"Wasting all your money on exotic caps in a low pass circuit usually makes no sense unless there is a specific design need for it"
Have you ever measured the losses through a typical electrolytic as compared to a group of ganged "exotic" caps? These losses become self-evident when the speaker drivers and cabinet alignment are up to the task. Problem is, most speaker systems lack proper transient response at low frequencies, hence the difference between caps getting lost in the "mud" & "smearing" that emanates from such a design.
If one has some parts laying about with ( near ) identical values in an "exotic" and electrolytic cap, and you have a variable voltage power supply with a meter, try the following.
Place the electrolytic cap on the power supply with the supply turned all the way down. Gradually bring the power supply up in steps, paying attention to the voltage as it tracks upward. So long as the cap is in decent shape and not ancient, bring the voltage up to or just below the rated voltage. Double check the voltage reading on the meter and let it sit for half an hour or so. Then come back and check your voltage reading again. In most cases, the voltage will have climbed noticeably higher than where you left it. Remove the cap, but don't discharge it.
Use the same approach as above, but replace the electrolytic with the "exotic" cap. Step the voltage up and watch as the cap tracks the voltage climb. Bring the voltage up to the same point where you originally stopped with the electrolytic i.e. just at or slighlty below the rated voltage on the electrolytic. Now let the cap sit for half an hour on the power supply.
When you come back, you'll find that the voltage is still very close to where you originally set it. Unlike the electrolytic, there should be very little "voltage creep" after the initial setting. That's because most "exotic" caps have FAR better transient response and / or "loading characteristics". This translates sonically into sharper transients with less time induced smearing. Remove the cap, without discharging it, and let it sit next to the electrolytic. Obviously, you don't want these laying about where someone can get "whacked" by touching the charged caps.
After several days, come back to the caps and measure their voltages. What you'll find is that the "exotic" cap has retained very close to the full charge whereas the electrolytic has lost quite a bit of voltage. This has to do with internal losses in the electrolytic. Just as the electrolytic has a higher voltage loss, it also has a higher signal loss too.
The end result is that, using "exotic" caps anywhere / everywhere in a crossover circuit, and they really don't have to be all that "exotic" in terms of price, is lower loss with improved transient response. As previously mentioned, the rest of the circuit / speaker design has to be up to the task of revealing the differences / benefits in such a change. Otherwise, one really is wasting their money using these "higher grade" parts.
A reasonable compromise here is to use an electrolytic and combine it with a group of ganged "exotic" caps in order to achieve the total capacitance value desired. This helps keep costs down quite a bit as compared to using all "exotic" caps without incurring the massive sonic losses associated with using just a single large electrolytic. While many will instantly think of "bypass caps" as found in a power supply, i'm talking about a MUCH higher ratio of "exotics" to electrolytics in terms of absolute values here.
Since music is of a transient nature, the "exotic" caps in such a set-up provide the speed, clarity and focus that most electrolytics lack. Dynamics become far punchier and tighter.
On the other hand, the electrolytic provides the core foundation for the storage capacity needed. We need this in order to keep the circuit working as electrically intended, both in terms of volume and in terms of actual crossover frequency, while keeping costs down.
This approach allows the sluggish response of the electrolytic, which is now smaller in value than the original cap used, and therefore less lossy, to be helped along by the faster response times and reduced losses of the "exotics". We now have the electrolytic carrying the brunt of the load at reduced costs without having to respond quite as quickly. That's because the "exotics" take over that part of the equation. The cost of the "exotics" has been reduced too, as we are only using them for a portion of the total capacitance used.
Depending on the caps used and what one wants to spend, one can play with the ratios here. Obviously, the faster and less lossy the cap used, the better the overall performance will be. As such, the more that we can reduce the overall volume of capacitance in the electrolytic, and shift more of it onto the "exotic" caps, the better the potential sound.
With that in mind, one big problem can arise from all of this. That is, "exotic" caps tend to be quite large in size for the given volume of capacitance involved. As such, one can eat up a good amount of internal volume within the cabinet if using gobs of "exotics". This can end up detuning the bass alignment somewhat, so one may have to play with adding additional damping material in the cabinet in order to achieve similar box volumes. That is, ADDING damping material makes the box seem bigger internally, not the other way around.
One more thing. Many "exotic" caps tend to have very long legs on them. These legs tend to ring or resonate quite badly. If one doubts this, loosely hold the body of the cap and "flick" the extended lead of the cap with your finger. Chances are, you'll not only feel the cap body shake quite a bit, but a very metalic ringing will be heard quite loudly. As such, you always want to keep the leads as short as possible AND damp them as they go into the body of the cap.
Since the leads are shorter going into the cap, there will be more heat transfer as connections are made using a soldering iron, so i recommend using some type of "clip on" heatsink device between the connecting point and the body of the cap. Something as simple as an alligator clip can work quite well.
One more thing. Before replacing older caps with newer "exotic caps", gradually bring the new caps up to rated voltage with a variable voltage power supply. Let them sit for a day or two and then check their voltages. If a cap has dropped down in voltage quite a bit, it needs further forming. Apply more voltage and re-check in a few days. If it is still losing voltage at a high rate, that cap may not be worth installing as an "upgrade". You may be able to return it to the place of purchase, so you don't want to cut the lead length down until actual installation takes place.
One more thing for sake of safety. After you're all done checking the caps and letting them sit for a few days, ALWAYS discharge them before handling them. A cap can hold quite a bit of energy i.e. enough to scare the hell out of you and / or possibly stop your heart. If you're unskilled in this area, read, learn and find someone nearby with experience to help you.
Using this appoach, the electrical "break-in" period of the new caps forming is drastically reduced. Since the caps were brought up to their rated voltage, or very near it, this is FAR more voltage than they would ever see in an actual loudspeaker circuit. The fact that this voltage remained consisted for a few days and was not just a temporary dynamic swing, like that of a music signal, further stresses and helps form the cap. This reduces both loss in the dielectric and helps it to fully "form".
On top of that, the "metalic ring" that many complain about with some "exotic caps" is reduced / negated. This is due to the damping of the leads at the entrance into the body of the cap. As far as damping material goes, the use of something along the lines of "Blu-tak" or the generic equivalent seems to work well and is easily molded.
Hope this helps and opens some ears / minds. Sean
>
