[SOLVED] Filing empty capacitor spaces on MB, good idea ?

RobbeyRobert

Reputable
May 10, 2015
92
0
4,660
9
Hi, to put this straight forward, what positive or even negative impacts could there be if you fill ALL capacitor holes on a PC motherboard.

Will it help in OC, PC stability or even increasing it's life capacity ?? Thanks. sorry if if my english isn't so good.
 
Firstly: which empty capacitor spaces do you speak of since they can be left empty for any of several reasons.

One reason is circuitry that would be needed on a different model board. Filling those spaces will do nothing as it would connect nothing to nothing unless you fill the other empty spaces or replace some components with different components.

But most likely those spaces are unneeded decoupling caps. When a board is in design engineers will often times scatter pad locations for decoupling caps on the off-chance they may be needed. If they left them out they were found to be unnecessary and adding them will do nothing. They generally leave in far more than are necessary and removing a bunch of them would have equal effect...which is to say nothing at all.

But these things are usually tiny MLCC's and hard to hand solder. If you make an error soldering them in, such as a leave a solder bridge, you'll short out what is most likely a power trace and destroy the board. And possibly your CPU and PSU in the bargain.
 
Firstly: which empty capacitor spaces do you speak of since they can be left empty for any of several reasons.

One reason is circuitry that would be needed on a different model board. Filling those spaces will do nothing as it would connect nothing to nothing unless you fill the other empty spaces or replace some components with different components.

But most likely those spaces are unneeded decoupling caps. When a board is in design engineers will often times scatter pad locations for decoupling caps on the off-chance they may be needed. If they left them out they were found to be unnecessary and adding them will do nothing. They generally leave in far more than are necessary and removing a bunch of them would have equal effect...which is to say nothing at all.

But these things are usually tiny MLCC's and hard to hand solder. If you make an error soldering them in, such as a leave a solder bridge, you'll short out what is most likely a power trace and destroy the board. And possibly your CPU and PSU in the bargain.
 

RobbeyRobert

Reputable
May 10, 2015
92
0
4,660
9
Firstly: which empty capacitor spaces do you speak of since they can be left empty for any of several reasons.

One reason is circuitry that would be needed on a different model board. Filling those spaces will do nothing as it would connect nothing to nothing unless you fill the other empty spaces or replace some components with different components.

But most likely those spaces are unneeded decoupling caps. When a board is in design engineers will often times scatter pad locations for decoupling caps on the off-chance they may be needed. If they left them out they were found to be unnecessary and adding them will do nothing. They generally leave in far more than are necessary and removing a bunch of them would have equal effect...which is to say nothing at all.

But these things are usually tiny MLCC's and hard to hand solder. If you make an error soldering them in, such as a leave a solder bridge, you'll short out what is most likely a power trace and destroy the board. And possibly your CPU and PSU in the bargain.
--
The two empty spaces are right under my CPU cooler, between the CPU and VRM, both of the holes are rated for 6V capacitors.
Thought if i filled those with some caps it would have an increase on system stability and durability or atleast some benefit, i can remember that i've read somewhere at some time in the past, even here on the forum were a lot of discussions, that they used to fill those emtpy spaces to gain better system stability, it was referenced a lot with OC'ing. Thanks !

Also, a side by question, i have 4 VRM's chips or whatever they are without a cooling system on them, right near the ones with a big cooler on their heads, those VRM's tend to get so hot at times that when touched, it can literally injur your skin. i would say above 80Celsius. Can i put a simple cooler on them without thermal pad or do i need a thermal pad to draw the heat and forward it toward the cooler, or should i leave them be ? they don't seem to be of silicon material like normal chips do.
 
Absolutely not you are not a computer engineer you did not design the circuits or the motherboard and you have no idea where capacitors are needed and where they are needed. Under any circumstances never do this.
 
--
The two empty spaces are right under my CPU cooler, between the CPU and VRM, both of the holes are rated for 6V capacitors.
Thought if i filled those with some caps it would have an increase on system stability and durability or atleast some benefit, i can remember that i've read somewhere at some time in the past, even here on the forum were a lot of discussions, that they used to fill those emtpy spaces to gain better system stability, it was referenced a lot with OC'ing. Thanks !

