Compound capacitor - my definition - multiple smaller capacitors in parallel to form a larger value capacitor.
Like many others, I've been playing around with this concept for power supplies and I like the results. The advantages are the smaller caps have higher self-resonance frequency and combining the caps in parallel reduces the ESR or DF of the combined cap in the same way parallel resistors have lower combined resistance. Since the high cost, big value power supply capacitors mainly seem to have low ESR, it seems like you can improve the performance of your power supply with parallel caps at a lower cost than the high priced caps.
On the downside, you have to be careful about impedance between the caps, especially careful about letting the leads between caps add to much inductance. Yet, it seems, if you are careful to keep lead length down, you can make a nice compound cap out of small caps.
I have also seen some folks trying compound caps with signal level and crossover caps, so I started to explore this possibility as well. I have forum friends that believe the best signal cap is a single, high quality cap and some of them have a great deal of experience. I am interested to hear opinions and experience with compound caps in crossovers, both positive and negative. In addition, what is the best way to combine caps to avoid adding inductance?
To help define the problem, I have a 22 uF cap in the signal path of my tweeter (DynaAudio D-28) as part of a Linkwitz-Riley 4th order acoustic crossover. At the moment, I have an Audyn Plus in that position that sounds pretty good. I would call the 'Plus' a medium price capacitor, but it is still $52 in a 22 uF.
A tempting approach for a compound cap would be to join 10 x 2.2 uF using Panasonic ECWFD caps. At less than $1 each, the price is certainly an improvement. Being smaller value caps, their self resonance is at about one million Hertz. The dissipation factor (DF) of the Audyn Plus and the Panasonic ECWFD 2.2 uF are about the same at 0.0005, but combining 10 Panny 2.2's should divide the dissipation factor by 10. From a spec standpoint, the Panny compound cap looks pretty good. Of course, good specs don't necessarily mean good sounding.
What do you think?
Like many others, I've been playing around with this concept for power supplies and I like the results. The advantages are the smaller caps have higher self-resonance frequency and combining the caps in parallel reduces the ESR or DF of the combined cap in the same way parallel resistors have lower combined resistance. Since the high cost, big value power supply capacitors mainly seem to have low ESR, it seems like you can improve the performance of your power supply with parallel caps at a lower cost than the high priced caps.
On the downside, you have to be careful about impedance between the caps, especially careful about letting the leads between caps add to much inductance. Yet, it seems, if you are careful to keep lead length down, you can make a nice compound cap out of small caps.
I have also seen some folks trying compound caps with signal level and crossover caps, so I started to explore this possibility as well. I have forum friends that believe the best signal cap is a single, high quality cap and some of them have a great deal of experience. I am interested to hear opinions and experience with compound caps in crossovers, both positive and negative. In addition, what is the best way to combine caps to avoid adding inductance?
To help define the problem, I have a 22 uF cap in the signal path of my tweeter (DynaAudio D-28) as part of a Linkwitz-Riley 4th order acoustic crossover. At the moment, I have an Audyn Plus in that position that sounds pretty good. I would call the 'Plus' a medium price capacitor, but it is still $52 in a 22 uF.
A tempting approach for a compound cap would be to join 10 x 2.2 uF using Panasonic ECWFD caps. At less than $1 each, the price is certainly an improvement. Being smaller value caps, their self resonance is at about one million Hertz. The dissipation factor (DF) of the Audyn Plus and the Panasonic ECWFD 2.2 uF are about the same at 0.0005, but combining 10 Panny 2.2's should divide the dissipation factor by 10. From a spec standpoint, the Panny compound cap looks pretty good. Of course, good specs don't necessarily mean good sounding.
What do you think?