Weird power on thing

[Brian Stirling]

I just completed a new PC build a couple weeks back and after moving the PC to it present location I have it and the rest of the gear (monitor, audio, label printer, desk lamp) all connected to a power strip that I turn off at night or when not in use to save power. Now, when I power on, the power comes on for a second or two, the drops off again for a second or two, then comes back on and completes the power up. If the power to the PC isn't turned off at night it does not do this double power dance. The details are:

Asus x99-Pro/USB3.1 motherboard
Thermaltake Touchpower TPD-070M power supply

My guess is one or the other of these components is doing this and I don't know which one it is. I don't believe this is a huge problem, but it would be better if power didn't cycle twice as often as needed. Anyone know what the deal is and is there a setting to prevent this. If worse comes to worse I guess I could switch the PC power to an unswitched outlet and then use the power strip to switch the other items.


Brian

 

[Red Squirrel]

I've had computers that do this too, I don't recall the mobos but it may have been x99 Asus boards.

I have two x99 Gigabyte based computers, and they do similar weird stuff, often times when they boot everything is DOG SLOW, like it's drawing the text slower than a bubblejet printer can print it on paper. Or it just keeps looping into different shades of black screens. (monitor off or monitor on but all black) sometimes with a blinking cursor in corner. A couple forced power cycle usually makes it normal again.

Not sure what causes this, maybe it's some kind of test it runs or something.

 

[Burpo]

Latest Flash dated 1-06-2016

X99-PRO BIOS 2101
Improve system stability.

https://www.asus.com/us/Motherboards/X99PRO/HelpDesk_Download/

 

[ReignQuake]

I've had that problem too and the boot process was excessive despite the fast boot features. I dont know what fixed it. I updated the bios a lot, both banks. I stopped using a Belkin 8 plug surge protector strip and started using an APC 5 plug surge protector strip. After that I moved in to a brand new property built last year. If it continues you should consider selling your house.

 

[Brian Stirling]

Latest Flash dated 1-06-2016

X99-PRO BIOS 2101
Improve system stability.

https://www.asus.com/us/Motherboards/X99PRO/HelpDesk_Download/

I built the box on 1/26/16 and updated the BIOS and drivers -- I'm at 2101.


Brian

 

[westom]

My guess is one or the other of these components is doing this and I don't know which one it is.

PSU powers off or on when ordered by a power controller. With power on, that controller decides if power is stable. Then either powers off or permits the CPU to execute.

BIOS obviously is completely irrelevant since BIOS is ignored until the CPU is permitted to execute. That only happens after the power controller decides to permit it.

Nobody can say anything about your symptoms without numbers that describe how that entire power 'system' operates. PSU is only one part of a 'system'. Useful replies are only possible after providing some three digit numbers from a meter, minutes of labor, and some requested instructions.

You computer must operate fine and happy even when incandescent bulbs dim to 40% intensity. Is your voltage dropping that much? If not, AC mains are irrelevant to your symptoms. To say what is happening requires information (numbers) that only you can first provide.

 

[balloonshark]

The MSI motherboard in my lower signature would boot twice if power was cut to it and when you made a change in the BIOS. It was perfectly normal for that board. Perhaps yours is similar.

 

[Brian Stirling]

PSU powers off or on when ordered by a power controller. With power on, that controller decides if power is stable. Then either powers off or permits the CPU to execute.

BIOS obviously is completely irrelevant since BIOS is ignored until the CPU is permitted to execute. That only happens after the power controller decides to permit it.

Nobody can say anything about your symptoms without numbers that describe how that entire power 'system' operates. PSU is only one part of a 'system'. Useful replies are only possible after providing some three digit numbers from a meter, minutes of labor, and some requested instructions.

You computer must operate fine and happy even when incandescent bulbs dim to 40% intensity. Is your voltage dropping that much? If not, AC mains are irrelevant to your symptoms. To say what is happening requires information (numbers) that only you can first provide.

I have the power strip connected to a Kil-a-wat power meter and under average usage the total power averages about 220W including PC, 24" monitor, Logitech external audio system and an LED lamp. Measuring the items separately then subtracting I get the PC box alone at about 144W.

Under heavy loading as when rendering video the total power will increase to an average of about 395W with peaks to about 460W so subtracting the the monitor etc the PC box max's out about 384W.

The power supply is properly sized at 750W -- I could have gone a bit smaller and many would have gone even bigger so the PS is properly sized for the task and power usage by the PC ranges from 19% to 51% of the PS rating. The power strip is capable of providing about 1800W but the max usage is no more than 460W or less than 26% or circuit capability.

