Jump to content
numlog

shielding for AC (power, audio)?

Rate this topic

Recommended Posts

Is it generally a good or bad idea to add passive shielding (no ground) to AC cables?

 

I recall reading adding shielding can sometimes create problems instead of improving things but I did some basic tests on a very short coax RCA cable and felt there could be some potential (alu foil tape then an outer JSSG shield) , its not possible to compare and difference is small so I would like to hear if anyone has experience or advice before continuing. Headphone cable and AC power cords are all over a meter and could be the biggest source of interference in the system

 

Twisted pair wiring for DC cables had great results for me, this wouldnt work the same way with AC would it?

 

Share this post


Link to post
Share on other sites

Twisted pair have result for non-DC only. If DC have pulses, twisted pair can give advantages.

 

Shielding have some theoretical bases for design (grounding lopps, as example). But finally, shielding must be adjusted by measurement tools only.

 

This work is effective, when you adjust full system as single complex, including power and signal cables, boards, cases, etc.

 

Read more information https://samplerateconverter.com/educational/power-conditioner

Share this post


Link to post
Share on other sites
2 hours ago, audiventory said:

Twisted pair have result for non-DC only. If DC have pulses, twisted pair can give advantages.

 

Shielding have some theoretical bases for design (grounding lopps, as example). But finally, shielding must be adjusted by measurement tools only.

 

This work is effective, when you adjust full system as single complex, including power and signal cables, boards, cases, etc.

 

Read more information https://samplerateconverter.com/educational/power-conditioner

The advice for DC twist pair was taken from this post and a lot of people are using it:

https://www.computeraudiophile.com/forums/topic/31554-diy-dc-power-cables/

 

it is worth trying and very simple to test.

 

It probably depends on the application too, LT3045 power supply for a DAC is fairly sensitive.

 

I will experiment with AC power and audio twisted pair, if DC had good results then AC must be great!

 

Also I dont agree passive shielding should only be done with measurements, it is harmless, simple and inexpensive process and results in the most extreme are either slightly bad or slightly good (and so far nothing bad has come out of USB style shielding on RCA). when we can establish the good practices almost anyone should be able to achieve good results, giving more power to anyone with some DIY incentive not just those with access to measurement tools and the training needed to use them.

Share this post


Link to post
Share on other sites
6 hours ago, numlog said:

Twisted pair wiring for DC cables had great results for me, this wouldnt work the same way with AC would it?

Hi Numlog

It's easy enough to find out . Simply get a length of 3 core mains cable, remove the 3 wires from the outer jacket ( the hard part) then  tightly twist them, followed by a layer of heatshrink cable, and then shrinking it with a heat gun, followed by another layer of heatshrink cable to help make up for the thinner outer protective layer.  (3 core mains cables already have a larger radius twist)

Don't forget that twisted pairs are also used in telecommunications cables, from 2 pair to main cables between Telephone Exchanges with  > 1,000 pairs.( Even the 2 pair cables have some  AC. mains rejection properties)

 

Regards

Alex

Share this post


Link to post
Share on other sites
6 hours ago, numlog said:
8 hours ago, audiventory said:

 

The advice for DC twist pair was taken from this post and a lot of people are using it:

 

Twisted pair defend by radiation (from and to cable).

  1. Cables interact with other cables and environment thru air via electro magnetic field.
  2. Current is created in conductor when electro magnetic field is altered.
  3. DC generate constant electro magnetic field.
  4. There is no current in neighbour cunductors by constant electro magnetic field.

Resume:

Thus twisted pair defend cables by constant electro magnetic field that don't generate current in the cables already.

So there is nonexistent issue is fixed.

 

Remember about exception, that I noted in previous post (DC have pulses).
 

 

 

6 hours ago, numlog said:

It is worth trying and very simple to test.

I'm not sure test of effectiveness of noise defending is so simple. There are professional measurement tools, special knowledges and experience are necessary.

 

 

 

6 hours ago, numlog said:

Also I dont agree passive shielding should only be done with measurements

Some coarse issues may be fixed by ears. But fine tuning demands technical instruments.

 

Share this post


Link to post
Share on other sites
43 minutes ago, audiventory said:

Resume:

Thus twisted pair defend cables by constant electro magnetic field that don't generate current in the cables already.

So there is nonexistent issue is fixed.

 

DC cables can have varying voltages in them due to instantaneous load requirements, especially with digital with it's fast rise and fall times.

This also applies to USB cables where it is best to screen the D+ and D- SEPARATELY form +5V and 0 volts to prevent induced voltages into the data pair, DESPITE it being a supposedly twisted pair.

