RFI : how does it affect mains circuitry of hifi equipment?

[F1eng]

Surely any designer, having discovered something which could adversely influence the performance of his design, in this case RFI, would make the necessary changes to the design to prevent this?
As a, now retired, designer it wouldn't occur to me not to have.

The whole idea that there are masses of hifi electronics for which such detail has not been dealt with seems quite ridiculous, in fact not credible. Unless a large number of hifi designers are not very good, which seems unlikely since the most sensible posts on this and most other threads come from people who have design and production experience of hifi products.
There are always cost constraints on designs though, so cheap kit may have parts missing, but surely it would be better to spend the money on a better product than buy something inadequate and "fix" it oneself with an aftermarket cable?

 

[JohnW]

Surely any designer, having discovered something which could adversely influence the performance of his design, in this case RFI, would make the necessary changes to the design to prevent this?
As a, now retired, designer it wouldn't occur to me not to have.

I suggest you request a copy of the Paul Miller article from Werner.

The HiFi industry is so amateurish that most do not understand the significance of the problem - not to mention have the equipment to perform the measurements.

One also has to strike a balance between filtering (LPF can and does effect the sound quality requiring high quality (= Expensive) components) - verse real life encounters of RF and not just under extreme "test" conditions.

It be a shame to sacrifice sound quality for the sake of a few extreme operating conditions - but where to draw the line.

Devices with transistor front ends tend to be affected worst then MOSFETs / JFets and Tubes which is one of the main reasons why I prefer the later devices.

 

[whatsnext]

It's a very british thing to criticise successful businesses, not that harping on about morals here is going to make any difference.

How are you on Drug dealers and distributors?

 

[stevec67]

Surely any designer, having discovered something which could adversely influence the performance of his design, in this case RFI, would make the necessary changes to the design to prevent this?

This is very much the crux of RFI and mains borne interference in general. It's a well known, extensively studied problem, and it's understood. This is not the wave particle duality/Higgs boson level of physics understanding, this is at the level of "how do you make a car engine that will go 10,000 miles between services and not wear out in 50k miles?" If you're en engineer working in this area, you know. Just because a bunch of accountants, teachers and food technologists on a hifi forum don't know how it works, it doesn't mean nobody does. I'd therefore venture that if it were something this simple then RA/Nordost/whoever would have been all over it by now, published the results and be making very solid claims about how it works. The fact that they aren't says that it isn't about RFI or mains borne interference.

In addition, if it were, why are there so few hifi items running on batteries? It's actually easier to design a component running from a battery (though I'll pass on valves!) as you don't need 240V in the case, you need 40-0-40 or similar, or sometimes just 12V.

There are always cost constraints on designs though, so cheap kit may have parts missing

Everything is built to a cost. The skill of the engineer is to cut costs where they make no difference. Instead of securing a case with 12 screws, use 4. Better still use 2 and secure the front edge by having it fit into a slot. Any fool can make a car to sell at £20k. It takes skill to make one that sells at £6k and still generates a profit, so take a bow Hyundai, Toyota and whoever else is churning out basic hatches that do a simple job. Back in hifi, I'll spend the money on metal film resistors where necessary but the one that limits the current to the power LED can be a crappy carbon film effort, saving 2p a unit. Same goes for boutique caps.

So there's your choices.
Choice 1: 50 years since the invention of the transistor, hifi engineers have NEVER thought of RFI and a bunch of guys on a hifi forum have hit on it, purely out of chance.
Choice 2: RFI is done and dusted, engineers know it's not a big deal and can be easily dealt with elsewhere for pennies. The enthusiasts on the forum are getting all excited thinking they have made a big discovery, when in fact they haven't. Except that they have started thinking about something new, new to them at least.

Which of these 2 is more likely?
Well, I know, it's pfm for a Nobel Prize, high fives all round!

 

[zarniwoop]

The plural of 'anecdote' is not 'data'.

I'm sorry but I'm going to steal that.

 

[formbypc]

The plural of 'anecdote' is not 'data'.

If there are sufficient anecdotes, they become 'significant anecdotal evidence', thereby forming the data for any study of them, surely?

 

[Steven Toy]

If there are sufficient anecdotes, they become 'significant anecdotal evidence', thereby forming the data for any study of them, surely?

They certainly become 'a body of evidence' that could allow a certain conclusion to be reached 'on the balance of probabilities,' which, as we know, just isn't enough for some even when the matter in hand is neither criminal in nature nor an issue of life and death.

 

[Blzebub]

They certainly become 'a body of evidence' that could allow a certain conclusion to be reached 'on the balance of probabilities,' which, as we know, just isn't enough for some even when the matter in hand is neither criminal in nature nor an issue of life and death.

