Would it hurt if you did not use your.....

[Elso Kwak]

RF Interference

Hi,
I once was the proud owner of a Mark Levinson JC1-AC moving coil cartridge preamplifier.
One day I heard mumbling voices behind my loudspeakers.
I went out to see if there was someone there behind the wall. Nobody!
I also had problems with the on/off switching of the refrigerator giving loud bangs in the loudspeakers.
So I installed a lowpass-filter on the inputs of the pre-pre consisting of a cap to ground few turns on a ferrite bead and again small cap to ground. This helped for the bangs.
But to my greatest surprise soundquality IMPROVED!
The voices were gone too, making me feel comfortable not having to call for psychiatric help
So attention to RF pays off. Real grounding to a pipe drilled into the earth may help too.

 

[Steve Eddy]

Re: Steve

I haven't had chance to dig out the plots yet, as my other half was not well at the weekend, which left me somewhat busier than expected.

Sorry to hear that. I hope she's doing better.

W.r.t. the earth think RF, think shielding, think non-linearity, think impedance beyond DC.

I have. And I still don't see what relevance an actual earth ground has in this respect.

The ability of an audio system to maintain linearity in the presence of RF is a difficult thing to achieve, particularly in high gain, low noise app's (e.g. MC phono).

Well, unless the RF is causing instability (i.e. oscillation), the system typically maintains linearity within the audio band. The problem is that EMI outside the bandwidth of the amplifying device can result in intermodulation products down in the audio band. But that's beside the point.

It's almost impossible to design a circuit to reject common mode RF signals that performs optimally at AF, hence shielding is the only solution that can prevent them getting there in the first place.

The only solution? You mean filtering can't work? Hmmm...

Mains earth can have a high RF impedance, depending upon how it is supplied, hence a good local earth (which may have higher impedance at DC than the mains earth) can be a massive benefit.

You seem to be operating from the assumption that shielding works simply by providing a low impedance path to earth ground.

Shielding works in two distinct ways. Reflection loss and absorption loss. Both of which are a function of the material used for the shielding, not whether the shielding is tied to earth ground.

Reflection loss works by way of creating an impedance mismatch between the impinging EM wave and the shield. The greater the mismatch (i.e. the reflection coefficient), the greater the reflection loss.

A parallel to this comes right out of transmission line theroy. If you recall, the power delivered to the load depends on the impedance match between the load and the characteristic impedance of the line.

If the load impedance is higher or lower than the line impedance, a certain amount of the signal will be reflected back to the source, the result being less power delivered to the load.

Same thing with reflection loss in shielding.

Absorption loss is largely due to eddy currents within the shield and just like skin effect in wire (which is also the result of eddy currents), is determined by the thickness of the shield.

In absorption loss, the energy of the impinging wave goes into producing the eddy currents which produce fields opposing and subsequently canceling the impinging fields.

So how exactly does an actual earth ground fit into this picture?

se

 

[Steve Eddy]

Re: Waterpipe Grounding

The waterpipe grounding is strictly FORBIDDEN in our country.
Just realise in the event of a malfuction the waterpipe carries lethal 220V!

The reason isn't because the water pipe would carry 220 volts (if the hot lead were to contact the pipe, the low resistance of the pipe would result in a very low voltage on the pipe itself). The reason is that if you happen to come into contact with the hot lead and touch a water pipe (or far more likely, a faucet), then you become the return path for that 220 volts.

You can touch the hot lead all you want and be perfectly safe as long as there's no low impedance return path between you and the neutral line (or earth ground seeing as the earth ground is tied to the neutral line).

An Uncle of mine used to show off doing this when I was a kid. To someone who didn't know the first thing about electronics at the time, it seemed pretty amazing.

se

 

[Elso Kwak]

Teaching

Hi,
I am a fully qualified teacher for.......chemistry in Highschool.
Old fashioned theory about teaching says you have to explain everything ad nauseum. If they don't understand it, you did not explain it properly.
Current theory says the teacher only provides the means and some clues for selfstudy, and organise the process. Let the pupil try and fall on his/her face but he/she will learn in the process!
Nelson Pass also mentioned "The willingness to experiment" words that keep ringing in my head.
You learn much more by designing or try to design something yourself instead of aping some recipe.
In view of all this I consider Fred a great teacher giving so many great hints but not complete "solutions". If he is tired of "force feeding" forummembers, I admit the same feeling sometimes creeps on me too, now and then.
Elso/not a good teacher

 

[Elso Kwak]

Re: Re: Waterpipe Grounding

The reason isn't because the water pipe would carry 220 volts (if the hot lead were to contact the pipe, the low resistance of the pipe would result in a very low voltage on the pipe itself). The reason is that if you happen to come into contact with the hot lead and touch a water pipe (or far more likely, a faucet), then you become the return path for that 220 volts.

