I've seen other posters on other sites use the same chart and EE 's keep telling them this chart isn't saying what they claim that ice power 1200 series phase drops off at 2khz. It reminds me of The Princes Bride " I don't think that word means what he thinks it means".
No Class D amps don't have horrible phase shift
There’s a regular poster on Audiogon who has routinely made 2 technical claims about the performance of Class D amplifiers. The first, about the audibility of the switching frequency, and the second, about Class D amplifiers having horrible phase shift. Further, as you will see below, he claims this is audible.
I’d like to take on the second claim, because we can do so purely on objective terms:
The chart that he posts to unfortunately does not support his claim. Nor could it. Unfortunately for this individual he’s misunderstood this chart, which to be fair, is almost NEVER provided by anyone. To understand what this is we need to understand how amplifiers work. Usually, in a linear, solid state amplifier which uses feedback there is output circuitry followed by a parallel RL network. So the performance of the amp is really the amp, this RL network and the speaker. Something like this:
Transistors --> RL Network --> Speaker
The RL network is there to prevent high frequency oscillation. Fun stuff happens when an amp maker excludes it for purity. :)
Anyway, the person’s chart (listed above) is labelled pretty clearly: "Figure 10: Output impedance vs. frequency".
Honestly I’ve never seen anyone plot the magnitude and phase angle of the network on the output. Why on earth B&W / ICEpower would ever ever chart that is kind of a surprise, but if you look closely at the magnitude it is in milliohms. Around 5 mOhms up until ~ 500 Hz. So, damping factor of around 1,600. Outstanding! My point is, this is NOT the chart of the signal at the speaker. It is at that output filter network I mentioned above. Yes, of course, linear amps and Class D have entirely different reasons for the output filter, lets not go down that for right now. This is a family establishment.
Point is, the measurement here is taken across the output stage, not across the output terminals. Confused?
To confound us all, B&O does not seem to have carried this chart forward, nor do they actually plot the output frequency and phase. For that, we need to look at another ICEpower amp, the 250 ASP. Anyone interested in digging in original sources see here:
https://icepower.dk/support/download-page/
For the 250 ASP, which I believe to be typical, phase shift is not 25 @ 2 K, but nearly zero. In fact it doesn’t reach 20-30 degrees until 20 kHz. OK, so not perfect, right? Trouble is, without knowing what these charts look for a normal amp, we can’t really say it’s bad. Stereophile doesn’t measure amp phase output it seems.
My point:
The phase of the output impedance is not the phase of the output and the poster who makes these claims is misinforming. So, this claim is technically simply not correct.
The further claim of the audibility of small amounts of phase shift is not taken up here.
I’d like to take on the second claim, because we can do so purely on objective terms:
[inflamatory content removed ] this is why many find the upper mids/highs objectionable on Class-D, why would I want to listen to an amp with this much phase shift already 25 degrees! at 2khz, and it get’s progressively worse all the way up to 70 degrees!!! at 10khz.
https://ibb.co/VYHFHsQ
The chart that he posts to unfortunately does not support his claim. Nor could it. Unfortunately for this individual he’s misunderstood this chart, which to be fair, is almost NEVER provided by anyone. To understand what this is we need to understand how amplifiers work. Usually, in a linear, solid state amplifier which uses feedback there is output circuitry followed by a parallel RL network. So the performance of the amp is really the amp, this RL network and the speaker. Something like this:
Transistors --> RL Network --> Speaker
The RL network is there to prevent high frequency oscillation. Fun stuff happens when an amp maker excludes it for purity. :)
Anyway, the person’s chart (listed above) is labelled pretty clearly: "Figure 10: Output impedance vs. frequency".
Honestly I’ve never seen anyone plot the magnitude and phase angle of the network on the output. Why on earth B&W / ICEpower would ever ever chart that is kind of a surprise, but if you look closely at the magnitude it is in milliohms. Around 5 mOhms up until ~ 500 Hz. So, damping factor of around 1,600. Outstanding! My point is, this is NOT the chart of the signal at the speaker. It is at that output filter network I mentioned above. Yes, of course, linear amps and Class D have entirely different reasons for the output filter, lets not go down that for right now. This is a family establishment.
Point is, the measurement here is taken across the output stage, not across the output terminals. Confused?
To confound us all, B&O does not seem to have carried this chart forward, nor do they actually plot the output frequency and phase. For that, we need to look at another ICEpower amp, the 250 ASP. Anyone interested in digging in original sources see here:
https://icepower.dk/support/download-page/
For the 250 ASP, which I believe to be typical, phase shift is not 25 @ 2 K, but nearly zero. In fact it doesn’t reach 20-30 degrees until 20 kHz. OK, so not perfect, right? Trouble is, without knowing what these charts look for a normal amp, we can’t really say it’s bad. Stereophile doesn’t measure amp phase output it seems.
My point:
The phase of the output impedance is not the phase of the output and the poster who makes these claims is misinforming. So, this claim is technically simply not correct.
The further claim of the audibility of small amounts of phase shift is not taken up here.
4 responses