It's my guess that if you maintain power to the PC box when it's off the PS maintains voltages in the caps etc but when you disconnect power the caps drain and when you then try to turn it on there's a voltage dip until the caps charge up.

My next test is to power off overnight then apply power to the PC box but leave it off for 30 minutes or so before trying to power up.


Brian

 

[westom]

The power supply is properly sized at 750W -- I could have gone a bit smaller and many would have gone even bigger so the PS is properly sized for the task and power usage by the PC ranges from 19% to 51% of the PS rating.

Many computer assemblers have no idea how electricity works. Therefore have no idea how to select a PSU. Help lines do not want to teach how electricity works. So we tell computer assembler to obtain a PSU twice the needed wattage. Your numbers demonstrate intentional oversizing. Even your AC mains are many times more powerful than what is required. All that is completely irrelevant to your symptoms and to what I discussed.

Again, a PSU is only one part of a power 'system'. That entire system must be confirmed using a recommended tool and instructions. That will measure something completely separate from what a Kill-A-Watt, PSU wattage, etc address. Nothing in those above measurements come close to or address a power controller and other relevant power 'system' components.

Maintaining power to a powered off PC means one supply inside a PSU is off while another remains on. That still does not help define your symptoms. Observations will not provide any useful answers. Those three digit numbers are the only information that can result in informed assistance.

 

[Brian Stirling]

Many computer assemblers have no idea how electricity works. Therefore have no idea how to select a PSU. Help lines do not want to teach how electricity works. So we tell computer assembler to obtain a PSU twice the needed wattage. Your numbers demonstrate intentional oversizing. Even your AC mains are many times more powerful than what is required. All that is completely irrelevant to your symptoms and to what I discussed.

Again, a PSU is only one part of a power 'system'. That entire system must be confirmed using a recommended tool and instructions. That will measure something completely separate from what a Kill-A-Watt, PSU wattage, etc address. Nothing in those above measurements come close to or address a power controller and other relevant power 'system' components.

Maintaining power to a powered off PC means one supply inside a PSU is off while another remains on. That still does not help define your symptoms. Observations will not provide any useful answers. Those three digit numbers are the only information that can result in informed assistance.

First of all, I'm not some neophyte to the tech industry and have worked in various capacities in the engineering world since the early 80's and that includes designing and building various electrical and electronic systems.

Second, you made reference, a couple posts ago, to needing to know more about the numbers and I provided the numbers I had and you were conspicuously non-specific about just what numbers you needed to form an opinion. So, at the risk of another non-answer answer perhaps you'd be so kind as to provide a hint of the numbers you feel are needed to form that opinion.

Third, a 750W PS is more than needed for my PC but less than many spec for a similar workstation. There wasn't much to be gained going smaller and the rating indicates better than 92% efficiency above 20% usage which is exactly where I'm at. But, do tell me what size supply you'd have recommended? The GPU alone is spec'd at 250W and the CPU about the same so adding them up you get 500W before adding the other items like: 9 fans, 2 water pumps for GPU and CPU, 32GB DDR4 RAM, PCH, PCIe SSD, an Optical drive, and 2x WD Black HD's. As I said the PS is more than needed but not overly so.

I have no idea whatever what you're babbling about on the mains issue -- it's a standard 15A 120V USA household circuit which works out to ... wait for it ... 1800W or 1800VA to be more accurate. I can't for the life of me figure what exactly you would have me do to more properly allocate mains power for the PC.

When power is available to the PS with the PC off there are some circuits that are powered on and others that are off. It's my guess that something within the normally powered portion is tripping off due to voltage dip when power is turned on shortly after power is applied from the mains. I generally prefer to turn off all power to eliminate vampire power loss where possible, but doing so causes this double power on issue.


Brian

 

[westom]

I have no idea whatever what you're babbling about on the mains issue -- it's a standard 15A 120V USA household circuit which works out to ... wait for it ... 1800W or 1800VA to be more accurate.

None of that was relevant. Not even remotely relevant. Stated quite clearly was what is required. You ignored all that to discuss something irrelevant.

Extracted was only what your numbers reported. Interesting that computer assemblers are told to buy PSUs that are double what is required. But your numbers say nothing about your problem - as made clear before you posted. Clearly stated was how to obtain useful three digit numbers. [quote}Nobody can say anything about your symptoms without numbers that describe how that entire power 'system' operates. ... Useful replies are only possible after providing some three digit numbers from a meter, minutes of labor, and some requested instructions.[/quote] I'm sorry you are so confused. But none of those numbers slobbered here are relevant or were asked for. Slobbered because apparently you want emotions - not facts. Now you can change your attitude. Read what was written rather than insult others because you still do not comprehend.

