i understand that many like very fine capacitors in their crossovers but how poor are electrolytic caps like m.d.l. sold by jantzen? my speakers are always short lived so if caps change their values slightly over time is nothing that bothers me, and series resistances of 0.2-0.3 ohm does not bother me either, what else is poor with electrolytics caps?
Nothing is poor with npe's , not until one knows these are employed. I am talking of solid brands you can find in a typical DIY shop, Intertechnik, Visaton, Bennic ,etc.
It has to do with how a electrolytic capacitor is constructed. Essentially part of conductive path the capacitor is essentially "salt water" or electrolyte. Imagine on end people are talking about 99.999% pure copper/silver cable and you have to pass it through salt water as a conductor.
Oon
Oon
Rod Elliott (ESP) says that a loudspeaker crossover is one place where bipolar electros should not be used, because distortion can be easily measured when there is a significant voltage across any electrolytic (bipolar or otherwise). Capacitor Characteristics
The impedance vs. frequency characteristics of bipolar electrolytic capacitors are linear over the audible frequency range and Rod says that resonance within the audio band cannot be claimed to be a problem.
He also says that it is a popular myth that electros have considerable inductance (claimed to be the case because of the way the foil is wound inside the can) and states that this is nonsense.
I offer the above simply as a stimulus for discussion. Don't shoot me, I am only the piano player!
The impedance vs. frequency characteristics of bipolar electrolytic capacitors are linear over the audible frequency range and Rod says that resonance within the audio band cannot be claimed to be a problem.
He also says that it is a popular myth that electros have considerable inductance (claimed to be the case because of the way the foil is wound inside the can) and states that this is nonsense.
I offer the above simply as a stimulus for discussion. Don't shoot me, I am only the piano player!
Last edited:
Yes
Sometimes with some drivers I even find NPE to sound better. We don't need always extra details at the price of some other defaults.
Same for MKT that sometimes gives better result than MKPs...
Many people looks at the number and the graphs and not always listening to in details...
I have nothing against good MKPs, they often are better, they just don't save a bad filter design (being not a designer myself but having refurbished enough filters to know)
Sometimes with some drivers I even find NPE to sound better. We don't need always extra details at the price of some other defaults.
Same for MKT that sometimes gives better result than MKPs...
Many people looks at the number and the graphs and not always listening to in details...
I have nothing against good MKPs, they often are better, they just don't save a bad filter design (being not a designer myself but having refurbished enough filters to know)
are you talking about normal electrolytic capacitors?
or crossover designed nonpolars?
quite a difference
Forgot : please guys consider the NPE from F&T : measure better in precision than NPE from Mundorf and Jantzen. They are often less than 5% while the Mundorf adverted 5% are often up to 12% ... at least all the ones I measured. And the F&T are nice both smooth and rough foil and cheaper ! Why asking more !
The main reason why i think they are worse than MKT or PP filmcaps is the instability and the higher ESR (so higher efficiency losses). But it's not that it's a disaster when you use good ones. I just won't do it for my own builds and would not advice it for project builds that need to last long in a good shape.
But when i repair a crossover with an elco, it's often an elco that replace it, or the owner has to pay me enough to redo the crossover totally from scratch (most won't do that). It's just easier (i don't have to compensate for the difference in esr), and elco's of today will always be better than the old ones from decades ago.
And good filmcaps don't have to cost a fortune, Wima MKP is also good, but i mostly use Jantzen Crosscaps because i can easely get them relative cheap.
But when i repair a crossover with an elco, it's often an elco that replace it, or the owner has to pay me enough to redo the crossover totally from scratch (most won't do that). It's just easier (i don't have to compensate for the difference in esr), and elco's of today will always be better than the old ones from decades ago.
And good filmcaps don't have to cost a fortune, Wima MKP is also good, but i mostly use Jantzen Crosscaps because i can easely get them relative cheap.
Passive crossovers have no place in a quality audio system; if you think about real-world drivers with impedances and outputs that you can't guess at and the amps and damping factors that drive them, just never made any sense. Active crossovers, even just plain old analog ones (like the ones I made in 1965) are way smarter.
So let's say you have a bunch of NPEs and some reliable way to "form" them and test them and you pick a few that come to match somewhere near your circuit requirement. If you don't test and make matched-pairs, are you happy with components off by very large tolerances? Put those tolerances into your sim and see what comes out Left and Right, bass and treble?
But doesn't the capacitance shift about in a salt-water and aluminum foil NPE, with the weather, forming voltage, age, wear, time since last use, loudness of the music, etc?
