Yes, it's the relative arrival time at the ears that matter. Besides, only a linear phase shift (constant delay) is necessary for composite signal integrity. Designs that achieve low phase-shift using a response similar to post #230 would suffer from an overshoot (and possible ringing) in the corresponding time-domain transient response.
Yes, but you would hear a difference at the xover point to the midrange if you swapped over the + and - of both your tweeters, as a hole or peak. that is the same as the phase shift does here in the Icepower 1200 on both channels.
https://www.diyaudio.com/community/...013590/?hash=cfa7aeff90ab22da542db2925974d489
And like I said
"The IcePower 1200 is now over 4 years old, they are not stupid, "if" that graph of the phase shift was wrong or mislabeled, they would have corrected it long before now. And it's the same phase shift on both mono 1200AS1 and stereo 1200AS2 modules on their own website data sheets"
Cheers George
Sorry not by me, just different amounts of phase shift in each situation in both channels. 180' vs 70' in or out of phase to the mids.
Cheers George
In the wrongly wired xover there is cancellation because you are superimposing the same signal - in two transducers - with different phase rotation.
Phase rotation in an amplifier is one single signal. It does not cancel with anything.
When both tweeters are wired out of phase with similar order xover, it's similar to have this amp phase shift powering both tweeters, phase shift depending on the roll off rate if similar, doesn't matter if it's done in an external line level xover, in the amp or at the speakers xover, phase shift is phase shift.
Looks like it's happening at around 10dB per octave which is similar to a tweeter 2nd order high pass xover , and if you swap both tweeters polarity around you will hear it, just like this amps phase shift.
Cheers George
Last edited:
but the amplifier has only a phase rotation, not a rolloff (not in the audio band anyway).
sigh
Phase shift angles are similar.
And it all boils down to what started this, "would you hear it" I say yes, as both channels are 70' out of phase at 10khz
Cheers George
This "phase shift", caused is by the Class-D's low pass low order output filter, to eliminate 600khz switching frequency noise from getting to your speakers.
Could be taken further up out of the audio band, if the switching frequency and output filter for it were at 1.5MHz instead of 600kHz,
Technics say they do this in their impossible to get $$$$$$$ flagship Class-D power amp the SE-R1 which uses the much faster GaN transistor technology. This the puts that phase shift higher than 20khz instead of right in the tweeters operating range.
Cheers George
Could be taken further up out of the audio band, if the switching frequency and output filter for it were at 1.5MHz instead of 600kHz,
Technics say they do this in their impossible to get $$$$$$$ flagship Class-D power amp the SE-R1 which uses the much faster GaN transistor technology. This the puts that phase shift higher than 20khz instead of right in the tweeters operating range.
Cheers George
Last edited:
You are conflating two related yet completely different effect:
1. one is the ADDITION of two copies of the same signal, not in phase, at the crossover frequency, which CAUSES a cancellation if the phases are too far apart
2. The other one is the effect of a low pass filter in class D amps, and any low pass filter CAUSES also some phase rotations, but it is the attenuation that is audible, not the phase rotation.
In the first case the attenuation in the audio band is desired anyway at the xover frequency otherwise there would be a bump - if the signals are in phase. In the second the filter has the cutoff frequency far enough from the audio band that the attenuation at 20hz is negligible — yet you state “right in the audio band”.
This, and your fixation in higher switching rate GaN class D amps, shows that you are absolutely clueless.
Your belief, and you are entitled to it, but without being "offensive" thanks, next time I won't hold back.
New class D amplifier
ICEpower Conductor
https://icepoweraudio.com/the-hunt-for-better-audio-performance/
PHASE: 15º at 20 kHz, much better than before but...
You know, the phase is important to me (this week I linearized the phase of my KEF Q100 coaxial modified with the help of rePhase: I notice the difference and... I like more the the sound!).
ICEpower Conductor
https://icepoweraudio.com/the-hunt-for-better-audio-performance/
PHASE: 15º at 20 kHz, much better than before but...
You know, the phase is important to me (this week I linearized the phase of my KEF Q100 coaxial modified with the help of rePhase: I notice the difference and... I like more the the sound!).
Measuring amplifiers
https://alpha-audio.net/academy/measuring-amplifiers/
Hypex Nilai: Frequency vs Phase
https://alpha-audio.net/academy/measuring-amplifiers/
Hypex Nilai: Frequency vs Phase
And when they measuered the linear MSB 202 power amplifier in that Amplifier Measurment's article, they also said.
"The phase measurement shows that up to 20 kHz there is almost no phase shift (max 14 degrees.) Class D does much less well here because of the filter that is needed. Every filter causes phase shifts. That includes a filter in an amplifier."
Cheers George
"The phase measurement shows that up to 20 kHz there is almost no phase shift (max 14 degrees.) Class D does much less well here because of the filter that is needed. Every filter causes phase shifts. That includes a filter in an amplifier."
Cheers George
Phase go back and forth in different units in a system. 2023 it's time to take a more scientific approach to a correct impulse behaviour and measure the whole system by letting the used source play a sweep and measure it in the listening position. Then one use a DSP to correct the whole system in one go so that it has falt FR and correct ideal phase response. Then the amp phase behaviour doesn't matter (as long as it is stable). This is how it is done.
Fiddling with only one individual part is so clueless.... why?
Probably much greater deviations in the speakers.
//
Fiddling with only one individual part is so clueless.... why?
Probably much greater deviations in the speakers.
//