EMI and RFI _ a very basic question.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi !
sorry for the trivial question.
Considering only the components inside a electronic device i wonder which are the parts who actually generate a decent level of EMI and RFI.

For now this is my list of main suspects:

EMI > mains voltage transformers, ferrite chokes, coils (?)

RFI > smps, rectifying diodes (?)

Any other suspect to add ?

Another question ... is there a cheap instrument to measure the RFI emission ?
Thanks and regards, gino
 
Last edited:
Henry Ott electromagnetic compatibility
Join here...
EMC & Compliance UK 2016

EMI and RFI are the same it is all electro magnetic interference, RFI....
You are entering a BIG world of pain...
EMI and signal integrity are two sides of the same coin, EMI tends to refer to interference from the environment or noise that a unit puts into the environment, signal integrity is the effect of low level EMI on a PCB that affects the quality of signals... I would join EMCUK and have a look at Keith Armstrong's archive pages, plenty of good stuff to read...
Also links here regarding switching circuits such as SMPS, class D amplifiers (a modulated SMPS)...
http://www.diyaudio.com/forums/power-supplies/193705-switcher-emc-design.html

Some more relevant links in this lot...

http://www.diyaudio.com/forums/software-tools/177231-pcb-design-related-links.html

Spent 30+ years learning about EMC and how to combat it, the PCB layout is the first and most important area to combat EMC issues... So its not something that can be easily answered in a thread. What I will say is ALL electronics radiates EM crud it is how you control this that is important... SMPS's always get a bad press in audio circles, yet digital itself by the way it works is noisy and one of the biggest culprits is often the system clock and harmonics derived from it, also the memory switching will also add to what is termed simultaneous switching noise....
Any sharp rise time switching signal will be problematic if the layout and design does not take it into account... Often on here though many disregard EMC engineering side of things because they think it does not matter for DIY Audio, why I don't know because you will never achieve ultimate fidelity unless you consider all aspects of EMC.
The answer to you question is ALL create EMC learning to control it is the skill and often layout is absolutely critical in achieving this....
 
Member
Joined 2014
Paid Member
Of course audio doesn't quite have the EMC problems others have. A few noted one
- certain models of vauxhall/opel would shut off at level crossings
-HMS sheffield. they had to turn off the radar to use the sat comms, so didn't see the missile coming
- USS Forrestal. Ships radar causes a missile on a plane to arm and fire across the deck.
- US black Hawk helicopter where the army variant didn't have the EMC pack and could be knocked out the sky with a garage door opener!
- CH148 cyclone currently grounded with EMC issues.

Be glad your hifi can't kill you...
 
Henry Ott electromagnetic compatibility
Join here...
EMC & Compliance UK 2016
EMI and RFI are the same it is all electro magnetic interference, RFI....
You are entering a BIG world of pain...
EMI and signal integrity are two sides of the same coin, EMI tends to refer to interference from the environment or noise that a unit puts into the environment, signal integrity is the effect of low level EMI on a PCB that affects the quality of signals... I would join EMCUK and have a look at Keith Armstrong's archive pages, plenty of good stuff to read...
Also links here regarding switching circuits such as SMPS, class D amplifiers (a modulated SMPS)...
http://www.diyaudio.com/forums/power-supplies/193705-switcher-emc-design.html

Some more relevant links in this lot...

http://www.diyaudio.com/forums/software-tools/177231-pcb-design-related-links.html

Spent 30+ years learning about EMC and how to combat it, the PCB layout is the first and most important area to combat EMC issues... So its not something that can be easily answered in a thread. What I will say is ALL electronics radiates EM crud it is how you control this that is important... SMPS's always get a bad press in audio circles, yet digital itself by the way it works is noisy and one of the biggest culprits is often the system clock and harmonics derived from it, also the memory switching will also add to what is termed simultaneous switching noise....
Any sharp rise time switching signal will be problematic if the layout and design does not take it into account... Often on here though many disregard EMC engineering side of things because they think it does not matter for DIY Audio, why I don't know because you will never achieve ultimate fidelity unless you consider all aspects of EMC.
The answer to you question is ALL create EMC learning to control it is the skill and often layout is absolutely critical in achieving this....

Hi and wow ... thanks a lot for the huge quantity of information.
I will try to read something but i am slow to learn.
Actually my need i think it is much more basic. Very basic.
I am looking for some kind of buzzer that put close to a component buzzes more when the component emits RFI.
For instance for temperature there are those handy and cheap infrared guns.
Is there a RFI gun ?
I mean, in a smps a capacitor or a resistor for sure they do not emits RFI or much lower than other components.
I need to spot those parts and try to shield them.
This is my main goal.
I have no chance to understand about how to design a pcb with attention to RFI issues ... no chance at all.
In general i am not so naive to hope to be able to design anything.
At least some modifications and not on smd parts for sure.
I would be more in the modding than in the designing of things.
In the end the equivalent of a Geiger meter for RFI that the more the RFI the more the buzz of the measurement.
This is the immediate goal.
Thanks a lot again. gino
 
Go to head-fi, ask which of their portable amps pick up handphone signals.
You'll get lots of replies. Buy the cheapest.

