[FONT=Helvetica, Arial, sans-serif]I just have one question because either I am misunderstanding electronics (which is quite possible, I'm a molecular biologist, not an engineer) or something is confusingly written in the DEQX manual—particularly the sentence in bold. Everything else is very clear to me.
[/FONT][FONT=Helvetica, Arial, sans-serif]
The manual says:
“As indicated in Figure 5 and Figure 6, a protection capacitor (a high voltage film
capacitor) in series with the tweeter is generally recommended. This will help protect it from DC offsets and turn on and turn off transients. If there is any question that the amplifier and upstream equipment is not perfectly behaved or that the tweeter is fragile, the capacitor is strongly recommended.
Unlike other systems, the DEQX will compensate for the roll off and phase shift introduced by the capacitor. The capacitor therefore does not have to be large (a common rule of thumb is to size it so that its roll off starts two octaves below the crossover frequency), but sized to roll off at the anticipated crossover frequency. Since it is relatively small in size, a very high quality capacitor can be used.”
So—the passive crossover in my speakers is at 2250 kHz – what capacitor do I get? It would be great if you could be very specific—like a link to a part.
[/FONT]
Last edited:
Hi, since you quoted from the DCX page, I guess you're doing active/multi-amp system, so the suggestion of putting a capacitor in line after the amp and before the speaker ( tweeter ).
That's because tweeters are not as robust as woofers ( the moving mass has to be very low because it has to vibrate from -say- 2 kHz to 20 kHz, and the wire of the coil is counted inside the Mms value, tw's wires are very thin...)
and they might burn in a fraction of a second if DC arrives at coil's terminals.
So a DC breaker such a capacitor in series is recommended. It has not to interfere with original filter's slope so it is better if its Fc ( cut ) is centered an octave below or two.
For normal passive crossover technique, a single capacitor ( which performs the high pass function according to C=1/(2*pi*f*R) ) is such a poor performer ( well, with tweeters having ferrofluid inside the gap that dampens the peak at Fs, it might perform good ) and going higher with power it will not guarantee any protection - a single cap ( or coil for the lowpass section ) generates a filter with 6 dB/octave slope and 6 dB is just *half the volume*
That's because tweeters are not as robust as woofers ( the moving mass has to be very low because it has to vibrate from -say- 2 kHz to 20 kHz, and the wire of the coil is counted inside the Mms value, tw's wires are very thin...)
and they might burn in a fraction of a second if DC arrives at coil's terminals.
So a DC breaker such a capacitor in series is recommended. It has not to interfere with original filter's slope so it is better if its Fc ( cut ) is centered an octave below or two.
For normal passive crossover technique, a single capacitor ( which performs the high pass function according to C=1/(2*pi*f*R) ) is such a poor performer ( well, with tweeters having ferrofluid inside the gap that dampens the peak at Fs, it might perform good ) and going higher with power it will not guarantee any protection - a single cap ( or coil for the lowpass section ) generates a filter with 6 dB/octave slope and 6 dB is just *half the volume*
Look at the impedance plot of your tweeter as published by the mfg to find the impedance at 2250Hz. Plug that into Anderson's formula, or use this basic chart to find the -3dB rolloff point for a single series cap.
1st Order / 6 db Butterworth Crossover Table
Andersonix' recommendation of 36uF is correct for 8 ohm tweeter with F3 at 560hz, two octaves below. The DEQX manual says you can use F3 at the actual crossover frequency of 2250, which would be estimated 10uF for 8 ohm tweeter, or 22uF for 4 ohm tweeter. Your actual impedance at 2250 may be lower than the nominal tweeter impedance, so check the mfg impedance graph if you want to be fussy. But as long as it's in the ballpark DEQX will correct.
Adding a cap to the tweeter introduces FR rolloff, and phase shift, so a protection cap is usually sized very large so that it's EQ and phase effects happen at very low freq (2 oct below,) but are insignificant at the design crossover freq where the cap acts only as a DC blocker.
DEQX allows you to use a smaller cap with F3 at the design crossover freq. DEQX will first characterize the tweeter separately to learn the net FR and phase of tweeter and cap together before it creates a crossover filter which fully compensates for the cap's presence. So you can use a smaller cap with F3 at the intended crossover freq to save money, or spend the savings on a better quality cap.
