Overall capacitance

[ReelMonza]

Noob here...

Is it possible to measure the overall capacitance of a particular mobo ?

I mean, it has (6) 3300uF 6.3v and (6) 1500uF 6.3v

Is there a way to know the overall capacitance values for that particular board ?

I read that if they're in parallel the overall capacitance will be the sum of all capacitances, while in series the overall capacitance will be the inverse sum of the applied capacitances, is that right ?

Thanks in advance.

[MD Willington]

You're correct about parallel and series caps.

http://www.allaboutcircuits.com/vol_1/chpt_13/4.html

MD

[ReelMonza]

Thanks, MD Willington.

Now how do I know if the caps are connected in series or in parallel ?

[willawake]

i am curious about your question. i think you want to install less caps on the board basically, using higher values instead. But you should consider that all caps are not connected as one circuit, groups of them or single caps are providing service to different areas of the board. without the board schematics or without good testing, you could be making a mistake and leaving a trace open if you dont install a cap in that position.

we already covered dealing with a row of caps which are connected in parallel in a previous post and i will recap here. This is a mod which is easier to use.

gg1978 said regarding an ABIT VL6 :

The first picture best illustrates what i did. The group of 5 caps on the outside edge of the board, between the ATX PWR connector and the CPU fan header, i replaced with (2) 3300uF Nichicon UPW's, and (1) 1500uF Nichicon UPW. The 3300's are the large can diameter, so there was no way to fit four on there, so i substitued the pair of 3300's for the (4) 1500's that were in those locations before. Then i put the 1500 in the other place. I believe all those are in parallel, so all of them could have been replaced by a 4700 and a 3300. Gives 600uF extra capacitance, and IIRC, that's the input side of the CPU VRM, although i would have to verify that with my DMM to be sure.

Given how shitty the original caps were, and how bad the ESR was on them, i doubt the possibly slightly higher ESR from using 3 caps instead of 5 will affect anything. The board will run 150FSB just fine..

KC8 explained how to place the caps :

all the + leads are tied together and so are all the - just install a cap in every other position.i have done this myself on my own boards. the new high quality caps are much lower esr than the taiwan cheapies ever thought of being.

It would be good if people could contribute whether we should always expect a row of mobo caps to be connected in paralllel. I think this would be the case.

[gg1978]

regarding the VL6

Hey all,

I verified all of the cap positions were in parallel before i did the "reduced # of caps" mod. As a general rule, most boards that have a bunch of caps in a "group" do tend to have them in parallel.. That's a general rule though, not a law..

My Abit BE6-II appears to have a row behind the CPU slot, and going by the fact that they're all in a row, you would assume they're all for the CPU VRM.. In fact, i doubt that, as the VRM topology of the board includes a variable +3.3V nominal VRM for the northbridge and SDRAM.. That VR appears to be behind the CPU slot, so presumably some of the caps there are for the memory/Northbridge VRM, not only the CPU. So assuming the caps all go in the same group can be wrong at times.. Most modern mainboards don't have the DDR RAM VRM close to the CPU VRM, instead putting it close to the memory slots themselves, so doing cap "grouping" on them tends to be easier. I'd still recommend verifiying the caps in a given group are all in parallel before recapping with less caps.

I still need to finish recapping the VL6 though..