The shame is all those brands have many caps, sometimes more than 8 !
Worse than Levis' with jeans !
Most of time two will be enough or justg one : the best they can do supposing they master how to do. Looking at the different subjective qualities of all those caps in a same circuit, I surmise the process not to be mastered but on the electrical side of course, an engineer stays an engineer !
This is off course a business and if it works it means there are customers but don't get me wrong : some are better than others, but the rank is not universal : I will not put a clear silver mundorf e.g. with an aluminium cone tweeter !
Worse than Levis' with jeans !Most of time two will be enough or justg one : the best they can do supposing they master how to do. Looking at the different subjective qualities of all those caps in a same circuit, I surmise the process not to be mastered but on the electrical side of course, an engineer stays an engineer !
This is off course a business and if it works it means there are customers but don't get me wrong : some are better than others, but the rank is not universal : I will not put a clear silver mundorf e.g. with an aluminium cone tweeter !
These discussions never go anywhere. I suppose an interesting experiment would be to put a high quality switch between caps and see if you can actually tell the difference. Or build a sort of Wheatstone Bridge with different caps in the arms and connect some headphones across the shunt. You would actually hear the difference if you got it in balance.
I tend to buy 250V or 400V MKP caps these days since the biggie 1000V types are so enormous, space consuming and expensive. And you can certainly design a crossover to present an easy impedance to an amplifier which may be more important.
But I was always struck by the rather brilliant observation of one of my lecturers (Prof. Tony Constantinides of Imperial College, who was just THE BEST on digital and analog filters...), that filter component tolerances only really affect the -3dB crossover point. Elsewhere filters are bandstop or bandpass, simple as that.
I tend to buy 250V or 400V MKP caps these days since the biggie 1000V types are so enormous, space consuming and expensive. And you can certainly design a crossover to present an easy impedance to an amplifier which may be more important.
But I was always struck by the rather brilliant observation of one of my lecturers (Prof. Tony Constantinides of Imperial College, who was just THE BEST on digital and analog filters...), that filter component tolerances only really affect the -3dB crossover point. Elsewhere filters are bandstop or bandpass, simple as that.
Oh dear. When someone posts a rather obvious joke and someone takes it seriously you know it's time to quit lol.
And I think it is rather safe to say here that any non linearities introduces by capacitors are far, far, smaller than those introduced by the loudspeaker elements themselves.
Now this isn't to say that different caps and types of cap don't have their place, they do. But once you've got a good quality, well formed polyprop cap of a suitable voltage in your xover, there really isn't any reason to go about changing it for something else.
As Pallas quoted. Boutique capacitors are just about the biggest waste of money you can get when it comes to loudspeakers.
And I think it is rather safe to say here that any non linearities introduces by capacitors are far, far, smaller than those introduced by the loudspeaker elements themselves.
Now this isn't to say that different caps and types of cap don't have their place, they do. But once you've got a good quality, well formed polyprop cap of a suitable voltage in your xover, there really isn't any reason to go about changing it for something else.
As Pallas quoted. Boutique capacitors are just about the biggest waste of money you can get when it comes to loudspeakers.
But I was always struck by the rather brilliant observation of one of my lecturers that filter component tolerances only really affect the -3dB crossover point. Elsewhere filters are bandstop or bandpass, simple as that.
This.
Once below the corner the reactive components are acting like a wire. Above it and the frequency response is then so far attenuated that you can't hear anything.
I find all this talk about audiophile boutique capacitors a laugh really. The BEST capacitor in the world is no capacitor. The way some people wax poetic about their $100 cap would imply that the difference between it and another sensibly priced good quality cap, is so amazing that you'd be hard pressed not to notice.
Well then why when you remove said capacitor does the sky not fall down? And by this I am referring to active systems where one only has a cap on the tweeter to protect it from DC. You are free to remove it of course, but at the risk of blowing a tweeter. Yet when you do remove it, is it an audio revelation? No. If someone removed the cap in your system and no one told you, from one day to the next of listening you'd think that nothing had changed. So much for the night and day difference.
Generally true, but... watch out that ESR doesn't bite you. I've recounted a few times how I made a speaker sound much worse by changing a nonpolar electrolytic to a good quality PP cap- the speaker designer, who was apparently much smarter than I, used the ESR as part of the transfer function. Lesson learned.
I think I understood the joke in German, Joachim. To do with the sky falling down. Which proves, just as I have always said, that the Germans have a rich sense of humour. Along with a tendency to invade Poland.
Of far more interest to me, is that solid state amplifiers are incredibly sensitive to speaker impedance between 5-20kHz. This is apparently because output transistors run out of open-loop gain at high frequencies, so there is less effective feedback in the closed loop.
Attenuators and impedance correcting tweeter Zobels have a dramatic effect on the high frequency energy and detail. I haven't fully formed a theory, but around 15R resistive at 10kHz seems to sound good.
Of far more interest to me, is that solid state amplifiers are incredibly sensitive to speaker impedance between 5-20kHz. This is apparently because output transistors run out of open-loop gain at high frequencies, so there is less effective feedback in the closed loop.
Attenuators and impedance correcting tweeter Zobels have a dramatic effect on the high frequency energy and detail. I haven't fully formed a theory, but around 15R resistive at 10kHz seems to sound good.
Last edited:
The better capacitor is, the more it sounds like...no capacitor (no cap always sounding the best to me)
I found Tony's Capacitor page pretty accurate and good as a guide for those who need it.
Most of the time i end up with mundorf supreme or jantzen silver-z though and don't bother with anything else.
As for ratings - for example ClarityCap ESA 250 and 630 are a lot more alike than if either is compared to some other brand...
I found Tony's Capacitor page pretty accurate and good as a guide for those who need it.
Most of the time i end up with mundorf supreme or jantzen silver-z though and don't bother with anything else.
As for ratings - for example ClarityCap ESA 250 and 630 are a lot more alike than if either is compared to some other brand...
Last edited:
Pretty much what I do in my own stuff- but I'm not trying to get the most bang for the lowest BOM cost. And for things like woofer Zobels, bipolar electrolytics work fine if you account for the losses in the design.
- Status
- Not open for further replies.