Computer(s) won't post.

[312tech]

I consider my experience level to be expert. I've hand-built plenty of deskstops and servers. I work in the IT field and have plenty of experience troubleshooting, and I've thought I've seen it all.

I moved not too long ago. I personally moved the computers myself so as to handle them with great care. Getting them setup wasn't my first priority, so they sat unplugged for a bit. Now that I have time, I've started to set them up. First computer, no video. No post, no beeps, and no power to the usb mouse and keyboard. I tried everything to get that machine going; pulling all the components, trying the machine empty. Re-seating everything including the cpu. I even had two spare brand new power supplies I tried in the machine. After couple hours, I just figured that it was a case of bad luck and that board must have just gone bad. But now I've tried a couple other machines that previously worked fine too. No video, no post, no power to usb (keyboard and mouse). I have in total 5 desktops and 2 servers.... I absolutely refuse to believe that somehow every PC I own got destroyed in the move. One, sure... definitely could happen. Two, eh... maybe. But 7 machines, all moved by myself with TLC... ruined? I mean, professionally, I'm an IT Manager and I oversaw a moving company move over 100pcs from one physical address to another and we didn't lose a single one.

I'm not an electrician. Is it possible that I'm not getting enough power from the house? What other reason can there be?

 

[dave_the_nerd]

Last spring, I had some problems with a power strip that would only let me turn on two computers at a time. Kind of an issue because I was trying to build a third.

I spent a couple days swapping motherboards and PSUs, finding a couple small PSUs that inexplicably worked fine (a 150w Pico and a 250 SFX) when the brand new full size ones (450w ATX) didn't, and generally tearing my hair out.

I tried a different wall outlet w/o a surge protector - almost by accident - and suddenly everything worked right.

It's concerning that you can't get ANY on, but if you're running them off the same circuit as the furnace or something, I could see that being a problem. Or your surge protector could be cutting out for some reason - maybe you have a dirty power signal and it's freaking out when the PC draws initial startup/surge power.

 

[312tech]

I probably should have mentioned I did try different outlets. Basement, a few different upstairs. I wish I had a volt-meter to check what's coming from the outlets... just to be sure. hard to believe that would be the problem... otherwise I would probably have problems will all of my electronics. also tried with and without surge.

Actually... I do recall I had to purchase a new wi-fi router after moving in. it worked fine before moving, but it was dead as a door nail when I got here. Meaning, it powered on, but it never "posted" either.

 

[Burpo]

Are you plugging the USB keyboard & mouse into USB3 ports? That has been a post problem for many boards..

 

[MtnMan]

Can you hear drives spin-up? Will optical drive drawers open/close? Any fan's spin-up? Any LEDs illuminate on case or MB?

 

[westom]

I moved not too long ago. ... First computer, no video. No post, no beeps, and no power to the usb mouse and keyboard. I tried everything to get that machine going; pulling all the components, trying the machine empty. Re-seating everything including the cpu. I even had two spare brand new power supplies I tried in the machine.

After a couple of hours, not one useful fact has been discovered. Using a meter, some requested instructions, and only minutes of labor could have provided numbers here resulting in a next reply that is informed. And without speculation. That identifies a suspect without doubt and probably the reasons why.

Meters are so ubiquitous as to be sold even in stores that also sell hammers - for less than the price of that hammer. Meters are sold even in Walmart.

First, power an incandescent bulb from a same receptacle. If that bulb does not dim to less than 40% intensity when the computer (attempts to) powers on, then power is more than sufficient. To say anything more and useful requires specific three digit numbers from a meter.

Two options exist. Identify a defect before even moving or disconnecting any parts. Or just keep replacing good parts until something works. The latter was done for hours with nothing useful. Maybe spend minutes to obtain numbers. Then others who better know this stuff can then post an informed reply.

 

[Puffnstuff]

Could they have been subject to an emp pulse during transit?

 

[312tech]

Can you hear drives spin-up? Will optical drive drawers open/close? Any fan's spin-up? Any LEDs illuminate on case or MB?

>> I do hear the drives receive power. not all of the machines have an optical drive, but the ones that do will open. Fans will spin, and depending on which machine... LEDs will illuminate. It never posts to BIOS though


Are you plugging the USB keyboard & mouse into USB3 ports? That has been a post problem for many boards..