Hi and thanks a lot for the kind advice
But a proper instrument ... would it be expensive ? As i said above a device like a Geiger meter for RFI.
I have to spot the sources of RFI on a pcb. And possibly with their intensity also to place some shields and evaluate the efficacy of these shields of course.
Thanks again, gino
 
Of course audio doesn't quite have the EMC problems others have.
A few noted one
- certain models of vauxhall/opel would shut off at level crossings
-HMS sheffield. they had to turn off the radar to use the sat comms, so didn't see the missile coming
- USS Forrestal. Ships radar causes a missile on a plane to arm and fire across the deck.
- US black Hawk helicopter where the army variant didn't have the EMC pack and could be knocked out the sky with a garage door opener!
- CH148 cyclone currently grounded with EMC issues.
Be glad your hifi can't kill you...

Hi and thanks a lot for the very interesting advice.
I hope my hifi will not kill me ... for this i avoid tubes for one.
However ... i have a bloody smps that spits out RFI and it is very badly shielded, actually not at all. It is completely exposed.
I just need something like a RFI sniffer maybe ?... that if i put it close to the
part i can understand if it emits so that i can put some shielding sheets on them and test again

22448978.jpg


i wonder if they are expensive.
I understand that smps are particularly prone to emit this kind of disturbs.
Is this true ?
Thanks a lot again, gino
 
Jim Brown, the go-to EMI/RFI expert has about 50 readable papers & Power Points:
Audio Systems Group, Inc. Publications

Hi and thanks a lot for the kind and valuable advice.
In this phase i just need some kind of probe, the put close to the emitting part starts to buzz or gives an indication of the RFI presence.
The idea is:
1) to spot the RFI emitting parts
2) try to shield them
3) test again to understand the benefits of the shielding
Thanks a lot again, gino
 
Use the radio in the medium-wave (AM) range, with it's built in ferrite antenna.
Try to tune in to the most sensitive spots in the frequency spectra.

Just turn on your device, close to your faulty smps and you will understand at once what I am talking about.
Can try also a scan through different frequncy ranges, try to tune through also the short wave bands. Though I find most useful MW.

Then move the unit around.

Try the diode bridges in your Linear! power supplies, of your pre-pre, pre or power amp..

Try the screen of your computer. The LCD screen of your laptop.
The huge TV set right in the middle of the hifi-console, so popular these days.

Try your amplified antenna cable descending from the attic.
The neon bulbs in the kitchen. The LED lights in your apartment. Even worse, the fluorescent bulbs of today, everywhere.

One could go on, forever..

Ps.: Forgot about the cordless phone bases, should be the first to depart with an elegant ballistic takeoff
 
Last edited:
Use the radio in the medium-wave (AM) range, with it's built in ferrite antenna.
Try to tune in to the most sensitive spots in the frequency spectra.
Just turn on your device, close to your faulty smps and you will understand at once what I am talking about.

Hi and thanks a lot again for the very helpful advice.
I would like just to precise that the smps is not really faulty ... it is just spitting RFI around because it has no shield at all.
I found it quite strange in an audio unit.
Even my pc has a shielded smps ... :confused: what the hell this people are doing ???

Can try also a scan through different frequncy ranges, try to tune through also the short wave bands. Though I find most useful MW.
Then move the unit around.

i have bought one cheap on ebay ... i am very curious to check.

Try the diode bridges in your Linear!

i have a specific question on this issue. Do this RFI affect more digital than analog ? because this is my understanding

power supplies, of your pre-pre, pre or power amp..
Try the screen of your computer. The LCD screen of your laptop.
The huge TV set right in the middle of the hifi-console, so popular these days.
Try your amplified antenna cable descending from the attic.
The neon bulbs in the kitchen. The LED lights in your apartment. Even worse, the fluorescent bulbs of today, everywhere.
One could go on, forever..
Ps.: Forgot about the cordless phone bases, should be the first to depart with an elegant ballistic takeoff

wow ... we are completely immersed in RFI then ? crazy ... :eek:
I am watching "Better call Saul" series ... with the lawyer obsessed by RFI and EMI and walking around wrapped in an aluminum sheet ... :D
However thanks a lot indeed again
I am buying an am radio on ebay right now.
I am very curious to test this methodology.
Thanks a lot again
Kind regards, gino
 
Last edited:
ginettot61 said:
Considering only the components inside a electronic device i wonder which are the parts who actually generate a decent level of EMI and RFI.
Are you asking about the components which generate EMI or those which radiate EMI?