You can touch the hot lead all you want and be perfectly safe as long as there's no low impedance return path between you and the neutral line (or earth ground seeing as the earth ground is tied to the neutral line).

An Uncle of mine used to show off doing this when I was a kid. To someone who didn't know the first thing about electronics at the time, it seemed pretty amazing.

se

Hi Steve,
if the waterpipe is NOT grounded which is current practice over here as they are using plastic piping outside the house, the waterpipe carries 220V in the event of a malfunction. Our electricity distribution system has neutral connected to earth so your body carries 220V as your feet are connected to ground and you open the hose to wash your hands. Do I miss something???

 

[Steve Eddy]

Re: RF Interference

I once was the proud owner of a Mark Levinson JC1-AC moving coil cartridge preamplifier.
One day I heard mumbling voices behind my loudspeakers.
I went out to see if there was someone there behind the wall. Nobody!
I also had problems with the on/off switching of the refrigerator giving loud bangs in the loudspeakers.
So I installed a lowpass-filter on the inputs of the pre-pre consisting of a cap to ground few turns on a ferrite bead and again small cap to ground. This helped for the bangs.
But to my greatest surprise soundquality IMPROVED!
The voices were gone too, making me feel comfortable not having to call for psychiatric help

Actually, that was just me playing a little joke on you. You see, I'd hidden a couple of walkie-talkies behind your curtains and well... I'm sorry I put you through all that trouble, but it was rather funny as you started questioning your sanity.

And you thought that filter actually worked. Ha! Filters don't work. Only shielding works.

So attention to RF pays off. Real grounding to a pipe drilled into the earth may help too.

Yes, attention to RF can pay off. And I hope nothing I said implied otherwise. As for the latter, I'm still waiting to hear a cogent argument in that regard.

se

 

[Elso Kwak]

Re: Re: RF Interference

Actually, that was just me playing a little joke on you. You see, I'd hidden a couple of walkie-talkies behind your curtains and well... I'm sorry I put you through all that trouble, but it was rather funny as you started questioning your sanity.

And you thought that filter actually worked. Ha! Filters don't work. Only shielding works.



Yes, attention to RF can pay off. And I hope nothing I said implied otherwise. As for the latter, I'm still waiting to hear a cogent argument in that regard.

se

 

[Steve Eddy]

Re: Re: Re: Waterpipe Grounding

if the waterpipe is NOT grounded which is current practice over here as they are using plastic piping outside the house, the waterpipe carries 220V in the event of a malfunction. Our electricity distribution system has neutral connected to earth so your body carries 220V as your feet are connected to ground and you open the hose to wash your hands. Do I miss something???

Ok, you originally said that the water pipe is forbidden from being connected to ground because in the event of a failure, the water pipe would carry a lethal 220 volts.

I'm assuming by "failure" you mean the hot lead somehow coming into contact with the water pipe.

If this is what you meant (and please correct me if I'm wrong), then the water pipe wouldn't carry 220 volts in such an event. The hot lead contacting the water pipe would effectively be a short circuit. And what's the voltage drop across a short circuit? Zero.

Of course that's ideally. The water pipe will have a finite resistance so there would have to be some amount of voltage across it. But by the same token, the mains cannot supply an infinite amount of current. So any voltage at the water pipe would have to be less than 220 volts.

So the real reason tying the water pipe to ground is so that you don't end up being the return path between the hot lead and ground (which again is tied to neutral at the service panel).

se

 

[Steve Eddy]

Re: Re: Re: RF Interference

I'll see your and raise you a .

Now you'll have to figure out if I'm holding a straight flush or just a pair of deuces.

se

 

[ALW]

Well, unless the RF is causing instability (i.e. oscillation), the system typically maintains linearity within the audio band. The problem is that EMI outside the bandwidth of the amplifying device can result in intermodulation products down in the audio band. But that's beside the point.

