[virtue]

Tommy,

A few "newbie" questions for you!

Do the ultrasonics that most tweeters can reproduce these days cause listening fatigue?

Is most source material bandwidth limited?

When Stereophile uses exotic rigs to shunt ultrasonics how should we think about that?

What else should we know about ultrasonics?

Seth

[Speedskater]

Well ultrasonics can cause IM distortion. And that distortion is in the normal audio band.

[AmpDesigner333]

Tommy,

A few "newbie" questions for you!

Do the ultrasonics that most tweeters can reproduce these days cause listening fatigue?

Is most source material bandwidth limited?

When Stereophile uses exotic rigs to shunt ultrasonics how should we think about that?

What else should we know about ultrasonics?

Seth

Do the ultrasonics that most tweeters can reproduce these days cause listening fatigue?
If you're referring to the ultrasonic switching ripple of a Class-D amplifier, then the answer is no.  The frequencies we're talking about (such as 1MHz) are WAY above the bandwidth of practical tweeters.

Is most source material bandwidth limited?
Yes, but source bandwidth is now extended thanks to higher sampling rates.  Also, don't be confused by the high rates of DSD.  In order for the waveform to be reconstructed by a DSD D/A, it must be filtered heavily.  That's another topic altogether.

When Stereophile uses exotic rigs to shunt ultrasonics how should we think about that?
Please provide a link to more info on this (what exotic rigs?).  It may just be for measurement purposes.  The high frequency ripple from Class-D amps can drive some measurement equipment (audio analyzers) nuts, and filtering it out before it gets to the analyzer is often required to get valid results.

What else should we know about ultrasonics?
A few quick things:
- Different switching amps have different ultrasonic "footprints".  A high frequency FFT can show these.
- The switching ripple of Class-D can dissipate power in the tweeter, "warming it up".
- Amps with heavy output filtering to reduce output ripple often cut into the audio band.

[*Scotty*]

How many filter poles do your amplifiers use in the output filter and what is the corner frequency? Do you find that there is an audible consequence to a higher parts count in the output filter?
Scotty

[barrows]

This is an interesting topic to me regarding class D amplifiers in general, as I suspect the design and implementation of the output filter is critical to the ultimate perceived sonic performance of the amplifier.

This:

"Do the ultrasonics that most tweeters can reproduce these days cause listening fatigue?"
"If you're referring to the ultrasonic switching ripple of a Class-D amplifier, then the answer is no.  The frequencies we're talking about (such as 1MHz) are WAY above the bandwidth of practical tweeters."

Seems like a possibly incomplete answer.  While I agree that no tweeter is going to be able to reproduce frequencies of the switching carrier (typically >300 kHZ or thereanouts on most class D amps) or its high harmonics.  We can see that these are radio frequencies, and given the slightest chance, they will go airborne.  Some believe humans can perceive such ultra high frequencies-yes this is getting into the realm of quasi science at this point, but I still feel it deserves consideration.  Perhaps some very sensitive humans are more disturbed by the presence of airborne RFI than others, and perhaps this airborne RFI can cause listening fatigue, and/or a general sense of being ill at ease.  In my experience with various class D amplifiers, attempting to to damp RFI has often produced subtle, but notable improvements in sound quality. 

As to high leverls of UHF producing IM distortion, this would be seen in the IM measurements, yes?  IM measurements for class D amps look pretty good to me...

[*Scotty*]

If you want to be disturbed by RFI, try looking at your AC power on your oscilloscope. A friend of mine measured 250mv of RFI noise on the AC power in his home. If your house wiring is not in a conduit, just Romex running in holes in the studs in your walls,then you are surrounded by a RFI carrying net of wire.
I think the switching amps contribution to radiated RFI via the speaker wires may be the least of our worries.
Scotty

[barrows]

Scotty:  I hear ya.  I have seen my AC on a 'scope, but it has been awhile.  I live in a somewhat rural area, share my AC transformer with no other homes, and have a fairly modest house, with limited at home producers of RF energy.  Additionally, there is no cell service up here either.  So, relative to a lot of people, the RFI pollution here is at a minimum.  Still tune an AM radio between stations and it is far from just quiet here!
Anyway, I do feel my home environment is a good one from an RFI perspective, hence I think I may be able to notice more subtle RFI sonic degradation than those living in more urban areas.  The worst is setting up at an audio show!

[*Scotty*]

I am in a 300 unit condo building on the 12th floor with who knows what in the air as well as in the walls. To be honest I haven't even looked at it. I built Felix power filters, Occam's design, and judging by the positive sonic results they have knocked down the mess quite a bit. It was much harder to hear circuit changes or cabling changes when the full load of RFI on the AC was unattenuated.
Scotty

[Speedskater]

When it comes to AC power lines and EMI/RFI interference, the two go-to experts are Bill Whitlock and Jim Brown. Both have been the AES committee chair on EMI/RFI interference. Their many papers contain much better information & advice than typical internet ramblings.

