I have been talking about loudspeakers as a load, at low, moderate, and at large drive levels.
An underhung woofer at a Fixed Frequency is a fairly constant load until either of two things happen:
1. The peak excursion causes some of the voice coil to go beyond the magnetic pole pieces.
2. The peak excursion causes either the outer suspension, or the spider to go beyond where Hook's constant is true (linear spring, versus non-linear spring).
But even at low and moderate levels where the excursion is linear, depending on the frequency, that load can be R, L, RL, or RC.
At large signals, there may be saturation of the crossover coil's core (if it is not an air core coil).
An underhung woofer at a Fixed Frequency is a fairly constant load until either of two things happen:
1. The peak excursion causes some of the voice coil to go beyond the magnetic pole pieces.
2. The peak excursion causes either the outer suspension, or the spider to go beyond where Hook's constant is true (linear spring, versus non-linear spring).
But even at low and moderate levels where the excursion is linear, depending on the frequency, that load can be R, L, RL, or RC.
At large signals, there may be saturation of the crossover coil's core (if it is not an air core coil).
That has worked for me.
However, in the past I had a Vector Network Analyzer to use.
I have measured some loudspeakers impedance and phase angles versus frequency on that VNA.
Then I used the VNA to measure the frequency response and phase of the amplifier into a resistive load.
I also measured the (electrical) frequency response and phase of the amplifier connected to the loudspeaker using the same VNA. Again, this was electrical, not acoustic.
It was quite enlightening.
However, in the past I had a Vector Network Analyzer to use.
I have measured some loudspeakers impedance and phase angles versus frequency on that VNA.
Then I used the VNA to measure the frequency response and phase of the amplifier into a resistive load.
I also measured the (electrical) frequency response and phase of the amplifier connected to the loudspeaker using the same VNA. Again, this was electrical, not acoustic.
It was quite enlightening.
The main thing I'd worry about is if the reactance of the speaker load could cause spurious oscillations or ringing in the power amp, due to it causing a shifting in time (phase shift) of the counter electromotive force (CEF), generated by the speaker and its passive crossover components, at certain frequencies, as perceived by the amplifier feedback loop.
Testing poweramps with sinewaves is conventional and a good starting place. I then check them with a 10kHZ squarewave and look for ringing at the output, into a non reactive load resistor. There should be very little ringing if any.
I haven't done it, but it might be interesting to then replace the load R with an actual speaker and look at what the speaker does to the 10kHZ squarewave at low levels (1 watt). It could get uglier at higher levels, but as has been stated before, speakers get very loud and burn out quicker than you think with continuous sinewaves or squarewaves.
Back in the 1970's I dealt with many solid state poweramps that had a very poor phase margin, and they would blow up too often, blowing the speaker too. When I got in with the big boys (EE's) at Tektronix in 1980, I learned all I could about phase margin in power amps, and to this day consider it one of the most important things to pay attention to. Speaker reactance is significant, but will vary from speaker system to spkr system, so it's hard to model it. There's a "tug-of-war" between the poweramp output, and the counter electromotive force generated by the reactive speaker. You always want the poweramp to win. Higher feedback is the superficial fix, but can be fooled.
One guy put a 0.5 ohm resistor at the output of his high feedback power amp, feeding the speakers, so the amp feedback would be significantly less affected by the reactance and CEF of the speaker. It throws away a few watts, and slightly changes the calibration of the passive crossover, but is likely to improve amplifier stability.
As far as distortion, reactance and CEF in a speaker, I generally leave that up to the speaker manufacturer since it varies from spkr to spkr, so is hard to characterize. Speaker ringing is reduced by a lower source impedance coming from the power amp, but that lower source impedance can also raise the frequency of the ringing to a less desirable place. We once put a 1 ohm R in series with a speaker system, and the bass then sounded better. Less boomy, more mellow... But we were also dealing with how the room reacted to the change, so as usual there's many variables at play. I hope I'm not too far off subject.
Testing poweramps with sinewaves is conventional and a good starting place. I then check them with a 10kHZ squarewave and look for ringing at the output, into a non reactive load resistor. There should be very little ringing if any.
I haven't done it, but it might be interesting to then replace the load R with an actual speaker and look at what the speaker does to the 10kHZ squarewave at low levels (1 watt). It could get uglier at higher levels, but as has been stated before, speakers get very loud and burn out quicker than you think with continuous sinewaves or squarewaves.
Back in the 1970's I dealt with many solid state poweramps that had a very poor phase margin, and they would blow up too often, blowing the speaker too. When I got in with the big boys (EE's) at Tektronix in 1980, I learned all I could about phase margin in power amps, and to this day consider it one of the most important things to pay attention to. Speaker reactance is significant, but will vary from speaker system to spkr system, so it's hard to model it. There's a "tug-of-war" between the poweramp output, and the counter electromotive force generated by the reactive speaker. You always want the poweramp to win. Higher feedback is the superficial fix, but can be fooled.
One guy put a 0.5 ohm resistor at the output of his high feedback power amp, feeding the speakers, so the amp feedback would be significantly less affected by the reactance and CEF of the speaker. It throws away a few watts, and slightly changes the calibration of the passive crossover, but is likely to improve amplifier stability.
As far as distortion, reactance and CEF in a speaker, I generally leave that up to the speaker manufacturer since it varies from spkr to spkr, so is hard to characterize. Speaker ringing is reduced by a lower source impedance coming from the power amp, but that lower source impedance can also raise the frequency of the ringing to a less desirable place. We once put a 1 ohm R in series with a speaker system, and the bass then sounded better. Less boomy, more mellow... But we were also dealing with how the room reacted to the change, so as usual there's many variables at play. I hope I'm not too far off subject.
Consider a low pass crossover, and what Might happen.
The speaker system's + input terminal connects to a series L, the other end of the L connects to the Driver and C which are in parallel to the speaker's - input terminal.
If the Driver does not load the L and C very much, then we have a high Q series network
(short circuit at LC resonance).
If the amp output is Zero Ohms (Very high damping factor), we have max current at LC resonance. And the high Q of the LC circuit is Retained.
If the amp has a moderate output impedance (moderate damping factor), the Q of the series LC crossover is Reduced, and the current is lessened.
Which one might be more likely to ring for a longer time?
Just asking.
By the way Bob R.,
It is a shame that there is no OTS meeting tonight.
The speaker system's + input terminal connects to a series L, the other end of the L connects to the Driver and C which are in parallel to the speaker's - input terminal.
If the Driver does not load the L and C very much, then we have a high Q series network
(short circuit at LC resonance).
If the amp output is Zero Ohms (Very high damping factor), we have max current at LC resonance. And the high Q of the LC circuit is Retained.
If the amp has a moderate output impedance (moderate damping factor), the Q of the series LC crossover is Reduced, and the current is lessened.
Which one might be more likely to ring for a longer time?
Just asking.
By the way Bob R.,
It is a shame that there is no OTS meeting tonight.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.