But that's the point - it does affect linearity within the audio band. The intermod issue is hardly beside the point if it's audible.

The only solution? You mean filtering can't work? Hmmm...

Filtering works at reducing the susceptibility, it also works at reducing the music. It's hard, I'd go as far as to say impossible, to provide a filtering solution sufficient to remove all RF related issues and not have a sonically detrimental side effect.

It's a source of frustration, but you only become aware of the problem if you use your ears and not your test kit.

So how exactly does an actual earth ground fit into this picture

A shield that is unconnected to earth is also known as an antenna It can increase susceptibility of a system, since it acts to capacitively couple signal into the system.

You can connect the shielding material to the 0V of your system, which may have no relation to actual earth, but the RF that appears there can then get into the 0V of that system and wreak havoc, since it is, in effect, floating. The inability of the system to then deal with those signals linearly, and the resultant non-linearities are clearly audible.

Elso,

"The willingness to experiment"

I like that - I'd like to add

"The willingness to listen" too, something that seems sadly lacking amongst many supposedly knowledgable engineers.

As a typical engineer, with a wide RF and analogue background I have an overwhelming desire to measure what I'm hearing. It is true though that this is not always possible using the readily accepted criteria. I'm not saying it's not possible, but sometimes the ears bring quicker results. The filtering point above is a classic case in point - try filtering havily at all system inputs (and that by definition includes the actual outputs too) and compare with a well designed system that filters much less.

Theory tells me it must be better filtered, practice and my ears tell me otherwise.

Andy.

 

[Elso Kwak]

Re: Re: Re: Re: Waterpipe Grounding

Ok, you originally said that the water pipe is forbidden from being connected to ground because in the event of a failure, the water pipe would carry a lethal 220 volts.

I'm assuming by "failure" you mean the hot lead somehow coming into contact with the water pipe.

If this is what you meant (and please correct me if I'm wrong), then the water pipe wouldn't carry 220 volts in such an event. The hot lead contacting the water pipe would effectively be a short circuit. And what's the voltage drop across a short circuit? Zero.

Of course that's ideally. The water pipe will have a finite resistance so there would have to be some amount of voltage across it. But by the same token, the mains cannot supply an infinite amount of current. So any voltage at the water pipe would have to be less than 220 volts.

So the real reason tying the water pipe to ground is so that you don't end up being the return path between the hot lead and ground (which again is tied to neutral at the service panel).

se

Hi Steve,
Sorry for the misunderstanding...
Just to sum up the practice here in Holland:
Safetyground i.e. the third prong of the mainsplug is connected to real ground here. That is a bar of conducting metal is drilled into the ground. Then the resistance to ground is measured and if it is not up to standard a second bar is driven into the ground.
All new homes also must have a saftey earth leakage switch that switches off the whole house if a few mA's find there way to real ground instead of to neutral. Please note that safety ground is NOT connected to neutral in the switchbox!
So if I ground my stereo equipment I ground to the safety ground.
The audible effect of this is very subtle but I don't like the trembling feeling when I touch a metal part of my equipment.

I noticed some improvement with my Sony CDP when I disconnected the case from the circuitground and connected the case/chassis to the safety ground. Now I have a all plastic Philips CDP (Well I could safety-ground the hood)

 

[till]

in theorie we have real earth grounds here for every house ................................................. in reality often null and earth are connectet were the cable comes into house - thats all.

 

[Elso Kwak]

Rf issues

But that's the point - it does affect linearity within the audio band. The intermod issue is hardly beside the point if it's audible.



Filtering works at reducing the susceptibility, it also works at reducing the music. It's hard, I'd go as far as to say impossible, to provide a filtering solution sufficient to remove all RF related issues and not have a sonically detrimental side effect.

It's a source of frustration, but you only become aware of the problem if you use your ears and not your test kit.



A shield that is unconnected to earth is also known as an antenna It can increase susceptibility of a system, since it acts to capacitively couple signal into the system.

You can connect the shielding material to the 0V of your system, which may have no relation to actual earth, but the RF that appears there can then get into the 0V of that system and wreak havoc, since it is, in effect, floating. The inability of the system to then deal with those signals linearly, and the resultant non-linearities are clearly audible.

