TOPIC:

In the documentation of the minidsp flex eight (as well as your other active crossovers), you indicate the need to use a capacitor to protect the tweeter. Is there a specific rule/equation to determine which capacitor to choose based on tweeter spec, amp wattage, and/or crossover frequency? And does this also apply with modern amplifiers, class D based for example on TI TPA3250 OR 3255 chips, which a priori have specific protection?

(I already know the use of capacitors for passive crossover design... But the idea here being to protect only the tweeter while limiting the influence of the capacitor on the useful frequency of this one.)

A capacitor will protect against steady-state DC.  Low frequency thumps and transients could still deflect a tweeter driver because of charging currents.
Most amplifiers now are pretty well behaved and won't cause any tweeter problems.
If you're not feeling confident, I suggest to use a value for a series capacitor so it forms part of your desired electrical crossover slope.

That said, I've been hooking amplifiers directly to various tweeters in various speaker systems for nearly forty years now and I've never damaged one.
I do know what I'm doing though.  Many audiophiles don't. 
YMMV.

Dave.

Thank you Dave,
I too have never used a capacitor on the tweeter in my previous active filters (nanodigi). And indeed by being careful and lucky it was never a problem.
I just wanted to know if there was a rule not on the electrical capacitance (uF value) but on the choice of capacitor voltage in this specific case according to the characteristics of the tweeter, the power of the amps, etc.

You'd use a non-polarized capacitor with a higher voltage rating than would ever be seen during normal usage.
Other than that, there are no 'rules' to speak of.

Audiophiles have a lot of opinions regarding speaker capacitors, but most of them are based on subjective evaluation and/or speculation. 

Dave.

So there is no big difference between using a capacitor as a protection or as a high pass in a passive filter. It makes sense. Except that it is desired here to limit its influence on the frequency response, therefore an electric capacitance chosen for this purpose without either exaggerating its value. Thanks again Dave.

I meant to say 'no' rules above.  (Edited my post.)

If you desire to limit its influence on the frequency response, then a larger value becomes necessary........thus, you are exaggerating it's value.
You're being diffuse in your queries here so I'm not sure I'm following.  But, I think you've got the gist of it.

Dave.

dreite wrote:

If you desire to limit its influence on the frequency response, then a larger value becomes necessary........thus, you are exaggerating it's value.

 

Yes that's what I meant but without exaggerating too much.
For example on an 8ohm tweeter, a 50uF capacitor gives -7dB at 200Hz and -26dB at 20Hz while giving -0.3dB at 2kHz (which is fine here) but does this protect the tweeter enough?
An 18uF gives -15dB at 200Hz, -35dB at 20Hz and -1.2dB at 2kHz (which is not so dramatic when performing an active high pass around 2.5kHz) which still seems safer.
Then 18uF remains a fairly high value for a tweeter. In passive filtering we would rather be around 4.6uF for -6dB at 2.5kHz (obviously simplifying, passive filtering of a tweeter requires in most cases more work and component than that).

On the other hand, it is also necessary to monitor the influence of the capacitor on the impedance and the phase. But if we take care to make the initial measurement with the capacitor in series, in active filtering it does not change much.

In short, it is not complicated but not simple either.

Actually, it is simple.  But, it's up to you.

Cheers,

Dave.