Ouch, great example you've found.
I am looking for components at Germany site, prices before VAT
Of course. An Aliexpress piezo speaker is cheaper and smaller compared to a HF driver of reasonable quality too. Nobody will argue. But what's your point?
It actually also really depends if you would like to go for rather (very) expensive consumer capacitors, or just get a bunch of "industrial" types from Digikey, Mouser, TME, Farnell, ....
Especially the lower voltage ones are VERY affordable.
Bipolars do have a tendency to have a higher ESR, but one way to compensate, is just buying a bunch and putting them in parallel.
Probably will be smaller and maybe ever cheaper.
For the amount of money some people spend on passives filters, you can actually build a very well performing active speaker, and probably have quite some money left to buy a decent wine to enjoy the music with.
If you think about it, whatever anomalous sound the cap in that position makes, would show up pretty far down in level, compared to the bass it allows to pass. If the woofer happens to have a lot of sensitivity at those higher frequencies - like it has a whizzer for some reason - maybe you could hear it. So there might be some dependency on the type of driver used for the woofer.
Another long shot is the character of that cap makes it back through the inductor, to perturb the output impedance of the amplifier, then its sound is heard through the midrange / treble. Remember the ear is logarithmic and can hear any dern thing even at remarkably small levels, which is what makes audio reproduction so much fun!
This view came from recapping a KEF crossover, which had all kinds of RLC networks just to make the impedance of the speaker seem flat from the amplifier's perspective - one of KEF's tricks. Did the type of cap in those networks - NP
electrolytic's, KEF's favorite - effect the sound too? Some say yes - which of couse depends on the type of amplifier and the magnitude of its output impedance.
I think KEF banks on some theory that after an NP electrolytic is "worked" by the music signal for a few minutes, you cant hear any difference. So you play the speaker for a few minutes then it starts sounding best - so what - who's going to know? Who's going to come back from vacation, startup their system and hear something so wrong in the first minute - and complain? Against the cost of a large handful of gigantic film caps?
OP's solution is probably optimal, short of going to a bank of film caps.
Another long shot is the character of that cap makes it back through the inductor, to perturb the output impedance of the amplifier, then its sound is heard through the midrange / treble. Remember the ear is logarithmic and can hear any dern thing even at remarkably small levels, which is what makes audio reproduction so much fun!
This view came from recapping a KEF crossover, which had all kinds of RLC networks just to make the impedance of the speaker seem flat from the amplifier's perspective - one of KEF's tricks. Did the type of cap in those networks - NP
electrolytic's, KEF's favorite - effect the sound too? Some say yes - which of couse depends on the type of amplifier and the magnitude of its output impedance.
I think KEF banks on some theory that after an NP electrolytic is "worked" by the music signal for a few minutes, you cant hear any difference. So you play the speaker for a few minutes then it starts sounding best - so what - who's going to know? Who's going to come back from vacation, startup their system and hear something so wrong in the first minute - and complain? Against the cost of a large handful of gigantic film caps?
OP's solution is probably optimal, short of going to a bank of film caps.
Last edited:
Well, back in the day, almost everyone made the impedance flat.
This was for a couple of reasons.
Probably just a little left over from the older tube amplifiers, since the output impedance of those wasn't so great.
But most of all, actually measuring and mostly simulating speaker filters/crossovers wasn't as easy as today.
So the easiest way of calculating a filter, is when you don't have to deal with some kind of complex impedance, but make the impedance flat (aka resistive).
In theory the filter curve would also be slightly steeper compared to a rising impedance. In practice that differences are rather small and can be easily compensated to a level that you see the difference somewhere around -24dB levels (so not really significant anymore)
Smart people, please answer questions I asked here in post
https://www.diyaudio.com/forums/multi-way/340233-bipolar-vs-mkp-2.html#post6242091
Thanks a lot.
https://www.diyaudio.com/forums/multi-way/340233-bipolar-vs-mkp-2.html#post6242091
Thanks a lot.
First of all, a resistor in series with the cap for a 2nd order electrical LP filter is a trick that will slightly alter the slope and can thus be used to reach the acoustic slope you aiming for, whatever it is. Saying you need a value of 12% of the nominal impedance of the driver is pure nonsense as you don't strictly need a resistor any time. You need a simulation software and measurements. You choose Watt rating depending on the application (i.e. how much power you expect will flow into that resistor).
Increasing the cap will bring up the slope, lowering it will bring down the slope. An aged electrolytic cap will show this effect, and also a higher ESR.
Ralf
Thanks for the reply!
Bringing up the slope means that higher frequencies that were suppressed before will be played mid-bass more, right?
This means that my small voices issue potentially can be solved if I will do right replacement! Great stuff if so!
The resistor in the parallel branch also changes the phase curve (obviously). This affects the (acoustic) summing of the channels (here bass and mid), often more than a dB or so. In effect you can use this resistor to optimize response, e.g. when faced with acoustic alignment issues. But that of course requires measurements and (for the sake of economics) a simulation app.
Hi,
In the impedance corection circuit, esr of lytics should be measured, especially if old in order to adapt the Z of the coil and resistor...
lytics ; plain & Smooth is the same terminology, just can differ on some datasheets according the brand. Difficult to find a smooth NPE above 100 uF.
JB capacitors are also not so bad.... The Jantzen smooth I suspect the same than Mundorf!
The MKT 73-16 from Russia are the best ever MKT for audio, maybe the ferous foil gives some vibration damping, but they worth a try for the price in the midrange.
In the impedance corection circuit, esr of lytics should be measured, especially if old in order to adapt the Z of the coil and resistor...
lytics ; plain & Smooth is the same terminology, just can differ on some datasheets according the brand. Difficult to find a smooth NPE above 100 uF.
JB capacitors are also not so bad.... The Jantzen smooth I suspect the same than Mundorf!
The MKT 73-16 from Russia are the best ever MKT for audio, maybe the ferous foil gives some vibration damping, but they worth a try for the price in the midrange.
If you use K73-16 in the LF and midrange sections, the capacitor is good. But I wouldn't use it in series with an HF speaker. The sound is too harsh.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.