 

There was a good explanation by John Swenson about this issue , linked to in the 3rd post on this page.

Share this post


Link to post
Share on other sites
4 hours ago, sandyk said:

DC cables can have varying voltages in them due to instantaneous load requirements, especially with digital with it's fast rise and fall times.

Yes. Variable load due audio-part-power circuits and switching pulses of power-supply-unit can cause some voltage altering, including fast in some cases. However, first way to solve it are coils and capacitors, connected into DC circuits. I meant it when wrote "DC have pulses".

 

4 hours ago, sandyk said:

This also applies to USB cables where it is best to screen the D+ and D- SEPARATELY form +5V and 0 volts to prevent induced voltages into the data pair, DESPITE it being a supposedly twisted pair.

USB cable have permanent AC component primarully. Quality of USB-cable twisted pair construction impact to losses  in the cable (along distance) and interferention defence.

 

Shielding is additional defence there.

 

The defence is symmetric:

good defence from external impact to cable conductors = good defence from radiation of cable conductors to external environment.

 

Also ferrite rings are set at cables to absorbtion of interference energy.

Share this post


Link to post
Share on other sites
3 minutes ago, One and a half said:

I would not place a shield over an existing no shielded AC cable, since a strand may end up in the wrong place.

Same rules apply for AC cables as far as signal cables are concerned, the garbage stays inside the confines of the screen.

I'm fanatical about avoiding coupling from AC cables to signal cables that can run parallel to each other , so all my AC cables for audio equipment are shielded types with the shield connected to ground (earth) at both ends. With AC cables mixing with small signals in close proximity behind the equipment rack, there's always a chance that a signal cable will receive a spike some day.

 

 Likewise. However, I don't feel the need for shielded A.C. mains cables, and try to keep mains cables well separated from my double screened interconnects, crossing them at 90 degrees wherever possible.

Share this post


Link to post
Share on other sites
51 minutes ago, audiventory said:

Also ferrite rings are set at cables to absorbtion of interference energy

 

This may be fine for printer cables, but usually NOT a good idea for high speed USB

51 minutes ago, audiventory said:

However, first way to solve it are coils and capacitors, connected into DC circuits.

 

I do not agree with that. I prefer to keep the output impedance of the power supply as low as possible, and the DC cables as short as possible, and with as low a resistance as possible, even if that means boring out the end of the D.C. connector a little if needed to accommodate heavier gauge copper wire. Add to that the John Swenson mods if possible without making the cable too stiff.

Share this post


Link to post
Share on other sites
1 hour ago, One and a half said:

I would not place a shield over an existing no shielded AC cable, since a strand may end up in the wrong place.

This happened with the RCA and its metal plugs in the beginning, leaving the shield 5mm or so away from the plug and tight wrapping with a sliver of insulation tape held it in place and forms a barrier, checking for a short between shield and plug ground is pretty easy with a multimeter. not a durable solution but RCAs usually arent moved around much. This would only apply to ungrounded shields

 

For power I agree, wrapping conductive materials around a power cable is a bad idea . 

Share this post


Link to post
Share on other sites
7 hours ago, audiventory said:

I'm not sure test of effectiveness of noise defending is so simple. There are professional measurement tools, special knowledges and experience are necessary.

 

Some coarse issues may be fixed by ears. But fine tuning demands technical instruments.

what would be involved in fine tuning of a passive shield? say you discover the shield is introducing some of its intereference back into ground through measurements, what physical changes to the shield could be made to address that other than simply removing it?

 

From my experience it can be the opposite, coarse issues can be fixed with measurements but a keen ear is needed for fine tuning,  the goal in the end is for it sound better not measure better. measurements, if done correctly, are always reliable and accurate but ears are not, this is the problem.

 

 

Share this post


Link to post
Share on other sites
12 hours ago, sandyk said:

Hi Numlog

It's easy enough to find out . Simply get a length of 3 core mains cable, remove the 3 wires from the outer jacket ( the hard part) then  tightly twist them, followed by a layer of heatshrink cable, and then shrinking it with a heat gun, followed by another layer of heatshrink cable to help make up for the thinner outer protective layer.  (3 core mains cables already have a larger radius twist)

Don't forget that twisted pairs are also used in telecommunications cables, from 2 pair to main cables between Telephone Exchanges with  > 1,000 pairs.( Even the 2 pair cables have some  AC. mains rejection properties)

 

Regards

Alex

The power cable for my amp and DAC is 10 AWG fine strand speaker cable pairs soldered to the plug and transformer on each end, its awkward but it has to be used to know if the differences were in the twists and not the cables. I managed about 2 twists per inch with amp cord, it sounds better. The DAC cord should have a bigger impact but to be able judge those differences some time is needed to familiarise with the current sound before making more changes. 

