odd motherboard voltage reading

[Per Hansson]

Re: odd motherboard voltage reading

jbanter; yes those are the ones
I just went with logic as not all can be seen int he pic, so if one is missing it's just me assuming things incorrectly

The board BIOS may be set to boot always when the power is resumed, or only if the previous shutdown was due to a power outage (i.e. you janking the cord since there is no button to turn off the computer with)
There is also a possibility that the small caps around the chipset area (below 330uF) are out of spec, this could also make it turn on by itself

[PCBONEZ]

Re: odd motherboard voltage reading

I think you mean remove and reinsert or reinstall.. Replacing it is kinda excessive..

English is your second language?
Let me help you then..

replace [rɪˈpleɪs]vb (tr)
1. to take the place of; supersede the manual worker is being replaced by the machine
2. to substitute a person or thing for (another which has ceased to fulfil its function); put in place of to replace an old pair of shoes
3. to put back or return; restore to its rightful place

[PCBONEZ]

Re: odd motherboard voltage reading

Another thing what can cause a motherboard to power on by itself, when the switch at the back of the PSU is turned on? The front panel connector is not even connected to the motherboard.

Two things:

- A BIOS setting that tells it to either start or return to previous condition after a power loss.

- A short in the circuit the switch goes to.
Look at the drawing. Note the IC pin marked with an X. When the switch is open the voltage on X is high, near +5vsb voltage. When the switch is shut the voltage on X goes low, probably less than 1v. [This is called to 'pull down' the pin.] When pin X is pulled down the IC sends a signal to the PSU for it to start. [The IC is usually integrated into the chipset.] The PSU then has roughly 50 milliseconds to get voltages within 80% or it will initiate a protective shutdown. [Turn itself back off because 'something' is wrong if the volts don't come up.] - - - Anyway, if pin X or the resistor between it and the switch has a ground it will 'think' it is pulled down... You see?
.

[PCBONEZ]

Re: odd motherboard voltage reading

VRM MOSFETS? It sounds like something I read before, can't remember now. Where is it located?
Suppose it has traits you mentioned, is it replaceable?

They are replaceable but they can be a bugger to get off when they are soldered on flat like they do these days. The metal chunk at one end is the heat-sink. The whole back side of that is soldered down so it takes a lot of heat.
.

[jbanter]

Re: odd motherboard voltage reading

- A short in the circuit the switch goes to.
Look at the drawing. Note the IC pin marked with an X. When the switch is open the voltage on X is high, near +5vsb voltage. When the switch is shut the voltage on X goes low, probably less than 1v. [This is called to 'pull down' the pin.] When pin X is pulled down the IC sends a signal to the PSU for it to start. [The IC is usually integrated into the chipset.] The PSU then has roughly 50 milliseconds to get voltages within 80% or it will initiate a protective shutdown. [Turn itself back off because 'something' is wrong if the volts don't come up.] - - - Anyway, if pin X or the resistor between it and the switch has a ground it will 'think' it is pulled down... You see?
.

I had to read few times but I think I sort of get it. The drawing does help though, thanks.
The bios setting is set to off after power failure. What are the likely suspects for this ground?

[jbanter]

Re: odd motherboard voltage reading

jbanter; yes those are the ones
I just went with logic as not all can be seen int he pic, so if one is missing it's just me assuming things incorrectly

The board BIOS may be set to boot always when the power is resumed, or only if the previous shutdown was due to a power outage (i.e. you janking the cord since there is no button to turn off the computer with)
There is also a possibility that the small caps around the chipset area (below 330uF) are out of spec, this could also make it turn on by itself

I'll have a look at such caps but the markings are so small on those smaller caps. It's a bit tricky.

[PCBONEZ]

Re: odd motherboard voltage reading

I had to read few times but I think I sort of get it. The drawing does help though, thanks.
The bios setting is set to off after power failure. What are the likely suspects for this ground?

There is an unseen resistor [or resistance] inside the IC chip between X and ground. Compared to the internal resistance the two resistors outside the IC chip are small and the one on the +5vsb side is the larger of the two.
- With the switch open the current path is through the 'X -to- +5vsb resistor' and the unseen resistance. The unseen resistance is higher so most of the voltage drops arcoss it and the voltage on X goes high.
- With the switch shut the primary current path is through the 'X -to- +5vsb resistor' and the 'X to ground resistor' because the 'X to ground resistor' is a very small ohm resistor. The 'X to ground resistor' doesn't drop much voltage and that pulls X nearly down to ground potential.

All the other things that will start the PC [wake-on-lan, wake-on-ring, wake-on-mouse, etc] do the same thing as the switch in that they put a ground on X through a small resistance.

- As the board has been worked on I would first look for solder splats or solder hairs shorting PCB traces. That would be the most likely suspect. [And about the only easy one to find.]
- Next look for SMD parts that shifted when caps were replaced.

- It's possible a bad small cap has failed in such a way it is effectively replacing the switch. [Short on ground side of X making X low.]
- It's possible a bad small cap has failed in such a way that it is pulling a whole segment of the +5vsb distribution down to near ground. [Short on power side of X making X low.]
- It's possible a bad cap in the power to the IC chip [or that section of the chipset] is affecting it's power and so affecting it's operation.
- It's possible a bad cap in the Reference Voltage for the IC chip [or that section of the chipset] is making the Reference Voltage wrong. The IC compares the voltage at X to some other voltage to know if it's high or low. That other voltage is the Reference Voltage.
- A bad small cap can also hose Vtt. Vtt is kind'a sort'a the power voltage IC chips use to make siganls to other chips. [Flashlight Anaolgy: Vtt is the battery, the IC chip the signal comes from is the switch, the IC chip the signal goes to is the bulb.]

Most of those are a real pain to circuit trace and track down. At some points you may find you are looking up IC schematics to figure out what to check next.

That's why I just replace the 85C caps when I see them. I don't want to deal with those kinds of time-eating problems. When you already have one of those problems and 85C caps present just replace them. It's not 100% certain to be the fix but it will -probably- fix the problem and save loads of time.

.

[PCBONEZ]

Re: odd motherboard voltage reading

.
Someone just showed up with a blown MOSFET like I mentioned earlier.
Happens but not too common.
https://www.badcaps.net/forum/showthread.php?t=10400
.

.

[kaniki]

Re: odd motherboard voltage reading

3. to put back or return; restore to its rightful place

Hmm.. that would be the original packaging or box.. right???? Wait, we are talking about an intel processor right... OK.. in the garbage then... AMD RULES!!!! HA HA HA..

j/k all... but had to pick...

[PCBONEZ]

Re: odd motherboard voltage reading

I haven't used AMD for my own gear since Slot A.
I'm not feeling a need to switch back.

[jbanter]

Re: odd motherboard voltage reading

Most of those are a real pain to circuit trace and track down. At some points you may find you are looking up IC schematics to figure out what to check next.

That's why I just replace the 85C caps when I see them. I don't want to deal with those kinds of time-eating problems. When you already have one of those problems and 85C caps present just replace them. It's not 100% certain to be the fix but it will -probably- fix the problem and save loads of time.

That's a handful of caps to replace.
Thanks for explaining.

By the way, I found out what caused my voltage spike. It turns out reseating the CPU and make sure it is fastened before locking it, fix the issue. At least so far, and I hope it stays that way.

[kaniki]

Re: odd motherboard voltage reading

nice to hear you found the problem.. good luck