Bill Whitlock:

2012  "An Overview of Audio System Grounding and Interfacing"
http://centralindianaaes.files.wordpress.com/2012/09/indy-aes-2012-seminar-w-notes-v1-0.pdf

2005  "UNDERSTANDING, FINDING, & ELIMINATING GROUND LOOPS IN AUDIO & VIDEO SYSTEMS"
AN007    Generic Student Seminar Handout
http://www.jensen-transformers.com/application-notes/

2010  "Ground Loops: The Rest of the Story"
ABSTRACT
The mechanisms that enable so-called ground loops to cause well-known hum, buzz, and other audio system
noise problems are well known. But what causes power-line related currents to flow in signal cables in the first
place? This paper explains how magnetic induction in ordinary premises AC wiring creates the small voltage
differences normally found among system ground connections, even if “isolated” or “technical” grounding is
used. The theoretical basis is explored, experimental data shown, and an actual case history related. Little
has been written about this “elephant in the room” topic in engineering literature and apparently none in the
context of audio or video systems. It is shown that simply twisting L-N pairs in the premises wiring can
profoundly reduce system noise problems.
http://xa.yimg.com/kq/groups/20963848/268252969/name/Whitlock-Fox+-+Ground+Loops+.pdf

***********************************

Jim Brown:

"Power and Grounding For Audio and Audio/Video Systems -- A White Paper for the Real World"
http://www.audiosystemsgroup.com/SurgeXPowerGround.pdf

"
RFI, Ferrites, and Common Mode Chokes For Hams"
ignore the word Ham, as much of the paper also applies to audio systems
http://www.audiosystemsgroup.com/RFI-Ham.pdf

About 50 more good papers & Power Points.
http://www.audiosystemsgroup.com/publish.htm

[AmpDesigner333]

Good points brought up here.

My 2 cents on RFI: Immunity is KEY!   (a.k.a. "lack of susceptibility").  Well designed components incorporate counter measures to handle conducted (through connections) and emitted (through the air) RFI.  There is no such thing as 100% immunity, but I believe it's possible to reduce susceptibility such that interference in practical environments will not cause audible effects.

The harmonics of the switching frequency go way out to 100s of MHz in measurable "spikes".

In amplifiers like Tripath, Ice power, Ncore, and Cherry (Maraschino and Classic), the switching frequency is not fixed, and generated RF is modulated by the audio in ANY Class-D.

Another quick note, any amplifier that uses its output devices as switches IS CLASS-D.  Class-xyz is just a marketing thing, and not recognized by the engineering community as an actual amplifier type class designation.

[AmpDesigner333]

When it comes to AC power lines and EMI/RFI interference, the two go-to experts are Bill Whitlock and Jim Brown. Both have been the AES committee chair on EMI/RFI interference. Their many papers contain much better information & advice than typical internet ramblings.

Bill Whitlock:

2012  "An Overview of Audio System Grounding and Interfacing"
http://centralindianaaes.files.wordpress.com/2012/09/indy-aes-2012-seminar-w-notes-v1-0.pdf

2005  "UNDERSTANDING, FINDING, & ELIMINATING GROUND LOOPS IN AUDIO & VIDEO SYSTEMS"
AN007    Generic Student Seminar Handout
http://www.jensen-transformers.com/application-notes/

2010  "Ground Loops: The Rest of the Story"
ABSTRACT
The mechanisms that enable so-called ground loops to cause well-known hum, buzz, and other audio system
noise problems are well known. But what causes power-line related currents to flow in signal cables in the first
place? This paper explains how magnetic induction in ordinary premises AC wiring creates the small voltage
differences normally found among system ground connections, even if “isolated” or “technical” grounding is
used. The theoretical basis is explored, experimental data shown, and an actual case history related. Little
has been written about this “elephant in the room” topic in engineering literature and apparently none in the
context of audio or video systems. It is shown that simply twisting L-N pairs in the premises wiring can
profoundly reduce system noise problems.
http://xa.yimg.com/kq/groups/20963848/268252969/name/Whitlock-Fox+-+Ground+Loops+.pdf

***********************************

Jim Brown:

"Power and Grounding For Audio and Audio/Video Systems -- A White Paper for the Real World"
http://www.audiosystemsgroup.com/SurgeXPowerGround.pdf

"
RFI, Ferrites, and Common Mode Chokes For Hams"
ignore the word Ham, as much of the paper also applies to audio systems
http://www.audiosystemsgroup.com/RFI-Ham.pdf

About 50 more good papers & Power Points.
http://www.audiosystemsgroup.com/publish.htm

Thanks for the links!

[AmpDesigner333]

How many filter poles do your amplifiers use in the output filter and what is the corner frequency? Do you find that there is an audible consequence to a higher parts count in the output filter?
Scotty

It's not necessarily the parts count in and of itself.  It's the quality of the components, layout, values of the parts, tuning, and most importantly, how it's being driven.  Our output filters vary from 2 to 4 pole and cutoff is >100kHz on Maraschino.  Classic Cherry amps are >50kHz.  We make custom parts for some designs.