Generation would include anything which switches and anything which oscillates. This includes diodes, followers (potentially) and anything else which might oscillate - whether intentionally or accidentally.

Radiation would include anything which is connected to the EMI generator, so could include almost any component or wire.

The idea is:
1) to spot the RFI emitting parts
2) try to shield them
3) test again to understand the benefits of the shielding
This assumes that the RF has been unavoidably generated (not always true), that it is only emitted by certain components (not always true) and that shielding is the only way to stop it getting out (not always true).
 
Ah, I forgot the most juicy situations..
Try what happens close to your SPDIF transmission channel, with big honky RCA-s and square mile area pigtail connections. Enjoy the music getting through, picked up by your radio!!
Try what happens in your XMOS usb receiver. Or along the USB cable..

All this do not count, I'm already hearing the resident expert opinion.. Ah yes, especially with the uV decision levels, at which level your high loop gain feedback circuitry is working..

Ciao, George
 
Are you asking about the components which generate EMI or those which radiate EMI?

Hi ! thanks for the reply. Is there a difference ? :confused:

Generation would include anything which switches and anything which oscillates.
This includes diodes, followers (potentially) and anything else which might oscillate - whether intentionally or accidentally.
Radiation would include anything which is connected to the EMI generator, so could include almost any component or wire.
This assumes that the RF has been unavoidably generated (not always true), that it is only emitted by certain components (not always true) and that shielding is the only way to stop it getting out (not always true)

Thanks a lot for the very helpful explanation.
If i understand rightly while there is little to do in order to avoid RFI generation (maybe some parts generate less ? so a parts selection can help ?) a good design of the circuit can mitigate the radiation ? maybe with short wires and so on ?
This is very interesting but much more than what i really need to try.
I will try now with an AM radio to spot the RFI emitters on the smps and then the idea is to cover with absorbing materials the parts and check again for the results.
I am quite sure that just a simple metallic lid could have tamed this RFI emission ... instead i see only a plastic partition between the smps and the circuits. Not the ultimate shield for sure
Thanks again, gino
 

Attachments

  • smps 1.jpg
    smps 1.jpg
    749.9 KB · Views: 137
  • barrier plastic.jpg
    barrier plastic.jpg
    639.1 KB · Views: 136
Last edited:
ginetto61 said:
If i understand rightly while there is little to do in order to avoid RFI generation
No, a lot can be done to avoid RF generation. Good circuits will already do this. For example, stopper resistors at valve grids and follower outputs.

a good design of the circuit can mitigate the radiation ? maybe with short wires and so on ?
Where RF generation is unavoidable (e.g. fast switching) then good design can reduce the radiation. Short wires may be part of this; small loops too, which may require longer wires!

I fear, Mr. ginetto61, that you are doing your usual thing of seeking simple recipe-like answers to complex questions.
 
No, a lot can be done to avoid RF generation. Good circuits will already do this. For example, stopper resistors at valve grids and follower outputs.

Hi ! what i meant it this is suppressing RF with design choices.
You cannot stop a follower from generating RF because it is its nature.
You told me. There are parts/components RF generators by nature.
The only way is taming them with adequate design.

Where RF generation is unavoidable (e.g. fast switching) then good design can reduce the radiation. Short wires may be part of this; small loops too, which may require longer wires!
I fear, Mr. ginetto61, that you are doing your usual thing of seeking simple recipe-like answers to complex questions

Yes i agree. I will play a little with this AM radio just for curiosity.
Maybe i could also try some passive shielding on the most buzzing parts.
Unfortunately the cheapest shield is copper tape that is conductive.
I do not want to short circuit anything.
As i understand that pulling out the smps from the box tames completely the noise this could be another option.
Again i do not understand why not an external adapter like so many other units. That would have solved the issue easily.
Thanks a lot again, gino
 
Your EMC starter...
http://www.hottconsultants.com/pdf_files/dipoles-1.pdf
http://www.hottconsultants.com/pdf_files/dipoles-2.PDF
http://www.hottconsultants.com/pdf_files/dipoles-3.PDF

Add Ralph Morrison to your list of required reading...
Amazon.co.uk: Ralph Morrison: Books, Biogs, Audiobooks, Discussions

The fields of electronics is a good starter....

I have as I said previously been learning about EMC from a practical (and academic)viewpoint for 30+ years, don't expect enlightenment overnight... The practical side is doing layouts and EVERY layout I do being tested quite often to the extreme for EMC, a lot of the stuff I play with has far tighter requirements than consumer products...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.