Capacitors in parallel

[Grizz]

I need to insall two capacitors on parallel....So I'm assuming you connect the positive to the positive and the negative to the negative.
But how do you install them both in to the pc board? Or can I install the one as normal and sort of piggyback the second one by soldering the leads of the one to the other.

thanks,

Scott

[demonstar]

Yes both caps +ve go together and both -ve go together. So the extra one has exactly the same orientation as the first one. It could be awkward trying to get them to stay. Is there no way you can buy a single cap. Could try twisting the leads together or drilling the pcb hole wider to get them both in.

[bkanber]

As demonstar mentioned, it's probably best for you to get just one capacitor with the right value, rather than two caps in parallel. The biggest argument for doing this (other than the obvious logistics of fitting them in there) is that when you have two caps in parallel, not only do their capacitances add, but so do their tolerances. If you have two caps, say both 1 uF +/- 20%, then putting them together yields 2 uF +/- 40%.. 40% tolerance is pretty big, and may be unacceptable in some situations. On the other hand, if you can get a single 2 uF cap, the tolerance will only be 20% (or whatever it's rated at).

This may or may not be a compelling argument, depending on the situation...

[Grizz]

Thanks, I'll try to find the proper cap the next time I order...but I'l go with running the two in parallel for the time being.

[Grizz]

I'm using a perfboard could I just put them side by side?

[Boogdish]

yeah, put them side by side.

[MikeH]

When I do this with caps or resistors, I leave the leads on one alone and just wrap the leads of the other around once up by the cap, solder, trim, and install the way I would normally.  Works great.

[dschwartz]

As demonstar mentioned, it's probably best for you to get just one capacitor with the right value, rather than two caps in parallel. The biggest argument for doing this (other than the obvious logistics of fitting them in there) is that when you have two caps in parallel, not only do their capacitances add, but so do their tolerances. If you have two caps, say both 1 uF +/- 20%, then putting them together yields 2 uF +/- 40%.. 40% tolerance is pretty big, and may be unacceptable in some situations. On the other hand, if you can get a single 2 uF cap, the tolerance will only be 20% (or whatever it's rated at).

This may or may not be a compelling argument, depending on the situation...

r u sure of that?

20% tolerance means that the value is X*(1-+0.2)
if you parallel 2 equal resistors, the resulting capacitance is X*(1+-0.2) + X*(1-+0.2)= 2X*(1-+0.2)   with is still 20% tolerance

[raulgrell]

dschwartz is right... Percentage tolerance only adds when the values are multiplied, not added. The ABSOLUTE tolerance adds, but the percentage tolerance stays the same.

[amz-fx]

Yes, the tolerance percentages do not add together...  it is easy to calculate: if each 1uf (nominal) capacitor can vary from 0.8uf to 1.2uf then the values of two in parallel will be 1.6uf (lowest) to 2.0uf (nominal) to 2.4uF (highest)....  which is +/- 20%

However, two capacitors in parallel have a lower ESR...  because the resistances are in parallel just like the capacitance, so if the ESR is 20millohm then two identical caps in parallel would have a 10milliohm ESR...  which is the main advantage to using two caps instead of one, as well as reducing parasitic inductance.

This is the theory behind the Multicap which is equivalent to 10 capacitors in parallel in a single component.

regards, Jack

[alanlan]

... as well as reducing parasitic inductance.

I wouldn't worry too much about parasitic inductance in the frequency ranges we're operating at though.

[R.G.]

NB
Nor ESR at the current ranges and frequencies used. Nor lead inductance. Nor package color.