Capacitor Mounting Question

[ceulrich]

I am building an amplifier based on instructions found in “A Complete Guide to Design and Build A Hi-Fi LM3886 Amplifier.” I am designing my own PCB, since I have not found a commercial one that will fit my existing heatsink and chassis. The design calls for a 4.7 µF input coupling capacitor. I am debating between using a quality electrolytic (Nichicon UFW) or a film type (MultiCap) mounted vertically – horizontal mounting is not a option due to board size constraints. The film type is probably preferred, but I am concerned about the vertical mounting. Any thoughts or advice would be appreciated.

Cheers,

ceulrich

 

[JMFahey]

You can vertically mount it, but it´s good Engineering to "fix" it to the board with some kind of glue (hot melt/contact cement/electronics grade silicone) to protect it against vibration.

Since it´s a relatively heavy component, and mounted on edge hanging from just one flimsy leg, vibration may eventually break it, so the glue avoids that.

 

[rayma]

If you do mount the cap vertically, add a spacer under the end with the short lead.
Use around a 1/4" spacer.

 

[scottjoplin]

The lead from the top creates a small loop and could be susceptible to picking up EMI.

 

[DF96]

You should not consider a 4.7uF electrolytic and a 4.7uF film cap as necessarily alternatives. It is quite possible that a film cap could have a smaller value, and so be easier to mount. What LF rolloff are you aiming at? What is the input resistance of the circuit?

 

[ceulrich]

Thanks everyone for your comments.

JMFahey: Yes, I will definitely do that.

rayma: I do not understand the reason for the spacer. My first thought is that this would make the component less stable. Could you please clarify?

scottjoplin: I never thought about that. Would it make any difference if that lead was wired to be the input or the output of the capacitor?

DF96: Yes, good thought. The design calls for 4.7µF with an input resistance of 20K, which puts the -3dB point at 1.69Hz, probably a bit of over kill. I happen to have some 3µF MultiCap units in my parts collection, so I was going to use them. That would put the -3dB point at 2.65Hz, still quite low. I could go down to 1µF for a -3dB point of 7.96Hz, but I still have a component length problem, at least with the parts I have.

Cheers,

ceulrich

 

[JeffYoung]

Have you considered a stacked film rather than a wound one? They're generally a lot smaller, particularly the polyethylene or polyphenylene sulfide ones.

 

[DF96]

0.47uF for a rolloff around 15Hz might be better.

Alternatively, just use the electrolytic. If you are worried about distortion then use a slightly higher value, say 10uF or 22uF.

 

[ceulrich]

JeffYoung and DF96: Good ideas, I would consider that if I did not have these MultiCaps laying around.

Thanks again for your suggestions.

Cheers,

ceulrich

 

[nigelwright7557]

0.47uF for a rolloff around 15Hz might be better.

Alternatively, just use the electrolytic. If you are worried about distortion then use a slightly higher value, say 10uF or 22uF.

I tend to use 10uf or 470nf poly caps on the input.
Never had any problems with either.