Hi Thanks for reading I've seen a few three-way speakers that use 400-500uf shunting caps in the bass filter.
IE A second order crossover comprising of large series inductor and a parallel connected 400-500uf capacitor.
Why does this large capacitor not damage the amplifier because of low impedance dangers?
The average reactance of a 400uf capacitor at 400hz is about one ohm, add that to the one ohm or so reactance of the inductor and you have a reactance of about 2 ohms, a possible low impedance danger ?
IE A second order crossover comprising of large series inductor and a parallel connected 400-500uf capacitor.
Why does this large capacitor not damage the amplifier because of low impedance dangers?
The average reactance of a 400uf capacitor at 400hz is about one ohm, add that to the one ohm or so reactance of the inductor and you have a reactance of about 2 ohms, a possible low impedance danger ?
No idea. If it were a second-order Linkwitz-Riley type of crossover, the absolute values of the reactances of the capacitor and inductor should be twice the loudspeaker impedance at the crossover frequency, but that means about 25 Hz crossover frequency when you have 500 uF and 6 ohm.
There are in fact (just have a look at Stereophile, or the manufacturer's own material for that matter) quite a number of 'high end' loudspeakers where impedance drops in the LF to around 2ohms. It's not uncommon to see this combined elsewhere in the range with a highly reactive load as well. Which is lousy design from my POV -they're obviously assuming it will be paired with big amplifiers that have the current and OA stability to handle it, in the same way that some software producers have historically tended to rely on cutting-edge hardware to prop up non-optimised coding. Such is the state a lot of very expensive commercial gear has got itself into.
2 ohms may not be too bad. Most amps can handle 4 ohm loudspeakers as the spec sheet usually states, but they mean a speaker with a nominal mpedance of 4 ohms, the minimum impedance tends to be about 2 ohms less.
There is usually a speaker resonant peak at 400hz which should be dampened down as much as possible, then the impedance dips before and after this peak, usually dipping down to 2 ohms at approximatly 1khz, for some speakers the impedance can go down to 1.8 or 1.5 ohms.
That's why i like to crossover at 1khz, this is usually at the point where the driver impedance is at its lowest which means a smaller inductor is required, big inductors tend to colour the sound.
This time around i am thinking about crossing over before the resonant peak, under 400hz, but this is where the current levels are higher and more care needs to be taken. So i'm guessing i need to rely on the reactance of the inductor to make sure the reactance does not go under 2 ohms. The reactance of the shunting capacitor is going to be very low at sub 400hz frequencies
There is usually a speaker resonant peak at 400hz which should be dampened down as much as possible, then the impedance dips before and after this peak, usually dipping down to 2 ohms at approximatly 1khz, for some speakers the impedance can go down to 1.8 or 1.5 ohms.
That's why i like to crossover at 1khz, this is usually at the point where the driver impedance is at its lowest which means a smaller inductor is required, big inductors tend to colour the sound.
This time around i am thinking about crossing over before the resonant peak, under 400hz, but this is where the current levels are higher and more care needs to be taken. So i'm guessing i need to rely on the reactance of the inductor to make sure the reactance does not go under 2 ohms. The reactance of the shunting capacitor is going to be very low at sub 400hz frequencies
There used to be a DIN standard stating that the impedance should not go below 80 % of the nominal impedance. 2 ohm for a supposedly 4 ohm loudspeaker is poor design.
I did have a quad 405 and 405 II, both amps have current limiters so low impedance is usually not a problem for them
It's quite common nowadays, if you look at the impedance curve and frequency response graph for most 4 ohm systems, the impedance goes down to 2 ohms and even less. Guess in the days when DIN came out it was mostly 8 ohm amplifiers, nowadays amps that can handle speakers with a nominal impedance of 4 ohms are abundant.
So...do you think with the low reactance of the shunting capacitor it is really up to the series inductor to provide the necessary reactance to make sure the speaker impedance does not dip too low. Signals that go through the inductor and drive unit should be fine. It's signals that choose to go through the shunting capacitor that could be a low impedance problem, or have i got it totally wrong
Probably the current limiters would not kick in at low volumes.
