Could you guys tell me how should be installed a capacitor to protect a tweeter in an active speaker? Should I just solder the capacitor in the middle of the cable that connects the tweeter to the speaker terminals like the one in the picture? Also, should it be on the + or - side?
An externally hosted image should be here but it was not working when we last tested it.
Possibly preventing the capacitor and diver resonance from corrupting the crossover slope...? 
Seriously, there are two ways of handling the effects of a passive filter (which is what this is) in an active design. Either you calculate/simulate the response and consider it when designing the active filter, or you just make the passive crossover frequency so low that you don't have to.
The latter may be perfectly ok if the purpose is to block DC from entering the tweeter, on the other hand the capacitor has to be bigger (=more expensive, possibly an electrolyte). Also, I like the thought of considering the summed effect of EVERYTHING, including the drivers, when designing loudspeakers. After all, it is the sum of everything we listen to.
Edit: to answer the original question, only one capacitor is needed, and it does not matter if it sits on the + or - side. The resistor can probably be removed if the capacitor is large enough.
Seriously, there are two ways of handling the effects of a passive filter (which is what this is) in an active design. Either you calculate/simulate the response and consider it when designing the active filter, or you just make the passive crossover frequency so low that you don't have to.
The latter may be perfectly ok if the purpose is to block DC from entering the tweeter, on the other hand the capacitor has to be bigger (=more expensive, possibly an electrolyte). Also, I like the thought of considering the summed effect of EVERYTHING, including the drivers, when designing loudspeakers. After all, it is the sum of everything we listen to.
Edit: to answer the original question, only one capacitor is needed, and it does not matter if it sits on the + or - side. The resistor can probably be removed if the capacitor is large enough.
here is a useful calculator to determine cap size/xover frequency, applicable to tweeter protection.
For my active setup, which is 4th order crossed over at 2250hz with an 8ohm tweeter, I'm using a 20uf cap.
For my active setup, which is 4th order crossed over at 2250hz with an 8ohm tweeter, I'm using a 20uf cap.
Hi everyone,
The picture I posted is not with my tweeter. I just found it searching for pictures showing how to wire a capacitor to the tweeter to prevent any DC from damaging it. For capacitors I'd use 100uF Mundorf Mcap which would give me a high-pass filter point to about 400 Hz which is far away enough from the actual crossover point.
Andrew, the filter point introduced by the capacitor should be at least 2 octaves away from the actual crossover point. If the value of the capacitor is to small it will cause phase shift problems. So for a crossover point at 3 kHz you would need for an 8 Ohm tweet a capacitor of at least 26.5 uF and for a 4 Ohm one double that.
Anyway, concerning my question... Should I just solder the capacitor between 2 cable parts and then connect the resulting cable with the capacitor inserted in it to the speaker terminal and the tweeter?
The picture I posted is not with my tweeter. I just found it searching for pictures showing how to wire a capacitor to the tweeter to prevent any DC from damaging it. For capacitors I'd use 100uF Mundorf Mcap which would give me a high-pass filter point to about 400 Hz which is far away enough from the actual crossover point.
Andrew, the filter point introduced by the capacitor should be at least 2 octaves away from the actual crossover point. If the value of the capacitor is to small it will cause phase shift problems. So for a crossover point at 3 kHz you would need for an 8 Ohm tweet a capacitor of at least 26.5 uF and for a 4 Ohm one double that.
Anyway, concerning my question... Should I just solder the capacitor between 2 cable parts and then connect the resulting cable with the capacitor inserted in it to the speaker terminal and the tweeter?
Does anyone disagree with the two octave rule?oceanic30 said:
Svante recommends reducing the active filter slope by one pole and integrating the protection cap into the required overall filter type. I like this, except for the accuracy required and the fixed Q=0.71 that must be compensated by altering the Q of the active filter. Although LR2 can be achieved with an active single pole (q=0.71) and passive single pole (q=0.71). LR4 and Butterworth require more thought.
yes and use hot melt glue or similar to secure the cap to something rigid, otherwise the solid core wire will fatigue and break.oceanic30 said:
What if the tweeter is a piezo type?
It is already, substantially, a capacitor. Is a protection capacitor also required after the filtered signal coming from the power amp.
What DC voltage can these piezo tweeters survive? Effectively no power is transferred and no heating occurs.
I have a pair of irreplaceable Audax HD3P that I would like to go active with.
The slight complication with the HD3P is that it is preceeded with an air-cored transformer and a 4r0 resistor to kid the amp on that it is an 8ohm load.
I fancy driving from a ClassA amp High Pass filtered at 6kHz and feeding the transformer direct. Maybe even omit the transformer.
It is already, substantially, a capacitor. Is a protection capacitor also required after the filtered signal coming from the power amp.
What DC voltage can these piezo tweeters survive? Effectively no power is transferred and no heating occurs.
I have a pair of irreplaceable Audax HD3P that I would like to go active with.
The slight complication with the HD3P is that it is preceeded with an air-cored transformer and a 4r0 resistor to kid the amp on that it is an 8ohm load.
I fancy driving from a ClassA amp High Pass filtered at 6kHz and feeding the transformer direct. Maybe even omit the transformer.
AndrewT said:
I'd asked various people about this, and read and heard the more the better (withing reason) -- although anything from 'around' 1.5 octaves is just fine if using a 4th order active. At the same time, was also advised that a slightly lower value film cap is better than a more ideally valued bipolar electrolytic cap due to long-term stability issues.
Cannot back this up with any experience or measurements, it's just what I've read and heard from various people (dealers included).
AndrewT said:
Mja... Dunno if I would propose decreasing the slope of the active filter as a general metod. I just meant that the effects of the passive components should be considered. IMO it is easy to neglect the acoustic effects of the driver, and when it comes to this hybrid type of active/passive filters, it is also easy to neglect the effects of the passive components. Adding a capacitor does not nessecarily result in one extra real pole, it will also result in a tendency to a peak near the driver resonance. It was this effect I had in mind when I commented on the parallel resistor.
Anyway, here is a simulation of a reasonably real tweeter with a rather high fs. The larger capacitor results in a cutoff of 700 Hz electrically according to textbook examples (assuming a purely resistive tweeter). The smaller capacitor results in a cutoff of 1400 Hz. It can be seen on the lowest curve group that the voltage across the tweeter terminals does not follow a standard 1st order HP filter function. In particular, the pink curve has a peak near fs, and it is larger than the light blue peak. This is because the parallel resistor reduces the impedance peak of the driver slightly, even though it does not eliminate it.
It can also be seen in the impedance curves (middle group) that the blue curves (=those with the resistor) presents an unnessecarily low impedance for the amplifier.
So, what is "best" is up to the designer, but IMO all of these effects as a whole should be considered. Changing one component might be just right to compensate for another etc.
In this case it might be easiest to stick with the dark red curve, ie a large enough capacitor in series with the tweeter. Surely it removes low frequencies good enough. However, there will still be a need to integrate the responses of the active filters with the responses of the drivers.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.