[Malvern] Electrical Intreference

[Steve Cashmore]

Hi Ian,

The symmetry of the filtering is a function of the particular filter 
circuit design.  Naturally it is important that it doesn't unnecessarily 
attenuate the side where the data communications equipment lies.  Thus 
it typically presents a high impedance (primarily inductive) at RF 
frequencies to the household side, and a low impedance on the other side 
to mop up whatever remains.  A smarter filter would present a high 
impedance at RF on both sides so as not to attenuate any power company 
signalling equipment.  To my knowledge the UK utilities aren't yet 
commonly using such signalling to domestic services.  Italy however has 
rolled out a large system to provide automatic remote meter reading 
using mains borne signalling.  I believe the new meters have built in 
filtering to prevent customer equipment from attenuating or interfering 
with the signalling system.

To answer your question more directly for your friend, there is still 
quite a strong RF attenuation in the 'wrong' direction even if the 
design is rather asymmetric.  Unless the filter is really pathetic I 
would be surprised if significant amounts of conducted RF made it 
through.  There could however be other poor building wiring and earthing 
practice that might partially bypass the filter.  Or equipment may be 
overly susceptible by reason of poor design.  There are well established 
legal emission and susceptibility levels with which all equipment sold 
within the EU must comply.  If you have an interference problem there 
are supposed to be channels through which you can apply for help, but I 
have no knowledge of how they work or how effective they are.  A little 
Googling should find the info.

It's not trivial to get a decent RF signal level on household mains 
wiring.  Until very recently the typical design of switch-mode power 
supplies in PCs, compact fluorescent bulbs etc used low RF impedance 
capacitors directly across the mains and from each leg to ground to 
reduce emitted switching noise.  This has the secondary effect of 
attenuating any existing RF data modulation.  The need to comply with 
current EU power factor and distortion legislation has helped alleviate 
this type of design practice.  But there are legions of stories in the 
USA of folks with remote control equipment that fails when certain other 
equipment is attached to the mains.  Wall wart type mains filters are 
available to try and reduce the attenuation caused by such equipment.

best regards,
-- 
Steve

Ian Pascoe wrote:
> Hi Steve
> 
> Well you kinda got the reasoning right!
> 
> My brother got strange interference on his TV and Hi-fi but at only certain
> times of the day.
> 
> A neighbour down the street who definately falls into the boys for toys
> bracket has recently had one of these multi room hifi's using mains cable
> for distribution installed.
> 
> He got himself an RF mains filter  and everything was hunky dory, but the
> people in the flat above started to experience the same thing, so he asked
> me and the rest is as they say history!
> 
> Mind you one other thought that comes to mind, that is does an RF filter
> work both ways?
> 
> So let's assume I have more money than sense and install one of these mains
> distributed systems, if I placed an in line filter between the meter and
> fuse board would it stop the RF going in both directions or only one?  Or
> put another way does it stop dead RF traffic in eithre direction once
> installed?
> 
> Cheers
> 
> Ian
> 
> -----Original Message-----
> From: malvern-bounces at mailman.lug.org.uk
> [mailto:malvern-bounces at mailman.lug.org.uk]On Behalf Of Steve Cashmore
> Sent: 28 January 2007 18:20
> To: Malvern at mailman.lug.org.uk
> Subject: Re: FW: [Malvern] Electrical Intreference
> 
> 
> Well the main purpose of any RF mains filtering before say a Hi-Fi
> amplifier is to reduce the effects of in-house equipment such as
> refrigerators etc from inducing clicks and pops into the audio chain.
> These are for mains conducted rather than radiated RF noise sources.
> Such noise is naturally diminished by distance so typically, but by no
> means always, such noise is greater from locally generated events.
> 
> One hopes that wireless induced noise would be naturally inhibited by
> good circuit design/bandwidth control and layout within the Hi-Fi
> equipment.  Sadly not always the case!
> 
> As to whether equipment SHOULD have such external filtering, then that
> is another question.  I would say only if you have a problem.  I
> understood your question was about preventing snooping on data
> deliberately induced onto mains wiring for communication purposes.
> 
> Common filtered extension sockets are mainly just transient suppressing
> devices designed to prevent damaging voltage spikes from destroying
> semiconductor devices in the protected device.  Most non-rural locations
> in the UK seem to enjoy a pretty good standard of clean electrical
> supply compared to say the USA, and improved PSU design practice have
> reduced the need for such devices IMHO.
> 
> that's enough rambling...
> --
> Steve
> 
> Ian Pascoe wrote:
>> Hi Steve
>>
>> OK, that all makes sense.
>>
>> Rather intrigued about the in line filtering before the distribution panel
>> in your home.
>>
>> If one of these were in place, would it be safe to presume that the amount
>> of RF captured by induction within the house is neglible or should things
>> like PCs audio systems etc still have their own RF filtering?
> 
> 
>> Ian
>>
>> -----Original Message-----
>> From: malvern-bounces at mailman.lug.org.uk
>> [mailto:malvern-bounces at mailman.lug.org.uk]On Behalf Of Steve Cashmore
>> Sent: 28 January 2007 14:23
>> To: Malvern at mailman.lug.org.uk
>> Subject: Re: [Malvern] Electrical Intreference
>>
>>
>> Hi Ian,
>>
>> Ian Pascoe wrote:
>>> For any sensitive electrical equipment it is generally recommended to
> have
>>> an RF filter somewhere on the mains feed to remove spikes that come in
>> over
>>> the mains power.
>>>
>>> If you are converting from AC to DC power will these spikes still be
> there
>>> after the conversion?
>>>
>>> If you transform the power down from 240v AC to say 12v AC does the
> spikes
>>> decrease in line with the reduction or do they stay the same?
>> Unfortunately fast rise-time spikes tend to capacitively couple their
>> energy directly to the output through components of the conversion
>> circuitry such as a transformer.  For that reason it's hard to predict
>> exactly what reduction there may be for a given circuit. It has much to
>> do with the physical construction of transformers and assembly of the
>> final product.
>>
>>> And lastly, I have read that if a neighbour uses the mains ring to
> provide
>>> data transfer for either audio or computer networks, that extra
>> information
>>> leaks back out into the mains distribution.  Would an RF filter remove it
>> or
>>> do you need something else?
>> Yes, an RF filter will remove the data modulation from external
>> radiation through the mains network.  Without such a filter, a reduced
>> signal may be present up to the transformer sub-station.  There are such
>> commercial domestic filters available although I can't point to a
>> specific example.  Some clip-on across the incoming house wiring making
>> no electrical connection, but have limited filtering, and others are
>> physically wired at the distribution board and provide a high degree of
>> isolation.
>>
>> Schaffner are a well known manufacturer of filtering devices in Europe.
>>   In the USA these RF filters are often available from whole house
>> remote control equipment suppliers.  Many of these systems use a low
>> frequency RF modulation on the house wiring for control purposes.
>>
>> best regards,