Is This VRM?

[HexaPro]

I just got another broken ASUS board, there's 5 bulging caps on it. They're UCC KZG 6.3V 3.300uF D10mm. I see that the caps is in VRM, but i need to make sure if it's right. I also need advice what caps can be use to replace them other than 3.300uF. Those 5 caps are the only 6.3V 3.300uF D10mm caps on the board.

Btw, i've seen more UCC caps bulging and never seen a bulging Nichicon. Is it UCC is more inferior than Nichicon?

[digge]

Re: Is This VRM?

The KZG series are plagued between certain years, someone else that remembers them might fill in.

To make it simple to choose new ones you can just get new KZG of the same size and everything. The fault was due to manufacturing error so if you buy new ones they should be fine. If im not misstaken HM have more or less the exact same spec, dont quote me on that though.

The more trickier route is looking at the spec sheet of KZG and look for a similar cap of another brand. The values you want to look at and match are ESR and ripple, apart from the normal values printed on them (volts and uF). What you replace with can be different depending on what you have access to.

[PCBONEZ]

Re: Is This VRM?

digge is mixing up two different things.

Nichicon HM ans HN were defective from 2001-2004 due to the wrong amount of electrolyte being added.
[You can find 100's of photos of bloated ones if you care to look.]

Chemicon KZG and KZJ use 'an advanced electrolyte' THAT SUX, and all of them are bad.
[There are two other Chemicon series that were rather rare [not many used, don't show up often] that seem to have the same problem.
Those are TMZ and TMV. Those do show up on Asus boards now and then.]

How long any of those manage to last depends largely on hours used and heat.
Sometimes they last years, other times months.

~~

On some boards you simply have to use a multimeter to know for sure what caps are in a VRM.
The group of three is obviously in VRM Vcore.
Those other two may or may not be. By location - can't tell for sure.

Use a DMM, if any of the two are in parallel with the three then they are in Vcore.
My guess would be that one is and the other isn't - but that's ONLY a guess.

Also note: It's not all that unusual for VRM caps to be mixed sizes.
They sometimes do that in Vcore for CPU heatsink clearance.
They also sometimes mix Lytics and Polys in either or both VRM-Vcore and VRM-High.
.

[digge]

Re: Is This VRM?

Opps, my bad, i stand corrected

Should have taken my coffe before replying

[HexaPro]

Re: Is This VRM?

The KZG series are plagued between certain years, someone else that remembers them might fill in.

To make it simple to choose new ones you can just get new KZG of the same size and everything. The fault was due to manufacturing error so if you buy new ones they should be fine. If im not misstaken HM have more or less the exact same spec, dont quote me on that though.

The more trickier route is looking at the spec sheet of KZG and look for a similar cap of another brand. The values you want to look at and match are ESR and ripple, apart from the normal values printed on them (volts and uF). What you replace with can be different depending on what you have access to.

Thanks digge

On some boards you simply have to use a multimeter to know for sure what caps are in a VRM.
The group of three is obviously in VRM Vcore.
Those other two may or may not be. By location - can't tell for sure.

Use a DMM, if any of the two are in parallel with the three then they are in Vcore.
My guess would be that one is and the other isn't - but that's ONLY a guess.

And how exactly do i do this? I've got a DMM but it's just a normal Digital Multi Meter and not a Low ESR meter. Can you tell me how to check it with a DMM if they are paralleled?

Also note: It's not all that unusual for VRM caps to be mixed sizes.
They sometimes do that in Vcore for CPU heatsink clearance.
They also sometimes mix Lytics and Polys in either or both VRM-Vcore and VRM-High.
.

I think i've seen that kind of combination a few times on some board. But will it work if it's mixed like that? What to consider before doing this mixed thing? Do you think that the caps on the left side are in the VRM too? I think it's the 12V rail, because it's 4x 16V 1.200uF Rubycon ZL.

I'm really frustrated here, this thing happened right after i order enough 6,3V 3.300uF caps just for one board while there are 3 board to fix (this board has a twin and that twin also has the same bulging problem with it's VRM but doesn't show any problems yet), and that Samxon caps will arrive after one week. What other reason for bulging cap beside hours used and heat? Bad PSU or elctricity?

[ipman]

Re: Is This VRM?

