What can I do to minimize electromagnetic radiation from power amplifiers?

[mnemonix]

Oh it's you again.

I think a few of these would definitely help.

 

[Wes]

Class D amplifiers do push a ton of switching noise back into the mains and back into all components connected to them. They also radiate switching noise from all their cables. Whether this is harmful or not, is something I've been wondering about. Not about human health, but about the added issues all this noise causes for interconnected components and whether it compromises performance. All the nice measurements you see are low pass filtered so the measuring equipment doesn't see the noise.

a metric ton, or what?

did you measure this? or is it calculated based on freq. dependent attenuation at some point down the AC cable?

I've seen 'people' post the same thing but have never seen either of the above lines of evidence pursued.

I suspect it is a non-issue or a minor issue. But I'd love to have some one else spend their time on it.

 

[raindance]

Easily measurable with a scope on signal lines and chassis of all connected gear. I've seen tens to hundreds of millivolts. I don't know if it's detrimental to performance of any gear, if it's at, say 400KHz, most audio gear isn't going to try to amplify that. Bear in mind, I'm not the OP, I'm simply pointing out that there's noise that exists and it isn't a fallacy. I'm not a 5G conspiracy theorist, I'm an engineer.

 

[raindance]

Additionally, in a past life, I designed switching power supplies for submarines and had to implement a method for passing EMI/EMC tests, so I'm familiar with methods to quell switching noise, just not sure whether it matters at all in this scenario.

 

[Trouble Maker]

Additionally, in a past life, I designed switching power supplies for submarines and had to implement a method for passing EMI/EMC tests, so I'm familiar with methods to quell switching noise, just not sure whether it matters at all in this scenario.

I can venture a guess or two why, but I am curious to hear from you since you worked on it. What is the problem with EMI/EMC in that situation? I'm especially curious about EMI, and also it pertains to this thread.

 

[Chromatischism]

Easily measurable with a scope on signal lines and chassis of all connected gear. I've seen tens to hundreds of millivolts. I don't know if it's detrimental to performance of any gear, if it's at, say 400KHz, most audio gear isn't going to try to amplify that.

Well...bet you weren't expecting that!

 

[Vini darko]

This is an expesive rabbit hole to go down.
Ways to reduce emi and radio noise.....
Dedicated line to listening room (fully shielded)
High quality mains conditioners that work both ways.
Faraday cage around room.
All cables running braid or twisted pair and shielded.
All electronics using potted transformers and copper plate internal shielding.
I hope your rich

 

[Chrispy]

Wow, rampant electromagnetic radiation.....where does this stuff come from?

 

[March Audio]

For example, a class AB amplifier with a linear power supply such as Dayton APA150 doesn't have electromagnetic radiation associated with class D PWM frequencies and SMPS noise.

SMPS noise in house wires contribute to electromagnetic radiation from house wires. Good switch mode power supplies have common mode choke and a few other dirty electricity suppression tricks. But, I don't know which switch mode power supplies do their best to not pollute house wires. I will have to measure.

Other than SMPS noise and electromagnetic radiation at class D chipset PWM frequencies, I suspect there could be electromagnetic radiation around audio frequencies because power amplifiers have to deliver audio signals to speakers. Speaker cables can be shielded. I don't yet know what it would take to suppress electromagnetic radiation at audio frequencies.

Do you know good ways to minimize electromagnetic radiation from power amplifiers?

You again.

I had to take you off ignore to see your post.

I will just repeat what I said last time.

https://www.audiosciencereview.com/...radiation-from-class-d-amps.13011/post-388351

 

[March Audio]

Class D amplifiers do push a ton of switching noise back into the mains and back into all components connected to them. They also radiate switching noise from all their cables. Whether this is harmful or not, is something I've been wondering about. Not about human health, but about the added issues all this noise causes for interconnected components and whether it compromises performance. All the nice measurements you see are low pass filtered so the measuring equipment doesn't see the noise.

No they don't.

They have to pass both conducted and radiated emissions tests.

Secondly noisy mains is a given. Any product should be designed to cope with it.

Measurements are filtered to prevent issues with the range switching etc of the AP so that accurate results are obtained.

 

[March Audio]

One possible issue, I’ve wondered about, with class D is the possibility of intermod with a ’noisy’ DAC output, perhaps playing DSD for example; could this produce tones in the audio band? We know that the March Audio Purifi amplifiers have a low pass filter in the input circuit and this could be one reason why, apart from not demodulating any spurious RF.

It's a possibility although unlikely. However any amp, class d or or not should have appropriate input filtering. There is a thread here where someone reported noises with their class d amp (not one of ours). At my suggestion implemented an input filter and problem went away.

 

[Mnyb]

Wow, rampant electromagnetic radiation.....where does this stuff come from?

The sun for example

 

[BillH]

The sun for example

Don't forget incandescent light bulbs!
60 Hz AC over a ~1/2 inch filament radiating into free space.
I'm so glad we don't have CRTs any more....

Hmm - what about those LED light bulb power supplies?

 

[Mnyb]

This is a measurable non problem , if you use an am radio or “electro smog meter “ ( that thing exists ) you are going to pick up something ofcourse you are .

Can only second March Audio , I had a big Rotel power amp that demodulated radio once in a while .

 

[mhardy6647]

Wonder why not LC or LCLC topology is more common ?

Plenty common in the vacuum tube amplifier domain ('specially DIYers' efforts).
Good chokes can be 'spensive, though.

Note the outboard P/S (far right) on the (borrowed!) "Berman" 76-based preamp (see, e.g., https://www.diyaudio.com/forums/tubes-valves/273755-berman-76-preamplifier-advice-please.html) I've been playing with lately, e.g.

DSC_0121 (2) by Mark Hardy, on Flickr

 

[raindance]

No they don't.

They have to pass both conducted and radiated emissions tests.

Secondly noisy mains is a given. Any product should be designed to cope with it.

Measurements are filtered to prevent issues with the range switching etc of the AP so that accurate results are obtained.

Sorry, you're wrong. Many of these amps are kits and don't have to pass anything. And I'm not making this stuff up!

 

[raindance]

I can venture a guess or two why, but I am curious to hear from you since you worked on it. What is the problem with EMI/EMC in that situation? I'm especially curious about EMI, and also it pertains to this thread.

In that situation, EMI and EMC can interfere with navigation, communications and combat systems. Not very desirable with life safety on the line.

 

[Tatteredmidnight]

In that situation, EMI and EMC can interfere with navigation, communications and combat systems. Not very desirable with life safety on the line.

In my experience, most houses don’t have navigation or combat systems, ymmv of course.

 

[MakeMineVinyl]

A Class AB output stage will draw half wave rectified, nonlinear currents from the supply rails. The resulting magnetic fields can induce distortions into the input circuitry of the amplifier.

HUH?

 

[March Audio]

Sorry, you're wrong. Many of these amps are kits and don't have to pass anything. And I'm not making this stuff up!

No, I am correct. The Hypex and Purifi PSUs and modules are CE approved. So are ICE etc. Whilst of course its not impossible for a manufacturer to completely screw things up, the fundamental components are fine and tested. You are extrapolating and generalising. I have no doubt some badly designed and untested SMPS ar class D amp could cause problems, but this is true of any product.

BTW I used to represent Keysight products and used to sell RF spectrum analysers for looking at precisely this sort of thing (yes I know what a LISN is )