Then there is the PC cord too.

Share this post


Link to post
Share on other sites
4 hours ago, sandyk said:

This may be fine for printer cables, but usually NOT a good idea for high speed USB

It is necessary to consider in context of certain cable construction and transmitted signal type. I can claim nothing only by cable speed and its general type.

 

 

4 hours ago, sandyk said:

I do not agree with that. I prefer to keep the output impedance of the power supply as low as possible, and the DC cables as short as possible, and with as low a resistance as possible, even if that means boring out the end of the D.C. connector a little if needed to accommodate heavier gauge copper wire.

Impedance impact of capacitors and coils is not matter for DC wires/cables. Because capacitors and coils imact to AC part of current. And AC should be just eliminated. For DC capacitors and coils are transparent.

 

 

3 hours ago, numlog said:

what would be involved in fine tuning of a passive shield? say you discover the shield is introducing some of its intereference back into ground through measurements, what physical changes to the shield could be made to address that other than simply removing it?

What is passive shield? I know simple shield. It eliminate electromagnetic radiation. Correct connecting shield to ground demands special skills.

 

To conrol of shielding result, oscillogram and spectrum in different system points (power or/and signal wires) are checked.

 

If the shielding eliminate noise or unnecessary oscillations in the system point, then shielding is effective.

 

 

3 hours ago, numlog said:

the goal in the end is for it sound better not measure better. measurements, if done correctly, are always reliable and accurate but ears are not, this is the problem.

 

Sound quality have 2 definitions:

1. High fidelity - close to original,

2. Nicer sound - subjective perception; sound enhancing.

 

Real record have distortions in recording and playback.

In case #2 level of distortions is not matter. But all these distortions are estimated. There are special processing (tape, tube, vinyl emulators) exists even.

For case #1, distortions also present. And sound depend on balance of the distortions.

Share this post


Link to post
Share on other sites
57 minutes ago, audiventory said:

What is passive shield? I know simple shield. It eliminate electromagnetic radiation. Correct connecting shield to ground demands special skills.

 

To conrol of shielding result, oscillogram and spectrum in different system points (power or/and signal wires) are checked.

 

If the shielding eliminate noise or unnecessary oscillations in the system point, then shielding is effective.

 

 

Sound quality have 2 definitions:

1. High fidelity - close to original,

2. Nicer sound - subjective perception; sound enhancing.

 

Real record have distortions in recording and playback.

In case #2 level of distortions is not matter. But all these distortions are estimated. There are special processing (tape, tube, vinyl emulators) exists even.

For case #1, distortions also present. And sound depend on balance of the distortions.

I mean there is no connection to ground or anything with a passive shield , passive is probably the wrong term.

 

The aim is for highest fidelity possible with computer audio. To gauge the most accurate and realistic sound it helps to be familiar with a large and varied selection of suitably* recorded music, again it's not perfect process but benefit of certain changes can always be reevaluated at a later time (after the wow factor wears off).

Some distortions can sound nice but I have only experienced this in analogue stage, for interference/noise/distortion etc. on digital side (PC, DAC) reducing these things always helps, subjectively and objectively.

Share this post


Link to post
Share on other sites
37 minutes ago, numlog said:

I mean there is no connection to ground or anything with a passive shield , passive is probably the wrong term.

I understand now that you meant. I don't know exact term too.

Shuch shield works like to ferrite ring, I think.

Share this post


Link to post
Share on other sites
1 hour ago, numlog said:

I mean there is no connection to ground or anything with a passive shield , passive is probably the wrong term

I agree not good term, but what is correct one? Not sure if are correct terms for audiophile use, but maybe following useful?

Define shield for these purposes - metal/conductive tube covering wires carrying power or signal to protect them from outside interference.

Define suggested wire/cable shield type terms-

 

1. 'Floating' shield - not connected to other circuits (your "passive")

2. 'Connected' shield - one connected at one end (or along length), probably need qualifiers for type of connection.

3. 'Through' shield - one connected at both ends.

 

Maybe people here use these words in future? Use of consistent words helpful to understanding, as attempted in these threads.

