The shield has to be connected at both ends for RF shielding, if you don't connect it at both ends for RF frequencies it becomes an antenna and wont shield any RF..... It also has to be connected where it enters the component.
You can have balanced (differential) routing and cabling for single ended signals but this is best done from source to destination, which means the PCBs have to be laid out to cater for this to get full advantage otherwise you can end up mixing balanced and single ended distribution techniques for the signal in question ... Best keep things simple otherwise you may end up with more problems, so use coax for single ended, low impedance shield (ie full 100% copper braid, low resistance as well is critical, helps avoid ground loops).
You can have balanced (differential) routing and cabling for single ended signals but this is best done from source to destination, which means the PCBs have to be laid out to cater for this to get full advantage otherwise you can end up mixing balanced and single ended distribution techniques for the signal in question ... Best keep things simple otherwise you may end up with more problems, so use coax for single ended, low impedance shield (ie full 100% copper braid, low resistance as well is critical, helps avoid ground loops).
Correct me if I am wrong, but any setup that is not fully dual-mono has at least one ground loop in it: You have two ground paths in the interconnect, one for the left channel one for the right. Two ground paths = ground loop?
Studio guys often advise connect the shield at one end only, assuming balanced interconnections.
Studio guys often advise connect the shield at one end only, assuming balanced interconnections.
With so many opinions it's no wonder folk get confused. Some say it's bad practice to use coax because it mixes any noise picked up by the screen with the signal which uses that screen as a return. Most cable manufacturers sell pseudo balanced single ended interconnects these days, they seem to work ok but as usual there is no data to show any advantage. I would like to go fully balanced but my kit is all single ended.
I made 5 meter interconnects earlier today using twin core screened mic cable. I connected the screen to signal ground at the source end (DCB1) and left it floating at the other. These replaced a set i made from coax, no noise is apparent but they appear to have slightly improved SQ, the sound is noticeably a bit more open.
I have to put this down to the mic cable having lower impedance because the alternatives - pseudo balanced connection - silver coated copper conductors - both signal and return using the same diameter wire.....
We all know that's just snake oil, right!
I made 5 meter interconnects earlier today using twin core screened mic cable. I connected the screen to signal ground at the source end (DCB1) and left it floating at the other. These replaced a set i made from coax, no noise is apparent but they appear to have slightly improved SQ, the sound is noticeably a bit more open.
I have to put this down to the mic cable having lower impedance because the alternatives - pseudo balanced connection - silver coated copper conductors - both signal and return using the same diameter wire.....
We all know that's just snake oil, right!
Last edited:
A word from German engineers:
"Cable screen not connected.
This screen will not prevent any interference, because the charge on
the screen, produced by interference, will remain and will effect the
central signal line. Also, the current induced by interference in the line
will flow through the sink, effecting the signal.
Cable screen grounded on source side only.
This screen will only prevent interference at low frequency signals. For
electromagnetic interference, where the wavelength is short compared
to the length of the cable, the screening efficiency is poor.
Cable screen grounded on either side.
It is effective for all kinds of interference. Any current induced in the
screen by magnetic interference will flow to ground. The inner of the
cable is not affected. Only the voltage drop on the screen will affect the
signal in the screen."
"Interference and Screening" paper, Radio Training Centre, Deutsche Welle Akademie
Of course, this works for balanced signal lines and properly grounded equipment. In home applications, chances are that a shield connected to both ends will create buzzing ground loops. In that case, leave the shield connected at source end.
"Cable screen not connected.
This screen will not prevent any interference, because the charge on
the screen, produced by interference, will remain and will effect the
central signal line. Also, the current induced by interference in the line
will flow through the sink, effecting the signal.
Cable screen grounded on source side only.
This screen will only prevent interference at low frequency signals. For
electromagnetic interference, where the wavelength is short compared
to the length of the cable, the screening efficiency is poor.
Cable screen grounded on either side.
It is effective for all kinds of interference. Any current induced in the
screen by magnetic interference will flow to ground. The inner of the
cable is not affected. Only the voltage drop on the screen will affect the
signal in the screen."
"Interference and Screening" paper, Radio Training Centre, Deutsche Welle Akademie
Of course, this works for balanced signal lines and properly grounded equipment. In home applications, chances are that a shield connected to both ends will create buzzing ground loops. In that case, leave the shield connected at source end.
Last edited:
A simple evidence is that an outer shield used as a return line will pick up interference. In a balanced cable, return line is protected by the shield.
This supports not using coax in single ended systems. I don't have the option of going fully balanced so is pseudo balanced a worthwhile compromise? It clearly has not added any noise but i'm not convinced it would help to remove any either.
One advantage is that the mic cable is VERY cheap.
I use balanced cables in my single-ended system, with shield connected at source end.