A Kill-A-Watt obviously provides no useful information. It is a superb tool. But nothiing even *implied* using one. Instead, read what is posted.

What I should have said is :"You scumbag. Why did you waste bandwidth with irrelevant numbers." Instead I was kind enough to explain only what those Kill-A-Watt numbers report. Why did you then reply with insult? Do you understand patience - hoping you eventually learn from your mistakes?

A Kill-A-Watt will not report anything constructive. Reason why was defined in my first post. You also did not read that. Three digit numbers are only possible using ... as is written in a previous and and quoted again here.

 

[Brian Stirling]

None of that was relevant. Not even remotely relevant. Stated quite clearly was what is required. You ignored all that to discuss something irrelevant.

Extracted was only what your numbers reported. Interesting that computer assemblers are told to buy PSUs that are double what is required. But your numbers say nothing about your problem - as made clear before you posted. Clearly stated was how to obtain useful three digit numbers. [quote}Nobody can say anything about your symptoms without numbers that describe how that entire power 'system' operates. ... Useful replies are only possible after providing some three digit numbers from a meter, minutes of labor, and some requested instructions. I'm sorry you are so confused. But none of those numbers slobbered here are relevant or were asked for. Slobbered because apparently you want emotions - not facts. Now you can change your attitude. Read what was written rather than insult others because you still do not comprehend.

A Kill-A-Watt obviously provides no useful information. It is a superb tool. But nothiing even *implied* using one. Instead, read what is posted.

What I should have said is :"You scumbag. Why did you waste bandwidth with irrelevant numbers." Instead I was kind enough to explain only what those Kill-A-Watt numbers report. Why did you then reply with insult? Do you understand patience - hoping you eventually learn from your mistakes?

A Kill-A-Watt will not report anything constructive. Reason why was defined in my first post. You also did not read that. Three digit numbers are only possible using ... as is written in a previous and and quoted again here.

Tell you what, I'll leave this as my final comment to you as there would be little point continuing. You ramble on about the mains being more than necessary then tell me none of that matters -- if it didn't matter why mention it?

Your first post said, among other things, "PSU is only one part of a 'system'. Useful replies are only possible after providing some three digit numbers from a meter, minutes of labor, and some requested instructions" and it should be obvious that unless you were more specific the meter you mentioned might well be a power meter, but there's no way to tell from your obtuse post. The part about "three digit numbers" was, once again, unclear. And finally, the reference to "minutes of labor" -- well, I guess I should have seen that you weren't serious from the beginning -- minutes of labor, you might as well have said 'color of the Sun' as that would be no less irrelevant. The point about '...some requested instructions' -- again, I'm left shaking my head at whatever your point was.

I've come to the conclusion you never meant to help and have been trolling all along -- good for you. So, reply if you must but this is my last correspondence with you...


Brian

 

[VirtualLarry]

Yeah, westom's a troll. All he does is inject oblique references to testing your voltages with a multimeter "to three digits", as if that's some magic tea leaves to read to determine if your PSU is bad.

 

[westom]

I've come to the conclusion you never meant to help and have been trolling all along -- good for you. So, reply if you must but this is my last correspondence with you...

Get a meter, request instructions. spend minutes executing those instructions. Then (and only then) are every question answered. Then learn how a computer really works. Instead, you want to attack another because you ignored directions in one sentence and wasted time with a Kill-A-Watt.

One simple sentence creates so much anger because you could not read it? Amazing.

Of course, I have every right to attack you with similar insults. But I don't care. Do what is recommended as if you never had a melt down. Then discover the technically informed ignored the tirade. Either learn or stew in your anger. Your choice.

PSU powers off or on when ordered by a power controller. That very first sentence defined what creates your symptoms - a controller.

 

[MongGrel]

Get a meter, request instructions. spend minutes executing those instructions. Then (and only then) are every question answered. Then learn how a computer really works. Instead, you want to attack another because you ignored directions in one sentence and wasted time with a Kill-A-Watt.

One simple sentence creates so much anger because you could not read it? Amazing.

Of course, I have every right to attack you with similar insults. But I don't care. Do what is recommended as if you never had a melt down. Then discover the technically informed ignored the tirade. Either learn or stew in your anger. Your choice.

I normally wouldn't say STFU in a tech thread, but please STFU.