>> no, these machines aren't ancient, but old enough to not have any usb3.

Magnetic pulse was something I thought of, but that's too unlikely. What I'm going to try next is two things. I'll get a volt-meter. I'm a homeowner now and it will come in handy there too. The culprit I suspect now is dead CMOS batteries. These machines, again while not ancient, are old enough that dead batteries in all of them is completely possible. They've also been without any power for a couple of months.


from another post I was reading...

>> Re: Can a bad CMOS battery cause a computer to not boot up?
A bios battery can definately stop a computer from booting. I fix computers as a hobby and I have had at least 8 machines in the past 5 years that have been completely dead. Of the 8, half were found to be caused by the battery having gone short circuit, which permenantly holds the CMOS in the reset state. The other half were due to the battery terminal voltage droping to a level that causes momentary loss of cmos data, sufficient to require a manual cmos reset to default settings.

I'll report back my findings once I get the new batteries.... and if that's not the trick I'll see what the volt-meter says.

 

[312tech]

Could they have been subject to an emp pulse during transit?

thought crossed my mind... but probably highly unlikely.

 

[westom]

I'll report back my findings once I get the new batteries.... and if that's not the trick I'll see what the volt-meter says.

If the computer is powered, then BIOS has power from the +5VSB wire. If failure occurred as he suggested, then the motherboard is defective by design.

CMOS batteries are tested best with a meter. Swapping parts without first observing a defect is shotgunning; replacing good parts until something works.

Measure battery voltage with a meter. If below 2.9 volts, then battery may need replacement in or after 6 months. To create your failure, battery voltage must be well below 2.7 volts. If battery voltage is higher, then move on to other suspects. Another reason to have a meter.

Batteries should be good for much more than 5 years without power. If batteries discharge in months, then batteries were defective for years or motherboard has a defect that is discharging batteries.

 

[VirtualLarry]

Batteries should be good for much more than 5 years without power.

I suspect that you lack experience in this area. An unpowered motherboard, with a fresh CMOS battery, would be unlikely to retain settings for more than 5 years.

Now, for a properly-designed motherboard, as long as the mobo has power (plugged into PSU, and PSU plugged into AC outlet), then the CMOS is powered by the +5Vsb line, and doesn't draw from the battery.

 

[westom]

I suspect that you lack experience in this area. An unpowered motherboard, with a fresh CMOS battery, would be unlikely to retain settings for more than 5 years.

Please learn why CMOS batteries, once only good for five years, are now often rated for 10 years.

CMOS consumes so little battery power that the CMOS life expectancy is called 'shelf life'. That battery discharges after more than five years whether inside a computer or on a shelf due to internal discharge. Computer consumes so little power from that battery that only internal discharge currents cause battery failure after maybe ten years.

Same battery discharge occurs whether a 5VSB exists or not.

We have seen where a CMOS battery discharged in months. A defective CMOS semiconductor created microamp currents that must not exist in any properly designed computer. CMOS and associated clock consumes that little power.

In the early 1980s, Dallas Semiconductor made a 'CMOS with clock' chip that was pin for pin replacement in an IBM PC-AT. Dallas datasheets defined that battery good for five years. Today, that same part is rated for 10 years. Please read datasheets before posting. Please learn about shelf life.

Meanwhile, the useful solution is to measure that existing battery with a meter. To see a defect long before replacing anything. If battery voltage is sufficient (above 2.7 volts for a 3 volt battery), then move on to other suspects.

 

[myocardia]

Information about motherboard CMOS power consumption can be found here: http://www.tomshardware.com/reviews/power-supply-protection-calculate-consumption,3066-10.html

 

[VirtualLarry]

Information about motherboard CMOS power consumption can be found here: http://www.tomshardware.com/reviews/power-supply-protection-calculate-consumption,3066-10.html

As I said, 5 years, for your typical coin-cell CMOS battery. And if it's significantly under 3v, it needs to be replaced.

220 000 µAh ÷ 5 µA = 44 000 hours = 5 years
Most lithium coin cell batteries are rated at 3 V; however, actual readings on a new battery are usually higher. If your system clock seems inaccurate (it runs slow, for example), check the voltage on the CMOS battery. The highest accuracy is obtained if the battery voltage is maintained at 3.0 V or higher.