You can check if they are in parallel with the DMM.
Select continuity/buzzer or ohm scale. Put one probe on one pin of the capacitor, acknowledge which one is it - plus or minus, and then touch the other probe to the plus or minus of the 3 caps group above.
If it beeps (in continuity mode) or shows low resistance (close to zero) then they are in parallel.

[PCBONEZ]

Re: Is This VRM?

And how exactly do i do this? I've got a DMM but it's just a normal Digital Multi Meter and not a Low ESR meter. Can you tell me how to check it with a DMM if they are paralleled?

There is no such animal as a low ESR DMM.
It's a simply resistance check just like checking if anything else is in parallel.
You are checking continuity [resistance checks] from one cap to another through the motherboard, not through the caps.
.......
If + side cap#1 has zero resistance to + side cap#2 ~ then the + sides are connected through the board.
~and~
If the - side of cap#1 has zero resistance to the - side of cap#2 ~ then the - sides are connected through the board.
...
When both sides are connected through the board they are in parallel.
It's that easy.
.
All the caps in pic are in parallel no matter where they are on the board.
All the + 'solder pads' [red] are connected to all the + solder pads.
All the - 'solder pads' [blue] are connected to all the - solder pads.
.

[PCBONEZ]

Re: Is This VRM?

I think i've seen that kind of combination a few times on some board. But will it work if it's mixed like that? What to consider before doing this mixed thing? Do you think that the caps on the left side are in the VRM too? I think it's the 12V rail, because it's 4x 16V 1.200uF Rubycon ZL.

I think you need to study VRMs a little bit.

The VRM-high [PSU side] caps and the VRM-low [CPU side, aka Vcore] caps are completely different situations.
They use different caps and they aren't interconnected as you seem to think.
.
Mixing caps on either high or low is fine as long as there is enough combined uF, high enough Ripple for each cap, and low enough combined ESR.

Look at the drawing here some more.
https://www.badcaps.net/forum/showpo...9&postcount=19
.

[Uranium-235]

Re: Is This VRM?

looks like a dual-phase vrm, the three ontop is probably the input VRM (from the PSU) and the ones on the side are the output vrm (to the CPU)

I can see a small cap and mosfet that are probably the vrm for the northbridge, not sure what those other two are for. but if they're KGZ, its probably best just to replace them with something similar (with simlialr values Panasonic or sanyo) and leave it at that

[HexaPro]

Re: Is This VRM?

You are checking continuity [resistance checks] from one cap to another through the motherboard, not through the caps.

So i have to stick the test probe to the MB line and not the soldered caps lead right?

If + side cap#1 has zero resistance to + side cap#2 ~ then the + sides are connected through the board.
~and~
If the - side of cap#1 has zero resistance to the - side of cap#2 ~ then the - sides are connected through the board.
...
When both sides are connected through the board they are in parallel.
It's that easy.
.
All the caps in pic are in parallel no matter where they are on the board.
All the + 'solder pads' [red] are connected to all the + solder pads.
All the - 'solder pads' [blue] are connected to all the - solder pads.
.

Phew... I was testing the caps like ipman told me and i got confused my self coz every caps lead on the board that i test are beeping. Stupid me

I think you need to study VRMs a little bit.

That's what i'm trying to

The VRM-high [PSU side] caps and the VRM-low [CPU side, aka Vcore] caps are completely different situations.
They use different caps and they aren't interconnected as you seem to think.

Mixing caps on either high or low is fine as long as there is enough combined uF, high enough Ripple for each cap, and low enough combined ESR.

Look at the drawing here some more.
https://www.badcaps.net/forum/showpo...9&postcount=19
.

Now i begin to understand your previous diagram, so the 16V act as VRM-high filter and the 2,5V-6,3V act as Vcore filter. And combining different caps value is ok as long i use the same or higher voltage and also have a enough combined uF and low ESR, and each cap must have enough Ripple to stay flat and alive.

Anyway i've replace the badcaps with used OST caps from a dead ASRock board, i know OST is a badcap but it's emergency and my customer can't wait another week. I found out that the board has 10mm and 8mm through hole option but the only caps i've prepared are those 10mm OST. The OST caps are still fine (no sign of bulging) but i was worried those OST wont work though, but now they work well with the ASUS board and it has stop restarting by itself. But there's a new problem, the PC runs very slow. I ran SpeedFan 4.42 and i think there something wrong with the PSU voltage. I've attached the picture too, what do you think BONEZ?

[PCBONEZ]

Re: Is This VRM?

