Some food for thought about the effect of high frequencies in audio power amplifier

[Scytales]

Of course RF can enter a unit as well through the output.
The mandatory EMI capacitor at the line output jack and the usual series resistor between the actual output stage and the output jack should usually suffice.

Fun (sad ?) fact : through studying a lot of schematics and actual implementations in various apparatuses, mostly CD/SA-CD/DVD players, I found that EMI capacitors at the output jacks are provisioned in the schematics and PC boards, but not populated in actually produced units.

I do not know why. Perhaps, the manufacturers skimp on that components to save cost as long as a specific unit pass EMI regulation without them ?

 

[nanook]

The EMI caps have to be right at the jack to be most efficient (a wire inside the unit will serve as antenna when the cap is located further away).
Some designer may fear that the capacitor at the line output may create a resonant system with the inductance of the cable and possibly a capacitor right at the input of the receiving unit.
Cost should not really be a thing, but you never know.

 

[Cbdb2]

View attachment 379620

This is interesting and tells the story. Some quick calcs. For every 5db drop in RF input the output drops by 20db! When the input is down to -35dbv (still 18mv RMS) output on the worst Amp is already -100dbv. Typical antennae pickup is less than a 1mv (-60dbv) which would produce an output at -180dbv. Thats from an antennae, a shielded coax should drop the input another 20db.
Yea if your next to a 10Kwatt radio transmitter etc. you might have a problem otherwise nothing to worry about.

 

[nanook]

Well, you either have a new car (with the radio built according to todays EMI requirements) or you may hear that your mobile phone is about to ring a couple of seconds before it actually does. This at least holds true for my car radio built about 24 years ago ;-)

 

[HaveMeterWillTravel]

I welcome the Audio world to the well-mature world of EMI/EMC testing!

There are already established and vetted testing standards out there. EN/IEC 60268 has been harmonized in most of Europe. 60268-3 pertains to testing amplifiers. It puts the FTC's testing to shame.

 

[Chrispy]

I welcome the Audio world to the well-mature world of EMI/EMC testing!

There are already established and vetted testing standards out there. EN/IEC 60268 has been harmonized in most of Europe. 60268-3 pertains to testing amplifiers. It puts the FTC's testing to shame.

Got some links to this world?

 

[HaveMeterWillTravel]

Got some links to this world?

Yes and No.

The primary difficulty with most Standards is that the sanctioning body or regulatory agency typically charges for them. Often quite a bit.

Here is an introduction video by TI on several IEC 61000-4-X standards relative to audio electronics. in this case, they are testing an ADC with integrated signal conditioning for use in an ATE (automated test equipment).

The IEC 61000-4-4 and -6 would be most applicable to this thread's topic. They start around the 10 minute mark.

 

[Scytales]

I just discovered some interesting efforts made by Paul Miller back in the '90s to assess the susceptibility of amplifiers to RFI. They were published in Hifi Choice and can be read on the marvelous Worldradiohistory.com website.

Paul Miller named that measurement 'Broad-Band Radio Frequency Inter-modulation Test'.

Some examples :

• Issue #85 (see page 84 and the tests of many amplifiers in the previous pages)
• Issue #92 (see page 94 and the tests of many amplifiers in the previous pages)
• Issue #104 (see page 76 and the tests of many amplifiers in the previous pages)

In issue #104, Paul Miller referred to a paper submitted to an AES lecture, but I was unable to find it.

Paul Miller's test seems more interesting than the simpler tests devised by the Japanese I quoted above, because he used a carrier RF signal of "only" 20 mV peak-to-peak (-43 dBV or 7.07 mV RMS) modulated to a depth of 100%, ie a much lower and non-steady excitation signal.

 

[Speedskater]

Two noise and interference experts:

Keith Armstrong

Cherry Clough Consultants

Cherry Clough Consultants is an independent company that helps manufacturers, installers, and users deal cost-effectively with interference, safety, and comply with European Directives and similar regulations

www.cherryclough.com

Henry Ott (RIP)

https://www.amazon.com/Electromagnetic-Compatibility-Engineering-Henry-Ott/dp/0470189304

 

[Drain.pipe]

I lived next to a Naval Air Station and was unable to receive broadcast TV (digital) without frequent interference, despite being a straight shot to the transmitters.

I worked for a company that was a couple blocks from an AM station. Their broadcast was received by the PA speakers.

I used to be in a band that rehearsed within about 1km of an AM transmission tower. Occasionally the base amp with would pop and the radio station would be clearly audible. A rock and roll back and forth a few times without usually correct it...

 

[AnalogSteph]

I used to be in a band that rehearsed within about 1km of an AM transmission tower. Occasionally the base amp with would pop and the radio station would be clearly audible. A rock and roll back and forth a few times without usually correct it...

Sounds like a bad solder joint that had a habit of turning into a diode junction whenever only the surface oxide layer was connecting both ends.