Did you find anything contrary, did you find the sound quality improved with the limiters removed even when lstening at low volumes, heard on a few of the other forums that removing the limiters is supposed to be a good thing
That would normally be the case in the low-pass branch of a crossover filter, but 500 uF would correspond to an unusually low crossover frequency, so I don't understand what those capacitors are used for.
I didn't notice any difference but don't like limiters and also removed them from my Krell KSA50 clone. They are fitted to prevent warranty claims and any subsequent adverse publicity in the HIFI press. They are not there to enable you to overload an amp without any audible side effects.
It is the reactance of the series inductor that isolates the parallel capacitor from affecting the other drivers as well as impeding unwanted high frequencies.
It is the reactance of the series inductor that isolates the parallel capacitor from affecting the other drivers as well as impeding unwanted high frequencies.
I haven't come across such high values. Usually the LF drivers seem to have less than 100uf.
I think in this case the 400-500 uF does not belong to the crossover filter, but the capacitance of an impedance correction circuit for the woofer section and therefore modifies the impedance differently.
Indeed, so presumably it's just part of the LP. You do sometimes see oversized input RC Zobels that are used to help reduced an LF resonant peak, without being specifically targeted as an LCR would be. 400uF - 500uF would be large even by those standards though, and of course that requires the resistor, so back to square one again... You might get a series cap of that size on a bass leg in a sealed box, in ye olde method of adaping the alignment of an acoustically undersized box, but that's a different matter entirely. I suppose it's possible that a high[ish] Q electrical order is being used to help EQ a given on-baffle response to a desired shape, since what might be peaking in the electrical domain isn't necessarily so acoustically, but we're in the realms of pure speculation & there's never been much profit in that.
Dropping to 2ohms is -inept, in most cases, at least as far as I'm concerned. My own blanket rule (just my own -I'm not suggesting universal applicability 😉 ) is an impedance minimum of 3ohms is the lowest permissible, and then only if it isn't associated with a large shift in the electrical phase angle.
Dropping to 2ohms is -inept, in most cases, at least as far as I'm concerned. My own blanket rule (just my own -I'm not suggesting universal applicability 😉 ) is an impedance minimum of 3ohms is the lowest permissible, and then only if it isn't associated with a large shift in the electrical phase angle.
Thanks! If I had a low impedance problem where I thought the impedance of the speaker may dip below 4 ohms and possibly damage the amp, could I place a resistor in series with the speaker to address this.
I’ve seen a few youtube videos in which youtubers are testing amps and placing an 8 ohm resistors across the amplifier output, rather than using a speaker, dummy load they called it.
Is this possible, because the resistor provides resistance not impedance, why is the amp not being damaged
I’ve seen a few youtube videos in which youtubers are testing amps and placing an 8 ohm resistors across the amplifier output, rather than using a speaker, dummy load they called it.
Is this possible, because the resistor provides resistance not impedance, why is the amp not being damaged
A design strategy where this often makes sense is a large midrange (6-8") crossing over to an even larger woofer, and a 12db/octave high pass on the midrange needs to boost output in the 80-150Hz range because the designer has to correct for baffle step. Using a large capacitor in the LC circuit raises the Q of the filter so that in that frequency band, the output gets boosted 3-6dB.
This is accomplished by a low impedance reactive LC circuit, which draws extra current from the amp and forces it to to be dissipated in the midrange driver.
This is often what's going on in the Stereophile reviews that @Scottmoose is referring to. Impedance dips down to 2 ohms in the upper bass or lower midrange.
I apply the exact same technique in the thread "Open Baffle Bass Boost: +4 to +7dB w/ Passive Xover, No DSP" to boost the bass of an Open Baffle woofer (instead of a midrange). It also protects the woofer from subsonic signals. The link discusses this in detail.