I took one channel of output from my portable CD player connected two different types of capacitors (caps in parallel). The other lead of each cap was tied to ground through its own resistor. Also, an amplifier was tied to each lead. Any difference in the signal would be amplified. I compared a 1uf electrolytic and 1uf film type.
The results were that the difference was so small I had to turn up the gain to hear a slight signal from the speaker. I didn't take an actual reading but it must have been ~100 db down. No way can any difference have been heard under normal conditions. The caps were either equally bad (coloring the signal the same) or equally good (transparent). That tiny signal that got through was probably the differences in tolerance of the cap's values and/or the resistors.
This of course was a small signal test. Larger voltages may have a different result. I would have to try one of those expensive brands to see if there is any more difference.
The results were that the difference was so small I had to turn up the gain to hear a slight signal from the speaker. I didn't take an actual reading but it must have been ~100 db down. No way can any difference have been heard under normal conditions. The caps were either equally bad (coloring the signal the same) or equally good (transparent). That tiny signal that got through was probably the differences in tolerance of the cap's values and/or the resistors.
This of course was a small signal test. Larger voltages may have a different result. I would have to try one of those expensive brands to see if there is any more difference.
I'm convinced that in certain types of circuits telling the difference between caps is nearly impossible. In other circuits, especially if the cap is used in a filter and at high voltages or currents, I'm not so sure. Almost all capacitor tests, including the ones I've done, are flawed. That said, I did some interesting tests here, FWIW.
CH
CH
It all depends on how the cap is being used, in particular how much signal voltage is across it. What load impedance were the caps feeding? If high, then all caps would sound the same as they would not be doing very much apart from blocking DC.
I doubt if you were resolving differences to -100dB - that would require some very fancy equipment and very careful setup.
I doubt if you were resolving differences to -100dB - that would require some very fancy equipment and very careful setup.
i never did any test like that, but it sounds fun 
anyways whenever possible i try to not use elco caps.
they change value over time, and tolerance figures are not so good anyways.
In xovers i never use elco caps, for the same reason.
probably one could find a -most probably inaudible- difference between film and elco caps, but i kindof never had any logical idea why is that a good thing.
anyways whenever possible i try to not use elco caps.
they change value over time, and tolerance figures are not so good anyways.
In xovers i never use elco caps, for the same reason.
probably one could find a -most probably inaudible- difference between film and elco caps, but i kindof never had any logical idea why is that a good thing.
1uF and 100K gives an LF corner at 1.6Hz. This means at 1kHz only 0.16% of the signal is across the cap. Given that distortion is likely to be level-dependent, this means that a cap which might generate 10% distortion with full signal (and few caps are that bad), would generate distortion at 0.016% of line level (-76dB). This could be near the noise level in your test amp.
I suspect you may be confusing dB as a measure of sound level (with respect to a particular reference sound pressure) and dB as a logarithmic expression of a signal voltage ratio.
I suspect you may be confusing dB as a measure of sound level (with respect to a particular reference sound pressure) and dB as a logarithmic expression of a signal voltage ratio.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.