Anyone who has been doing PC tech support for some time could have told you those figures, just from experience.

Westom, the Dallas combo RTC / battery specs are not relevant. That's a separate modules, that combines the CMOS memory, RTC, and a battery module, that is NOT a CMOS coin-cell battery.

 

[westom]

the Dallas combo RTC / battery specs are not relevant.

Dallas RTC chip in early 1980s is same technologies found in today's computer CMOS. What was a 5 year life expectancy long ago (in early 1980s) in that Dallas chip is standard in today's computers. Battery life expectancy is that battery's shelf life. If a battery does not last that long, then a hardware problem exists (as explained earlier) or battery was inferior when purchased.

Arguing about a minor and irrelevant number is irrelevant to the OP's startup problem. Back to relevant numbers and the OP's issues.

Use a meter, as originally described, to measure that CMOS battery. Meter makes shotgunning unnecessary. Numbers were provided. Battery above 2.7 volts does not explain a problem. Battery at 2.9 volts should be replaced sometime in the next six plus months. Otherwise move on, using a meter as described earlier, to identify other suspects. All takes less time than was wasted arguing futilely about a battery's 5 year life expectancy.

 

[312tech]

I replaced the CMOS battery, and the computer(s) all booted right up without issue.

For the record, three of the motherboards were MSI k9a2 platinum, circa 2007? so, a little old. motherboards were without any power for probably 8 months. Makes perfect sense to me now...

Generally speaking, never really had a problem with CMOS batteries before. Probably because I used to upgrade my computer on a yearly basis.

 

[312tech]

Please learn why CMOS batteries, once only good for five years, are now often rated for 10 years.

CMOS consumes so little battery power that the CMOS life expectancy is called 'shelf life'. That battery discharges after more than five years whether inside a computer or on a shelf due to internal discharge. Computer consumes so little power from that battery that only internal discharge currents cause battery failure after maybe ten years.

Same battery discharge occurs whether a 5VSB exists or not.

We have seen where a CMOS battery discharged in months. A defective CMOS semiconductor created microamp currents that must not exist in any properly designed computer. CMOS and associated clock consumes that little power.

In the early 1980s, Dallas Semiconductor made a 'CMOS with clock' chip that was pin for pin replacement in an IBM PC-AT. Dallas datasheets defined that battery good for five years. Today, that same part is rated for 10 years. Please read datasheets before posting. Please learn about shelf life.

Meanwhile, the useful solution is to measure that existing battery with a meter. To see a defect long before replacing anything. If battery voltage is sufficient (above 2.7 volts for a 3 volt battery), then move on to other suspects.

From this experience, I would say 10 years might be possible only if the computer is plugged in and receiving power.

 

[westom]

From this experience, I would say 10 years might be possible only if the computer is plugged in and receiving power.

RTC and CMOS memory consume so little power that battery life expectancy is defined almost entirely by battery shelf life. That is same years - whether in a computer or stored on a shelf. Shelf life does not change whether a computer has AC power or not. Numbers in a next paragraph demonstrate same.

Assume his 0.22 A-hr battery. CMOS/RTC consumption at three volts and 25 degrees C is typically 300 nanoamps. IOW that is power for maybe 80 years - assuming no discharge due to shelf life. Discharging (due to shelf life) is why that battery is only rated for ten years. Actual life expectancy would be a function of battery quality. A better lithium battery has a longer shelf life.

CMOS consumes so little power that a residual charge in a dead battery is sufficient to maintain CMOS long after AC power was removed. But when not powered for a month, then a residual voltage drops too low. IOW that battery may have expired many years ago. But a dead battery was not apparent until after a month of no AC power.

So how many years was that lithium battery in that computer?

 

[Yayo3p]

Are you using the same monitor while testing all the computers?

 

[Charlie98]

Following the lines of an EMP... were the computers stored next to a magnetic field... like a big speaker or something?

 

[lakedude]

It has been solved already, it was the batteries.

http://forums.anandtech.com/showpost.php?p=37928226&postcount=16

 

[bigboxes]

It has been solved already, it was the batteries.

http://forums.anandtech.com/showpost.php?p=37928226&postcount=16

Lot of experts in this thread. LOL