It's still OK to use a larger cap that is "out of the way" at the crossover freq, but it is not OK to use a cap that is too small (whose F3 is higher than your intended design crossover freq,) because then DEQX will compensate for the additional rolloff by sending extra power to achieve flat tweeter FR and that will waste amp headroom.
1st Order / 6 db Butterworth Crossover Table
Andersonix' recommendation of 36uF is correct for 8 ohm tweeter with F3 at 560hz, two octaves below. The DEQX manual says you can use F3 at the actual crossover frequency of 2250, which would be estimated 10uF for 8 ohm tweeter, or 22uF for 4 ohm tweeter. Your actual impedance at 2250 may be lower than the nominal tweeter impedance, so check the mfg impedance graph if you want to be fussy. But as long as it's in the ballpark DEQX will correct.
Adding a cap to the tweeter introduces FR rolloff, and phase shift, so a protection cap is usually sized very large so that it's EQ and phase effects happen at very low freq (2 oct below,) but are insignificant at the design crossover freq where the cap acts only as a DC blocker.
DEQX allows you to use a smaller cap with F3 at the design crossover freq. DEQX will first characterize the tweeter separately to learn the net FR and phase of tweeter and cap together before it creates a crossover filter which fully compensates for the cap's presence. So you can use a smaller cap with F3 at the intended crossover freq to save money, or spend the savings on a better quality cap.
It's still OK to use a larger cap that is "out of the way" at the crossover freq, but it is not OK to use a cap that is too small (whose F3 is higher than your intended design crossover freq,) because then DEQX will compensate for the additional rolloff by sending extra power to achieve flat tweeter FR and that will waste amp headroom.
Last edited:
Hi,
6 dB are four times the power, hence a first-order highpass is sufficient spectral thermical protection. That is, if it is fit for X and really works. But as a passive circuit consisting of only a capacitor fit for X, it usually pronounces lower ringtone of tweeter hence is less suited as spectral thermic protection, i think, too. So i would fit it in fact for X/2, not for X.
For absolute thermical protection i recommend to insert a slow-acting wire fuse, which is gonna break at thermical power handling, which is around 5W for a 25mm diameter 3mm height voice coil on aluminium former. Finally a poly-fuse would also do, but for the developer it is better to know for sure, when power handling has been reached, and so more thoroly evaluate the characteristics of the loudspeaker.
Uli
For normal passive crossover technique, a single capacitor ( which performs the high pass function according to C=1/(2*pi*f*R) ) is such a poor performer ( well, with tweeters having ferrofluid inside the gap that dampens the peak at Fs, it might perform good ) and going higher with power it will not guarantee any protection - a single cap ( or coil for the lowpass section ) generates a filter with 6 dB/octave slope and 6 dB is just *half the volume*
6 dB are four times the power, hence a first-order highpass is sufficient spectral thermical protection. That is, if it is fit for X and really works. But as a passive circuit consisting of only a capacitor fit for X, it usually pronounces lower ringtone of tweeter hence is less suited as spectral thermic protection, i think, too. So i would fit it in fact for X/2, not for X.
For absolute thermical protection i recommend to insert a slow-acting wire fuse, which is gonna break at thermical power handling, which is around 5W for a 25mm diameter 3mm height voice coil on aluminium former. Finally a poly-fuse would also do, but for the developer it is better to know for sure, when power handling has been reached, and so more thoroly evaluate the characteristics of the loudspeaker.
Uli
Last edited:
You must be paraphrasing something in the manual, because I can't find that language. Then again I don't really know what I'm talking about
So what value of capacitor should I use?
You actually predicted and answered my next question about phase, thanks!
I would use the smallest capacitor value I reasonably could, which means putting its rolloff right at the crossover. While this offers better protection, smaller capacitors are normally less expensive and better performing. Additionally this makes the capacitor a part of the crossover rather than a redundant part.
You must be paraphrasing something in the manual, because I can't find that language. Then again I don't really know what I'm talking about
Check your original post.
Aww that's no fun!
1. Look up the impedance at 2250Hz on your tweeter's datasheet. Make sure you lookup the right tweeter, as some are available in both 4 ohm and 8 ohm versions.
2. Use the formula provided above by Andersonix, using the impedance at 2250Hz for R.
3. Round up your "C" solution to the next larger standard cap value.
It should be a film cap with >100V voltage rating, or at least higher than your tweeter amplifier's voltage rails.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.