Elso,

"The willingness to experiment"

I like that - I'd like to add

"The willingness to listen" too, something that seems sadly lacking amongst many supposedly knowledgable engineers.

As a typical engineer, with a wide RF and analogue background I have an overwhelming desire to measure what I'm hearing. It is true though that this is not always possible using the readily accepted criteria. I'm not saying it's not possible, but sometimes the ears bring quicker results. The filtering point above is a classic case in point - try filtering havily at all system inputs (and that by definition includes the actual outputs too) and compare with a well designed system that filters much less.

Theory tells me it must be better filtered, practice and my ears tell me otherwise.

Andy.

Hi Andy,
When an engineer like Fred, Jocko or you speaks I do pay attention and listen very carefully...
I don't have the experience that RF filtering reduces the music. My little example with Radio Iriwan illustrates that the music is enhanced. I agree there are other ways of reducing RF-susceptibility like reducing bandwith, shielding etc. It was Fred who pointed at a nice article on the subject by Analog Devices. "RFI can cause rectification in sensitive analog circuits. Outputdrivers can be jammed too." (page 50 of Grounding in High Speed Systems, Walt Kester & James Bryant.

As a sidenote I could not reduce the bandwith of the circuit of the JC1-AC as the circuit was potted.

 

[ALW]

Elso

I don't have the experience that RF filtering reduces the music

You are of course going to find that a system that is not suffering RFI (compared to one that is) will sound better.

My point is that if you can eliminate the filtering and get rid of the RF, it will sound better.

RFI is an ever increasing problem to audio, in an ever noisier world and certain topologies are more susceptible. A bipolar, low noise, non-degenerated stage will be orders of magnitude more susceptible than a lower gm FET input stage, or a degenerated bipolar one.

Unfortunately the application often dictates the topology, rather than the other way around.

It's always better, from an engineering perspective to stop RF getting in in the first place, rather than let it in, and then filter it.

In non-audio app's both methods should be applied.

Andy.

 

[Elso Kwak]

in theorie we have real earth grounds here for every house ................................................. in reality often null and earth are connectet were the cable comes into house - thats all.

Hi Till,
That situation would trip the Ground Fault Switch or am I wrong?

 

[Elso Kwak]

Re: Elso

You are of course going to find that a system that is not suffering RFI (compared to one that is) will sound better.

My point is that if you can eliminate the filtering and get rid of the RF, it will sound better.

RFI is an ever increasing problem to audio, in an ever noisier world and certain topologies are more susceptible. A bipolar, low noise, non-degenerated stage will be orders of magnitude more susceptible than a lower gm FET input stage, or a degenerated bipolar one.

Unfortunately the application often dictates the topology, rather than the other way around.

It's always better, from an engineering perspective to stop RF getting in in the first place, rather than let it in, and then filter it.

In non-audio app's both methods should be applied.

Andy.

Hi Andy,
RFI can not be avoided in this world full of TV & radio transmittors, mobile phones, mobile phone relay stations, garagedoor openers, wireless keyboards and mouses, babyphones and who knows what other gadgets......
In "civilized world" my uncle concluded, you are never more than 100km from a powerfull FM-transmittor because of the excellent coverage. One of the "blessings" living in a civilized world......

In my example of the moving coil cartidge it is very difficult to omit the RF to creep in, as the cable is not shielded all the way to the cartridge and the cartridge is also not completely shielded.

 

[grataku]

There are a number of reasons why the gnd should not be connected to the waterpipe, among others, the watersystem cannot be assumed to be a perfect gnd, and in most cases it isn't, second it's easy to run into the problems with pipe corrosion due to the nonzero gnd potentials that would be a very bad thing environmentally and economically. As far as shock risk, I only have seen this problem in the US where they don't use superfast differential swithes that would really prevent any problems of the kind. Two coils, one on the hot one on the neutral, if the two currents are not the same (even by a few uA) it means the current is getting discharged through other paths and the switch trips in a few ms.
I think it's pretty normal for the powerstation to connect the star of the three phase to gnd. It's too dangerous to leave it floating i the event of a transformer malfunction. Since in every house you have a live and a neutral, the neutral comes from gnd at the transformer secondary star connected while the high V primary is connected in a triangle arrangment at the substation, pardon my lack of technical english terms.


