I am about to build a speaker system, based on 2 Audiotechnology woofers + 1 Scanspeak Berylium tweeter.
The Amplifier is Hypex FA123.
In order to block DC, Hypex recommends a capacitor in series with the tweter.
As I see it, Hypex is not the only manufacturer, who recommends this, although, I see it as a disadvatage, removing one of the main arguments for an active system: no cross-over components.
My question, is, how to calculate the capacitor.
Some recomends that the cross-over frequence is in the range of half of the cross over, used in the Hypex module.
What do you think?
The lower the frequence, the more expencive capacitor!!
Any alternative ideas?
I did not yet buy the Hypex amplifier.
Thank you in advance for your help.
The Amplifier is Hypex FA123.
In order to block DC, Hypex recommends a capacitor in series with the tweter.
As I see it, Hypex is not the only manufacturer, who recommends this, although, I see it as a disadvatage, removing one of the main arguments for an active system: no cross-over components.
My question, is, how to calculate the capacitor.
Some recomends that the cross-over frequence is in the range of half of the cross over, used in the Hypex module.
What do you think?
The lower the frequence, the more expencive capacitor!!
Any alternative ideas?
I did not yet buy the Hypex amplifier.
Thank you in advance for your help.
Ordinarily I'd use the smallest capacitor possible without having to boost to compensate for it. In other words, instead of using a fourth order filter, use third plus this capacitor.
Hi Allen, and thank you for your reply.
With your sugested solution, I will need, to know my cross-over frequence in advance, limiting the flexibility, having an active system.
Or, do I misunderstand you?
With your sugested solution, I will need, to know my cross-over frequence in advance, limiting the flexibility, having an active system.
Or, do I misunderstand you?
A first order filter, 3 db/octave, requires a lot fewer $$ crossover components.
Knowing the crossover point comes from the brochure that comes with the drivers. Plus what design you are copying. I'm copying a Peavey SP2g 2004.
Say you want to crossover at 2 khz with a 8 ohm tweeter. First order is one capacitor series the tweeter. This means a 8 ohm capacitor at 1 khz.
1/(2*pi*f*C)=Z so C=1/(2*pi*f*Z) C=1/(6.28*1000*8)=19.9E-6 or 19.9 uf
The capacitor would be 4 ohms @ 2000 hz. A 3 db octave woofer crossed at 1000 hz would add some sound @ 2000 hz.
The 250 v 20 uf capacitors @ parts-express are $8 for generic & $14 for precision.
You can buy a number of 10 uf or 4.7 uf caps and a rotary switch and add or subtract capacitance to taste. That is how my KLH23 speaker came. 3 crossover positions. No crossover (inductor) on the 10" woofer at all. The tweeter was 3" paper cone.
Additional design consideration is the tweeter power limit. I've blown 2 tweeters with a 35 w/ch amp (dynaco ST70). My current amp will put out 25 vac max. The RX22 tweeter is rated 60 W. I'm putting an 1157 brake light bulb series the tweeter, 1 ohm cold, followed by back to back 20 v zeners across the tweeter to clamp the voltage to 21.
Voila! no need to power up two power amps the 12 hours a day I listen to music. Passive 3db/octave crossover does the job for $26 a channel. Including .82 mh inductor series the woofer. We'll see how it sounds in test.
Knowing the crossover point comes from the brochure that comes with the drivers. Plus what design you are copying. I'm copying a Peavey SP2g 2004.
Say you want to crossover at 2 khz with a 8 ohm tweeter. First order is one capacitor series the tweeter. This means a 8 ohm capacitor at 1 khz.
1/(2*pi*f*C)=Z so C=1/(2*pi*f*Z) C=1/(6.28*1000*8)=19.9E-6 or 19.9 uf
The capacitor would be 4 ohms @ 2000 hz. A 3 db octave woofer crossed at 1000 hz would add some sound @ 2000 hz.
The 250 v 20 uf capacitors @ parts-express are $8 for generic & $14 for precision.
You can buy a number of 10 uf or 4.7 uf caps and a rotary switch and add or subtract capacitance to taste. That is how my KLH23 speaker came. 3 crossover positions. No crossover (inductor) on the 10" woofer at all. The tweeter was 3" paper cone.
Additional design consideration is the tweeter power limit. I've blown 2 tweeters with a 35 w/ch amp (dynaco ST70). My current amp will put out 25 vac max. The RX22 tweeter is rated 60 W. I'm putting an 1157 brake light bulb series the tweeter, 1 ohm cold, followed by back to back 20 v zeners across the tweeter to clamp the voltage to 21.
Voila! no need to power up two power amps the 12 hours a day I listen to music. Passive 3db/octave crossover does the job for $26 a channel. Including .82 mh inductor series the woofer. We'll see how it sounds in test.
Last edited:
Michaelbreer, the preparation work for a passive or active system should be largely the same. There shouldn't be such a limit to flexibility in either case. Of course I understand the methods tend to lend themselves to various opportunities.
So I would guess the limit is either with the tools you are using, or with your present expectations.
Perhaps you could explain how/where you see the problem happening so we could make a suggestion?
So I would guess the limit is either with the tools you are using, or with your present expectations.
Perhaps you could explain how/where you see the problem happening so we could make a suggestion?
The tweeter already displays a high pass function in itself. You have to take that into account anyway. While Allen’s solution is the most elegant, you could consider using a cap that changes the natural second order rollof to a decent (B3) third order one. That might need adding some tweaking (notch) in the active filter setup though. You’d still have to compensate for it when setting the desired X-over, but that would be easier.Michaelbreer said: