so I'm only grabbing files that have new date stamp, so I end up with a "base" + increments for every day of the week. It's a bit of a hassle if I want to recreate a single restoration, but it's easier on the bandwidth than trying to do a full robocopy. The real backups (Norton) are also incremental. The reason for the robocopy method is that I often need to fall back to previous versions of some files my wife uses (she was Ned Ludd's wife in a previous life and saves/overwrites/deletes seemingly indiscriminately... having the robocopies lets me fall back when she screams bloody murder "all my work is lost!!!!111"). Otherwise, I'd just do real incremental backups.

As for the cloud, I wasn't thinking full backups, but a synchronous system... WRITE in real-time to the cloud. READ in real-time from the cloud. Let the cloud provider manage the backups on their end, where they have high speed bandwidth, and not try to do "personal backups" across the wire.
 
#30 ·
Will do. I tried pulling the manual from Dell the other day but kept getting 404 on both it and the other PDF listed on the support page (spec sheet maybe??). I figured they had retired the docs and failed to remove the links.


Are you using WinRoboCopy? Or just Robocopy?

BTW, one reason I stopped buying Seagate was the clustering of failures I got from them. So I went with Western Digital when I was buying more then 3 or so. WD tend to fail in singles rather then clusters. In my experience.
Robocopy from the CL (I'd share my switch settings with you but they're on the currently dead machine :( ). I've been developing HA storage systems for many years (I'm retired)... from 1st and 2nd gen Dell PE Cluster offerings, to EMC Symmetrix and Clariion systems, and then for years at LSI, building the OEM boxes branded by just about everybody. In all those years, my development organizations touched literally HUNDREDS OF THOUSANDS of drives... about 90% Seagate (of course, business class stuff... not desktop, and usually FC and later, SATA)... and our dropout rate was TINY. Same for field incidents. However, clearly, the days of Seagate superiority are over. See this Backblaze case study data for some shocking numbers (like one Seagate drive at 120% fallout rate!!!!) --

To replace this latest Seagate failure (and, again, the reason I bought the stupid USB 3.0 card that led to my current situation), I bought an ioSafe Solo G3. It arrived the other day, and is physically and visually impressive. It is a bit pricey/GB, but I'm hoping for piece of mind (it's currently $350 w/free ship at Costco). I have no idea what brand spindle is inside it, but I don't care... I just want it to work, and look forward to actually getting it up and running soon (assuming I can recover my Inspiron machine).
 
#29 · (Edited)
I haven't used the cloud for backups myself, but just recently read a comparison of various backup methods and Cloud wasn't highly spoken of. Mainly because of speed over the internet.

Can't find the article.

Seems we're always running into this bottle neck problem with backup systems. I'm sure there isn't a perfect method but Robocopy seems like one of the best nuts & bolts solutions.

Are you using WinRoboCopy? Or just Robocopy?

BTW, one reason I stopped buying Seagate was the clustering of failures I got from them. So I went with Western Digital when I was buying more then 3 or so. WD tend to fail in singles rather then clusters. In my experience.
 
#31 ·
You have vastly more experience with industrial quality drives then me. My small business (I've been retired for 9 years now) went with the savings we got from consumer drives. In that realm, I'd see more Seagates fail then WD. AND, to me the failures of the Seagates were often devastating, since so many times I could not just attach them as slaves and recover a clients files. The WD, when they'd fail as a Master, would often fail in such a way that I could use them as Slaves and recover data. Then I'd just toss them so I don't know if they were totally recoverable...in any case I couldn't trust them in a clients machine. Or in mine. Then, with the WD's, I could remove the card, clean some contacts and sometimes be back in business. Never had that happen with a Seagate.

Anyway, it's our personal experiences with a product or products that leads to our own biases.
 
Status
Not open for further replies.
You have insufficient privileges to reply here.
Top