No they don't. Scientific data is gathered via experiments designed to test a hypothesis.

[YOUTUBE]OL6-x0modwY[/YOUTUBE]

 

[formbypc]

No they don't. Scientific data is gathered via experiments designed to test a hypothesis.

If the hypothesis is based on what people hear, then surely the best experimental protocol is to gather large numbers of people and hence large amounts of data (perhaps anecdotally) on what they hear?

 

[Blzebub]

If the hypothesis is based on what people hear, then surely the best experimental protocol is to gather large numbers of people and hence large amounts of data (perhaps anecdotally) on what they hear?

I think you could perhaps do some research.

 

[Steven Toy]

No they don't. Scientific data is gathered via experiments designed to test a hypothesis.

[YOUTUBE]OL6-x0modwY[/YOUTUBE]

That's why I put single inverted commas around 'body of evidence' like I just did again.

 

[Blzebub]

That's why I put single inverted commas around 'body of evidence' like I just did again.

It's not "evidence", it's anecdote. The two are entirely different things.

 

[Arkless Electronics]

Interesting to see that in spite of many pages of foo about mains cables over the past week or so and speculation about RFI that there has been zero interest in my post #96!

 

[whatsnext]

If people are interested I'll do a post on "reference" section with full instructions and part numbers for removing RFI... (I may need a moderator to access that?)... but please folks.. read up on electronics before talking bollocks about it... and I mean that in the nicest possible way

Or you could see it as not many want to do anything practical themselves if they believe they have RFI. Maybe preferring to buy a solution themselves.

 

[Arkless Electronics]

Or you could see it as not many want to do anything practical themselves if they believe they have RFI. Maybe preferring to buy a solution themselves.

Maybe solutions that really work, DO follow the laws of physics, only cost about a fiver and can be carried out by anyone with a screwdriver in a few minutes just aren't foo approved enough!?

 

[Werner]

that there has been zero interest in my post #96!

Just do it then. Don't forget that a forum like this also acts as public library.

 

[zarniwoop]

If there are sufficient anecdotes, they become 'significant anecdotal evidence', thereby forming the data for any study of them, surely?

Significant anecdotal evidence may act as a trigger for a scientific study, but the anecdotes themselves can't form part of the data (unless it was a psychological study of how people can be influenced by purely anecdotal evidence, but that wasn't the context).

If the hypothesis is based on what people hear, then surely the best experimental protocol is to gather large numbers of people and hence large amounts of data (perhaps anecdotally) on what they hear?

The outcome is guaranteed to be inconclusive. Many would report hearing something and many would report not - no progress has been made.

They certainly become 'a body of evidence' that could allow a certain conclusion to be reached 'on the balance of probabilities,' which, as we know, just isn't enough for some even when the matter in hand is neither criminal in nature nor an issue of life and death.

You are missing the point. The "some" for whom anecdotal evidence is not enough include the scientists and engineers whose work ultimately results in the stack of boxes and wires you use to listen to music. That stuff doesn't get designed and built by people operating on anecdotes and a balance of probabilities.

 

[neil]

If people are interested I'll do a post on "reference" section with full instructions and part numbers for removing RFI... (I may need a moderator to access that?)... but please folks.. read up on electronics before talking bollocks about it... and I mean that in the nicest possible way

Sounds like an excellent idea. I'm sure one of the mods could "make it so".

I'm assuming you mean for RFI coming in on the mains feed here and not RFI that might come in directly into the equipment.

 

[misterdog]

Interesting to see that in spite of many pages of foo about mains cables over the past week or so and speculation about RFI that there has been zero interest in my post #96!

Lining the listening room with lead sheet ?

Seriously though I have just installed a new cart. to my TT. and bought an AV preamp with XLR inputs allowing me to use as a volume control.

The cable length from your Arkless 640 ) is around 6 Metres. and by necessity
(monoblocks and ESL's) crosses several mains cables, 1 sheilded 1 not.

On first listen I noticed a hum which had not been noticed before. Then I realised that I had not connected the shield from the 'phono' leads to the XLR's.
Connected them up- bingo, quiet as you like, sounding mighty fine.

EMI/RFI ? I have no idea, and I really don't care. Sheilding does work though.

With mains cables, shielding COULD also prevent any EMI/RFI being radiated OUT OF the cable and not into my delicate but sumptuous phono cables.

I have been meaning to try a copper lined plywood case for your phono stage for similar reasons - one day.

 

[Arkless Electronics]

Sounds like an excellent idea. I'm sure one of the mods could "make it so".

I'm assuming you mean for RFI coming in on the mains feed here and not RFI that might come in directly into the equipment.

For RFI coming in from the mains yes but also other sources of ingress. Little or no RFI will come directly into the equipment in most cases.