Hi
I am designing the class-d audio amplifier. I have a question on the loadspeakers. I see some papers assert as below:
.............
Loudspeakers are often modeled as resistors but in reality they are highly reactive. A typical 8 loudspeaker can have impedance minima as low as 4 and also behave nonlinearly.
.............
I wonder if the loudspeaker is 4 Ohm load, then can we think the minimum impedance is as low as to 2 Ohm? Can we have this conclusion?
I'm no loudspeaker expert, but conventional speaker systems do have large impedance variations with frequency. For example, I measured two different home speakers. Both models have one woofer, one midrange, and two different tweeters. I think the type A-III was "4 ohms" nominal, and the type C was "8 ohms" nominal, but I don't have the original brochures.
Yes, you can see that this a "bassreflex" design, having two resonance frequencies instead of one. Closed enclosure would have only one low frequency resonance. Regarding the original question, the nominal impedance is mainly defined by drivers voice coil resistance. Values below nominal impedance are caused by crossover circuit, the said half would be a realistic guess, but not a lower limit you can rely on.
It's been many years since I had the type C's, but I vaguely recall a small tuned port with relatively low output. The type A-III is sealed - only one bass resonance hump.
Frequency points with low impedance and large phase angle are stressful on linear amps with BJT output drivers. My old Heathkit amp has series pairs of BJTs to reduce secondary breakdown stress. That's less of a concern with switching amps.
It's been many years since I had the type C's, but I vaguely recall a small tuned port with relatively low output. The type A-III is sealed - only one bass resonance hump.
Frequency points with low impedance and large phase angle are stressful on linear amps with BJT output drivers. My old Heathkit amp has series pairs of BJTs to reduce secondary breakdown stress. That's less of a concern with switching amps.
Hi echo47,
Thanks for the reply. It is very helpful to me.
Since the impedence varied vs. the frequency. In the simulation and designing of D amolifier, I always modeled the 8 or 4 Ohm loudspeaker as 8 ot 4 Ohm resostor in the simulation. It did not contains the frequency characteristics. Is there any risks in it?
I don't have much experience with class-D amps, sorry. Hopefully someone else with more experience can offer their advice.
However, it seems like a good idea to run some simulations using difficult reactive loads (low impedance at large phase angle), to be sure your amplifier remains happy and stable even though the output voltage and current are out of phase with each other.
Class-D is operating with reactive load (output filter) anyway and would completely ignore the load . The shown impedance are rather uncritical, cause they have no impedance dips as sometimes observed with crossover. They have basically the voice coil resistance as a minimum, friendly load to an amplifier. They could be at the most an issue for an overdriven and undersized class AB amplifier.