RF in audio amps? Can someone explain to me why their presence should be ruled out a priori? Even in non patological cases. There obviously are transients all over the place that under normal circumstances just decay away. Since on the average are not there it doesn't mean they don't negatively impact the sound.

 

[Steve Eddy]

But that's the point - it does affect linearity within the audio band. The intermod issue is hardly beside the point if it's audible.

Intermod products in the audio band from RF above the audio band does not constitute a nonlinarity in the audio band. It constitutes a nonlinearity above the audio band.

And I said it's beside the point because the point of this discussion is earth ground.

Filtering works at reducing the susceptibility, it also works at reducing the music. It's hard, I'd go as far as to say impossible, to provide a filtering solution sufficient to remove all RF related issues and not have a sonically detrimental side effect.

Well, seeing as no electronic circuit has an infinite bandwidth, every electronic circuit is in effect a filter. So I guess it just depends on where the cut-off is.

A shield that is unconnected to earth is also known as an antenna It can increase susceptibility of a system, since it acts to capacitively couple signal into the system.

Excuse me?

We're talking about a proper shield here, yes? One which encloses the circuit such as an equipment chassis or the shielding around an interconnect, yes?

If so, then no, it is not also known as an antenna (whether it's tied to earth ground or not) and it does not act to capacitively couple EMI into the system. Just the opposite.

An enclosing shield is also known as a Faraday cage, which in fact acts to prohibit capacitive coupling into the circuit.

When EMI impinges on the shield, the electrons in the shield alternately move toward or away the source of interference, cancelling the E field at that point of the shield. When the E field is positive, electrons move toward it. When it's negative, they move away from it.

The net effect is that the E field within the enclosure remains 0.

You can connect the shielding material to the 0V of your system, which may have no relation to actual earth, but the RF that appears there can then get into the 0V of that system and wreak havoc, since it is, in effect, floating. The inability of the system to then deal with those signals linearly, and the resultant non-linearities are clearly audible.

If we're talking about an enclosed shield, i.e. a Faraday cage, then how exactly does it "get into" the 0V of the system?

An amplifier amplifies the voltage potential it sees between its input and its ground reference. If the ground reference is tied to the shield, whether the shield fully encloses the circuit or not, how do you get any potential difference between ground and any other part of the circuit?

se

 

[SY]

Unless you're sitting right next to a megawatt LORAN transmitter, RF is pretty easy to filter out and shield against without aural penalty, as long as you're talking about aural penalty and not "aural penalty." As a teen, I got more experience with this than I'd care for, trying to combine my twin hobbies of ham radio and DIY audio (and with a 1 MHz clear-channel AM station just a kilometer away). Blessed grid stoppers and mica caps! Without those, the banging from the speakers every time I'd key the CW transmitter would deafen you.

In my last three locations, RF just hasn't been an issue at all. But, to be honest, my old California Kilowatt so-called "linear" amplifier is a few decades behind me, and I live in a metro area of under 200,000 people.

The earth ground thing is... odd. Important if you're building antennas, though.

 

[Steve Eddy]

Re: Re: Re: Re: Re: Waterpipe Grounding

Sorry for the misunderstanding...
Just to sum up the practice here in Holland:
Safetyground i.e. the third prong of the mainsplug is connected to real ground here. That is a bar of conducting metal is drilled into the ground. Then the resistance to ground is measured and if it is not up to standard a second bar is driven into the ground.
All new homes also must have a saftey earth leakage switch that switches off the whole house if a few mA's find there way to real ground instead of to neutral. Please note that safety ground is NOT connected to neutral in the switchbox!

You mean your safety ground isn't tied to the center tap of the distribution transformer? It's just stuck in the ground? That doesn't seem right.

So if I ground my stereo equipment I ground to the safety ground.
The audible effect of this is very subtle but I don't like the trembling feeling when I touch a metal part of my equipment.

Hehehe. Hey, it's a cheap buzz.

I noticed some improvement with my Sony CDP when I disconnected the case from the circuitground and connected the case/chassis to the safety ground. Now I have a all plastic Philips CDP (Well I could safety-ground the hood)

Well disconnecting the circuit ground from the case can eliminate certain ground loops and interchassis currents.

se