add pigtails and lay capacitors sideways on mobo?

[bulk88]

Can I unsolder capacitors, solder pigtails onto caps, solder pigtails back into the mobo, and basically move then 3/4 inch and lie them sideways on a mobo or is the wire/trace length of the caps important to the CPU voltage regulator circuitry?

Caps are uncomfortably bent from a heatsink upgrade on a dual Socket A mobo system i'm working on.

 

[kayson]

Can I unsolder capacitors, solder pigtails onto caps, solder pigtails back into the mobo, and basically move then 3/4 inch and lie them sideways on a mobo or is the wire/trace length of the caps important to the CPU voltage regulator circuitry?

Caps are uncomfortably bent from a heatsink upgrade on a dual Socket A mobo system i'm working on.

The wire length is important. Longer wires means larger inductance which essentially means the caps cant "react" as fast to voltage changes. But whether that'll cause issues is dependent on a million things. If you have through-hole capacitors and its an older motherboard, I doubt it would be a problem at all.

 

[TazExtreme3]

The leg length isn't critical in motherboard applications. They are in some audio applications though. (don't quote me on that, but I think it's correct)
I would personally buy new caps and orient them where you want them with the heatsink installed, remove the caps and place heatshrink on the legs to insulate the exposed legs atop the mobo where the extra length was required, fit them back in the holes, then solder them in and clip any extra leg material.

 

[cyberfish]

Depending on their use in the circuit, the length may be important. For example, like kayson pointed out, lead inductance can make a decoupling cap pretty useless. The added resistance can also increase the time constant significantly. It can even cause things to oscillate, etc. Bad things can happen.

If they are energy bank capacitors, it's probably fine. But you have no way to tell short of analyzing the circuit.

EDIT: and oops. 2 months thread revival FTW!

 

[Bobnova]

PWM switching speed (and hence, the frequency of wave the caps are there to smooth) varies between 200khz and 1mhz, if that tells you EE people anything about lead length importance.
I have absolutely no idea, personally.

 

[kayson]

PWM switching speed (and hence, the frequency of wave the caps are there to smooth) varies between 200khz and 1mhz, if that tells you EE people anything about lead length importance.
I have absolutely no idea, personally.

It's actually not the pwm switching speed that makes it an issue. The feedback in the regulator circuitry would compensate for that. The problem is when you get a current spike from the cpu. It has high frequency components that make inductance matter.

In any event, a 200khz-1mhz signal wouldn't really have an issue with the longer lead length.