So i have to stick the test probe to the MB line and not the soldered caps lead right?

Stick test probe in the solder for the cap.
Use a pointy probe to get through the oxide and/or conformal coating.

Now i begin to understand your previous diagram, so the 16V act as VRM-high filter and the 2,5V-6,3V act as Vcore filter. And combining different caps value is ok as long i use the same or higher voltage and also have a enough combined uF and low ESR, and each cap must have enough Ripple to stay flat and alive.

Yes.

Anyway i've replace the badcaps with used OST caps from a dead ASRock board, i know OST is a badcap but it's emergency and my customer can't wait another week.

Customers are always having an emergency and if you use cheap caps they will have another one.
In their mind the second emergency will be your fault.
[They think like that even if you tell them it's coming.]

But there's a new problem, the PC runs very slow. I ran SpeedFan 4.42 and i think there something wrong with the PSU voltage. I've attached the picture too, what do you think BONEZ?

Don't use SpeedFan fan for that. It relies on sensors that cost like 5 cents.
Check the PSU voltages [with the mobo at the BIOS screen and with at least one drive attached to PSU for some more load] by sticking probes from your DC Volt-meter where the wires go into the PSU-to-Mobo connectors. [Some PSUs don't regulate well without enough load. It's easier to attach a drive to start for any PSU than to figure out if the PSU needs one.]
- If the PSU is bad just change it for now because repairing PSUs can eat up some time and doesn't always work. You can look at the PSU later.
.
Also check the CMOS battery. [Should be 3 volts or so.]
SpeedFan says it's flat so you should at least check.
.

[HexaPro]

Re: Is This VRM?

Customers are always having an emergency and if you use cheap caps they will have another one.
In their mind the second emergency will be your fault.
[They think like that even if you tell them it's coming.]

I'll replace it when my Samxon's arrived this week, i think those OST would last alot longer than just a week.

Check the PSU voltages [with the mobo at the BIOS screen and with at least one drive attached to PSU for some more load] by sticking probes from your DC Volt-meter where the wires go into the PSU-to-Mobo connectors. [Some PSUs don't regulate well without enough load. It's easier to attach a drive to start for any PSU than to figure out if the PSU needs one.]
- If the PSU is bad just change it for now because repairing PSUs can eat up some time and doesn't always work. You can look at the PSU later.
.
Also check the CMOS battery. [Should be 3 volts or so.]
SpeedFan says it's flat so you should at least check.
.

Ok, i'll try that tomorrow. Do you have any diagram for 24-pin ATX2 pin out? Incomplete info = disaster, right :P

[digge]

Re: Is This VRM?

Ok, i'll try that tomorrow. Do you have any diagram for 24-pin ATX2 pin out? Incomplete info = disaster, right :P

http://pinouts.ru/Power/atx_v2_pinout.shtml

[PCBONEZ]

Re: Is This VRM?

http://www.smpspowersupply.com/connectors-pinouts.html
Remember those are shown pin-side so for cable-side you will have to reverse right and left.
If cable is standard colors:
Yellow = +12v
Red = +5v
Orange = +3.3v
Purple = +5vsb
Blue = -12v
White = -5v [optional now]
Black = ground wires
Pin 8 [gray] and pin 16 [green] are control signals and don't need checked for this.

Also check at the 2x2 pin "P4 Connector" they are calling +12v2.
.

[HexaPro]

Re: Is This VRM?

http://pinouts.ru/Power/atx_v2_pinout.shtml

Thanks digge

http://www.smpspowersupply.com/connectors-pinouts.html
Remember those are shown pin-side so for cable-side you will have to reverse right and left.
If cable is standard colors:
Yellow = +12v
Red = +5v
Orange = +3.3v
Purple = +5vsb
Blue = -12v
White = -5v [optional now]
Black = ground wires
Pin 8 [gray] and pin 16 [green] are control signals and don't need checked for this.

Also check at the 2x2 pin "P4 Connector" they are calling +12v2.
.

Nice info, it's really detail PCBONEZ. I've changed the caps with Samxon GA 6,3V 3.300uF and I add one more cap to the empty post on the upper side. Now everything is good, and the slow loading problem was caused by faulty either ODD or IDE cable, everytime I plug the Optical Drive the PC would run very slow. So I change it to a newer SATA DVD-RAM and it's working smoothly now.

[PCBONEZ]

Re: Is This VRM?

Glad you got it running. - Congrats!