Switching from coax to balanced cable might be beneficial, provided that it is an overall upgrade in cable quality, because I have heard some really fantastic coax intecrconnects which did not seem to suffer from any "nakedness". Cheap cable means certain compromises on quality: for instance, cheaper cables use spiral instead of braided shields which offer less interference screening. They also tend to use copper instead of iron shield which is less efficient.
Switching from coax to balanced cable might be beneficial, provided that it is an overall upgrade in cable quality, because I have heard some really fantastic coax intecrconnects which did not seem to suffer from any "nakedness". Cheap cable means certain compromises on quality: for instance, cheaper cables use spiral instead of braided shields which offer less interference screening. They also tend to use copper instead of iron shield which is less efficient.
The stuff i used has silver coated high purity copper conductors twisted alongside cotton string shielded with copper braided screen. http://www.vdctrading.com/content/VanDamme-Cable-Specifications/VanDamme-Microphone.pdf
I paid around £1.20 per meter off ebay and it sounds just fine to me.
I was under the assumption that copper made a better screen, i think one of the Rane articles stated this too but i could be mistaken.
I paid around £1.20 per meter off ebay and it sounds just fine to me.
I was under the assumption that copper made a better screen, i think one of the Rane articles stated this too but i could be mistaken.
Because the two cables are in close proximity to each other for the entire part, the loop area is almost zero.
You can have as many signal channel cables as you like. If they all start at the same place and end at the same place and follow the same path there is no loop area.
note that in the video they measured the voltage between the separated and insulated core and the metal edging of the bench. The core and the metal edging were separated by a large gap.
This is an excellent example of why we should look at minimising LOOP AREA.
If they had measured the difference in voltage between the two cores (which may have been twisted inside that screen), then the numbers would have been MUCH smaller. Even when the load was still near infinity.
With the small loop area of the two cores, there would be little improvement of the measured voltage difference by grounding the screen. But adding the screen ground would attenuate a bit.
Now if they had loaded the cores with a typical load impedance, the measured voltage would be lower still.
We have a range of tools available to us to minimise interference.
Distance between receiver and transmitter.
Loop area between the conductors.
Twisting of the conductors.
Load and/or source impedance.
We should use any and if necessary all of these tools to make the interference voltages low enough to allow our circuits to work adequately.
We cannot get perfection, just "good enough".
This is an excellent example of why we should look at minimising LOOP AREA.
If they had measured the difference in voltage between the two cores (which may have been twisted inside that screen), then the numbers would have been MUCH smaller. Even when the load was still near infinity.
With the small loop area of the two cores, there would be little improvement of the measured voltage difference by grounding the screen. But adding the screen ground would attenuate a bit.
Now if they had loaded the cores with a typical load impedance, the measured voltage would be lower still.
We have a range of tools available to us to minimise interference.
Distance between receiver and transmitter.
Loop area between the conductors.
Twisting of the conductors.
Load and/or source impedance.
We should use any and if necessary all of these tools to make the interference voltages low enough to allow our circuits to work adequately.
We cannot get perfection, just "good enough".
Last edited:
On another note, that video was a precursor to the Health Benefits, or otherwise, of grounding humans.
They appear to have deliberately set up that experiment to show how bad the received field was compared to a grounded reference.
I think there was a sales pitch in there that was targeting an uninformed audience.
We are better than that.
We can inform ourselves, we analyse what we see and what we are told. We "sort the wheat from the chaff". We allow ourselves to become "informed".
That is what we as DIYers should be doing BEFORE we make decisions.
They appear to have deliberately set up that experiment to show how bad the received field was compared to a grounded reference.
I think there was a sales pitch in there that was targeting an uninformed audience.
We are better than that.
We can inform ourselves, we analyse what we see and what we are told. We "sort the wheat from the chaff". We allow ourselves to become "informed".
That is what we as DIYers should be doing BEFORE we make decisions.
Well said Andrew! It is foolish to waste precious time trying to obtain perfection, the law of diminishing returns will always come into play. Good enough is fine by me.
Hey, walking barefoot at the beach is awesome for your health.
I could point you in the direction of several beaches which could seriously challenge that statement
Right....
You have no RF screening, silver plated copper is not going to make much difference and the screen connected at one end only is an antenna...
All serious stuff recommends screens connected for RF at both ends (that is a RF connection), this is standard EMC engineering.
Henry Ott, Rane, Tony Waldren etc. etc. All have numerous information on EMC engineering witch is what this is.
RF connection at both ends, high impedance for the low frequency so no buzz... RF is a big problem and lack of proper screening is allowing the RF in making the sound generally 'bigger' the illusion of an improvement in the SQ...
There is balanced (differential) routing cables and there is balanced (differential) signals...
Balanced signal - balanced wiring best
SE signal - balanced wiring very good if done correctly.
SE signal - SE wiring can be excellent (coax etc) or bad, but is the most prone to noise pick up so requires the most thought on shielding....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.