2
\$\begingroup\$

I recently broke a Nichicon 100µF 10v Fine Gold High-Grade Audio Capacitor off my motherboard's audio circuit and replaced it with a generic 100µF 50v capacitor I found at a local electronics store that best matched its dimensions. The new capacitor is rated for 115C vs 85C for the old one so I assume the quality is the same if not better.

However, upon testing my front audio connection I am noticing an extreme imbalance in my left and right audio channels. The left channel is nearly imperceptible and the difference is too large to overcome through balancing controls in my audio drivers.

Could this have been caused by not replacing my capacitor with an identical type? From some searching, I was given the impression that as long as I kept the capacitance constant I could increase the voltage rating when replacing a capacitor.

\$\endgroup\$
8
  • \$\begingroup\$ I'm not familiar with the specs of audio capacitors, but if the package size is similar at 1/5 the rated voltage, there's a reason. It may have low ESR or some other property that makes it work better in an audio circuit. If a regular capacitor was OK, the manufacturer would have used one. \$\endgroup\$
    – K H
    Feb 19, 2021 at 7:07
  • \$\begingroup\$ Where was this capacitor put? On the power line or in the audio path? It is possible that you are suffering from a mismatch in ESR, which is a big thing for power delivery. Also in the audio path usually they are not polarized and should be replaced with similar parts. \$\endgroup\$ Feb 19, 2021 at 7:07
  • \$\begingroup\$ The manufacturer actually advertises "Nichicon Fine Gold Series Audio Caps on the product page, so I'm split between thinking these caps were necessary or that they were included as a cheap way to build marketing cachet. I have no way of knowing whether the cap I replaced was on the power line or the audio path, but the original was polarized. \$\endgroup\$ Feb 19, 2021 at 7:17
  • \$\begingroup\$ "Could this have been caused by not replacing my capacitor with an identical type?" - No. Either it's faulty, or you broke a track on the motherboard, or something else is causing the massive drop in audio level. \$\endgroup\$ Feb 19, 2021 at 9:50
  • \$\begingroup\$ How do you test? Headphones or connection to amplifier or amplified speakers? It is highly unlikely that you or anyone else would notice any difference which capacitors there are, as long as they are both 100uF. Perhaps something else is broken, like ESD from soldering iron may have killed the analog output. The audio grade capacitors on PC motherboards is simply a marketing gimmick, any capacitors will do if the circuit is otherwise designed in a way that it does not have noise or distortion. \$\endgroup\$
    – Justme
    Feb 19, 2021 at 10:34

3 Answers 3

1
\$\begingroup\$

Such a high-value capacitor can find two usage areas in a MoBo's audio subcircuit:

  • Bypass capacitor for the audio codec chip's ADC or DAC supply.
  • Coupling capacitor for Line Out channels -- for better bass response.

Regarding your problem, most likely the 100uF capacitor that you have replaced was a coupling capacitor. Some old motherboard designs have 100uF coupling caps along with a low-value (e.g. 75R) series resistor, for both left and right channel line outputs.

The capacitor you used as a replacement has the same size, same capacitance but x5 rated voltage. This may tell us the original one has very low ESR and show very low impedance at audio frequencies compared to the other. So this may lead to some audible imbalance between the channels. So you should either replace the broken with the original one or replace the other original one (because there should be another 100uF) with the cheaper one.

The "audio-grade" thing is almost always a marketing hype, but I should say that some good quality electrolytic capacitors perform better (e.g. degrade less) in audio frequencies and even high-temperature environments.

\$\endgroup\$
3
  • 3
    \$\begingroup\$ ESR will not cause a gross imbalance in level on a line level audio signal. \$\endgroup\$
    – user16324
    Feb 19, 2021 at 16:10
  • 3
    \$\begingroup\$ If the new capacitor is somehow losing half of the signal (which is difficult to imagine) it would correspond to a 3dB drop in sound output, which you would hardly notice. That suggests that something else is afoot. \$\endgroup\$
    – Frog
    Feb 19, 2022 at 5:11
  • \$\begingroup\$ Thank you for your comment, I wanted to close the loop on this and mention that I fixed the problem by replacing the other channel's capacitor to keep things symmetrical. It was a strange phenomenon but thankfully one with a simply solution. \$\endgroup\$ Jun 25, 2023 at 14:56
0
\$\begingroup\$

Quality in passive electronic components are to do with

  • Initial accuracy in value
  • Drift with temperature
  • Drift with voltage/current
  • Drift with frequency
  • Drift over time.

And if it's rated 10 times the voltage, then it is unlikely to be as accurate.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ You've missed out the other key parameter of a quality component - the value and stability of the parasitic elements. In electrolytic capacitors the key parameter is the ESR (equivalent series resistance) \$\endgroup\$
    – D Duck
    Jun 24, 2023 at 10:25
0
\$\begingroup\$

In your case it most definitely makes no difference at all. Some claim, that the leakage will be higher, which very well might be the case, but it does not matter. It is far outside of any noticable margin.

If you broke a capacitor of the VRMs of the CPU or GPU, than the values would matter much more.

For audio applications at low power these are expected changes:

  • Slightly more or less noise (higher leakage = more noise)
  • Different Frequency rolloff on output capacitor (more capacity = lower bass)

For power decoupling it will not matter for a headphone amp, as all decoupling will be mostly done by smd ceramic/foil capacitors. So no worries in your case at all. It will work just as fine if you do not mix up the polarity.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.