Also, a side by question, i have 4 VRM's chips or whatever they are without a cooling system on them, right near the ones with a big cooler on their heads, those VRM's tend to get so hot at times that when touched, it can literally injur your skin. i would say above 80Celsius. Can i put a simple cooler on them without thermal pad or do i need a thermal pad to draw the heat and forward it toward the cooler, or should i leave them be ? they don't seem to be of silicon material like normal chips do.
Forget about putting the capacitors in the space below the CPU. Just forget...it's not gonna help and be far more likely that you'd do damage in that area in particular.

As far as the VRM. If the chips you refer to are the FET's then yes, putting some heatsinks on naked FET's can often times assist in stability when operating at the limit. Go looking for things like these:

https://www.amazon.com/copper-memory-heatsink/s?k=copper+memory+heatsink

and double-sided thermal tape.

But if you're instead touching the chokes, or as some call them inductors, then putting heatsinks on them won't help anything.
 
Reactions: RobbeyRobert

Azzyasi

Distinguished
Jan 24, 2011
119
10
18,715
16
Really bad idea!

In an optimistic way, if you had the electronics know-how and the schematics in front of you we can argue what the non-present components might be used for. Without a diagram you can't read the traces since it's likely triple and quadruple layer.
I bet even with the actual diagram in front you/me/most on this forum might not have adequate knowledge to speculate what the function of that capacitor is.
 
Reactions: RobbeyRobert
Really bad idea!

In an optimistic way, if you had the electronics know-how and the schematics in front of you we can argue what the non-present components might be used for. Without a diagram you can't read the traces since it's likely triple and quadruple layer.
I bet even with the actual diagram in front you/me/most on this forum might not have adequate knowledge to speculate what the function of that capacitor is.
I do happen to know a LITTLE bit about the functioning of the VRM. And the place he's talking about sounds suspiciously like the output filtering capacitors of the VRM. Now what I am aware of is those output capacitors, acting along with the series inductors and resistors, form an RCL network that has significant influence on the switching frequency of the buck regulator circuit.

Now even not knowing exactly how OP's buck regulator functions I do know switching frequency has a terrible lot to do with the regulator's efficiency. So alter it in an unthinking manner and it will be far more likely to ruin efficiency and that means increasing heat output even more and that could only degrade performance. That's why just don't even think about adding or changing values of capacitors in that area...unless you have a great knowledge of buck regulator design.
 
Reactions: RobbeyRobert

RobbeyRobert

Reputable
May 10, 2015
92
0
4,660
9
I do happen to know a LITTLE bit about the functioning of the VRM. And the place he's talking about sounds suspiciously like the output filtering capacitors of the VRM. Now what I am aware of is those output capacitors, acting along with the series inductors and resistors, form an RCL network that has significant influence on the switching frequency of the buck regulator circuit.

Now even not knowing exactly how OP's buck regulator functions I do know switching frequency has a terrible lot to do with the regulator's efficiency. So alter it in an unthinking manner and it will be far more likely to ruin efficiency and that means increasing heat output even more and that could only degrade performance. That's why just don't even think about adding or changing values of capacitors in that area...unless you have a great knowledge of buck regulator design.
Initially i've wanted to do this operation on a xeon machine, but the risks are too high and so are the costs for that thing. Now, just for fun and out of plain curiosity i am going to ignore all your warnings and fill another 2 empty capacitor holes on an AMD machine, again right next to the cpu. Want to find out what impacts it's going to have on heat out put and performance, i don't really expect to have my PC blown away... but that's still to be debated.

If nothing happen i'll let them sit there but if it's going to work hotter or even performance issues then i just simply will remove them.
 

Azzyasi

Distinguished
Jan 24, 2011
119
10
18,715
16
Have fun!
Post results here please, I'm curious about results.
Experimenting with stuff you afford to mess and don't care too much about is a path. I usually do my homework before and try something that I mostly will know what to expect from in theory.
 

ASK THE COMMUNITY

TRENDING THREADS