Aesthetically, NPEs are just the kind of junk stuff I don't like in my house (if I could live without heavy induction motors), and certainly not in a tricky part of my audio system... ever.
Last edited:
I've often wondered if people that comment about 'improvement' when swapping out the npe caps for film are only hearing the slight increase in signal due to the lower esr. They then assume they are hearing more detail due to the 'better' capacitor.
I think that is true in a ton of cases, although not all.
Besides, replacement typically happens in 20/30/40 years old cabinets, where electrolytics have certainly dried up, more or less
*Anything* will work better there ... even fresh electrolytics 😉
Not dissing "non electros", of course: if you can afford them, have enough space and want them, I´m not the one who will disuade you.
"Salt water" is a despective and very innnacurate name, electrolyte chemistry is high Science and in fact getting it wrong can lead to worldwide disaster, we had enough of those examples by now, and as accurate as complaining that cars run on rotten dinosaur juice extracts.
What matters is how that "salt water barrier" measures.
And actual construction details.
A 1 cm diameter 10 cm long glass tube will show TERRIBLE conductivity between its ends compared to a 1 cm diameter 10 cm long copper bar, that´s for sure, but when said barrier is 0.1 mm thick and a square foot wide (gap between electrolytic capacitor plates filled with electrolyte) then resistance may drop to milliohms .... the actual ESR we measure in that cap.
Then the actual problem boils down to the load resistance at the other end of said capacitor.
A 6.5 ohm DCR 8 ohm tweeter?
Measure capacitor ESR and calculate how many dB attenuation you get.
Tweeters often are attenuated by a series resistor anyway, so factor ESR into the needed value.
Besides, replacement typically happens in 20/30/40 years old cabinets, where electrolytics have certainly dried up, more or less
*Anything* will work better there ... even fresh electrolytics 😉
Not dissing "non electros", of course: if you can afford them, have enough space and want them, I´m not the one who will disuade you.
"Salt water" is a despective and very innnacurate name, electrolyte chemistry is high Science and in fact getting it wrong can lead to worldwide disaster, we had enough of those examples by now, and as accurate as complaining that cars run on rotten dinosaur juice extracts.
What matters is how that "salt water barrier" measures.
And actual construction details.
A 1 cm diameter 10 cm long glass tube will show TERRIBLE conductivity between its ends compared to a 1 cm diameter 10 cm long copper bar, that´s for sure, but when said barrier is 0.1 mm thick and a square foot wide (gap between electrolytic capacitor plates filled with electrolyte) then resistance may drop to milliohms .... the actual ESR we measure in that cap.
Then the actual problem boils down to the load resistance at the other end of said capacitor.
A 6.5 ohm DCR 8 ohm tweeter?
Measure capacitor ESR and calculate how many dB attenuation you get.
Tweeters often are attenuated by a series resistor anyway, so factor ESR into the needed value.
Rod Elliott (ESP) says that a loudspeaker crossover is one place where bipolar electros should not be used, because distortion can be easily measured when there is a significant voltage across any electrolytic (bipolar or otherwise). Capacitor Characteristics
The impedance vs. frequency characteristics of bipolar electrolytic capacitors are linear over the audible frequency range and Rod says that resonance within the audio band cannot be claimed to be a problem.
He also says that it is a popular myth that electros have considerable inductance (claimed to be the case because of the way the foil is wound inside the can) and states that this is nonsense.
I offer the above simply as a stimulus for discussion. Don't shoot me, I am only the piano player!
Good for Elliott and what a great stimulus you've given us. I'd say he speaks pure nonsense.
Here is measured distortion of Kef R3 at 105dB (10V) at 1m with Klippel NFS measuring system:
Here is writing from Kef R series whitepaper:
"Mention should be made regarding the two parallel capacitors C3 and C4. The total value is large and low-loss types would be prohibitively expensive, so the bulk of the value is made up of an electrolytic type and around 10% of the total value is a polypropylene type, wired in parallel - the so-called bypass configuration."
Here is the schematics of the crossover from which you can see that the bipolar elco is right in front of midrange:
If there is any distortion rise, i can not see it. Actually, the distortion is lowest in its bandpass. Kef uses inexpensive brand M.D.L. bipolar elco capacitors.
Last edited:
Kryptonite caps are best, they're so unique, so one cannot actually buy them, reallllly expensive stuff
After replacing the electrolytic on the tweeter in my 30 year old speakers I was amazed at how much I hadn't been hearing with them for some time. There are probably a lot of people out there listening to speakers that aren't what they used to be but that could be brought back for just a couple dollars in capacitors.