 

 

1 hour ago, numlog said:

To gauge the most accurate and realistic sound it helps to be familiar with a large and varied selection of suitably* recorded music

Agree strongly, but recorded & Live music!

Share this post


Link to post
Share on other sites
6 minutes ago, look&listen said:

I agree not good term, but what is correct one? Not sure if are correct terms for audiophile use, but maybe following useful?

Define shield for these purposes - metal/conductive tube covering wires carrying power or signal to protect them from outside interference.

Define suggested wire/cable shield type terms-

 

1. 'Floating' shield - not connected to other circuits (your "passive")

2. 'Connected' shield - one connected at one end (or along length), probably need qualifiers for type of connection.

3. 'Through' shield - one connected at both ends.

 

Maybe people here use these words in future? Use of consistent words helpful to understanding, as attempted in these threads.

 

 

Agree strongly, but recorded & Live music!

floating shield is a good engineering term, free spirit shield would be audiophile grade term

Share this post


Link to post
Share on other sites
6 hours ago, audiventory said:

For DC capacitors and coils are transparent.

 

 No they aren't transparent .

With DC , filter capacitors, their type (Low ESR etc.) and value help to determine how an Analogue or even a Digital  component sounds. A series inductance changes (increases) the Power supply's output impedance which again determines to some extent how an analogue component especially will sound.

 At the component end of a DC  supply, there will always also be some type of additional parallel capacitor across it's D.C. input. This also helps to determine how the component will sound, because every power supply does have inductance in it's output leads as well as some amount of output impedance.

NO PSU is perfect !

Share this post


Link to post
Share on other sites
7 hours ago, numlog said:

The power cable for my amp and DAC is 10 AWG fine strand speaker cable pairs soldered to the plug and transformer on each end, its awkward but it has to be used to know if the differences were in the twists and not the cables. I managed about 2 twists per inch with amp cord, it sounds better. The DAC cord should have a bigger impact but to be able judge those differences some time is needed to familiarise with the current sound before making more changes. 

Then there is the PC cord too.

 

 Unfortunately, the more twists in the cable the less flexible it becomes .

Share this post


Link to post
Share on other sites
6 hours ago, sandyk said:

With DC , filter capacitors, their type (Low ESR etc.) and value help to determine how an Analogue or even a Digital  component sounds.

Single electrical circuit is designed  different way for AC and DC.

 

Capacitor have almost infinite resistance for DC. By Ohm law is is equivalent that there is nothing.

 

Coil have almost zero resistance for DC. By Ohm law is is equivalent that there is nothing too.

 

In power supply units (PSU) capacitors and coils are included to DC circuits to compensation of short time deviations of voltage due:

 

1. altering of input voltage of the PSU;

2. altering of output load of the PSU (including reason of ocsilations into audio circuit).

 

In this case DC voltage is considered as constant and the deviations are considered as AC. This AC may penetrate to audio circuits.

 

To compensate long term deviations, active electronical circuits (stabilizer) are used.

Share this post


Link to post
Share on other sites

Audiventory

Please stop spouting irrelevant Technical stuff to both numlog and myself and treating us both as coming from completely non technical backgrounds  .

I am not a newbie in this area, and I do have a reasonable background in electronics , partly due to 43 years in Telstra where I was a Principal Telecommunications Technical Officer with 43 years experience, and >60 years as a DIY Electronics hobbyist, so please get into the practical side of things, or is your experience limited to mainly as a Software person, and we are having Language related problems here ? 

Quote

 Coil have almost zero resistance for DC. By Ohm law is is equivalent that there is nothing too.

 

That depends entirely on the amount of inductance of the coil, with designs such as in some PSU designs from John Swenson using inductors of 10mH. 

 

Quote

To compensate long term deviations, active electronical circuits (stabilizer) are used.

I have attached a copy of an  active electronic stabilizer that I have further developed (and use) from an original design by the late John Linsley Hood. It not only has a capacitance multiplier section, (around 2 FARADS in my version) ) but it also has active cancellation of ripple to WAY past 300KHZ.

 The original circuit was only published ** with a current limit of 100mA, but this version is capable of higher performance to > 2Amps in either polarity. For improved performance at voltages as low as +3.3V and +5V, the 10K resistor has been replaced by a 1.4mA Current Regulating Diode.

Alex

 

** Electronics Today International April 1994

JLH PSU - revised drawing 2   8-1-2009  .jpg

Share this post


Link to post
Share on other sites
50 minutes ago, sandyk said:

I was a Principal Telecommunications Technical Officer with 43 years experience, and >60 years as a DIY Electronics hobbyist

 

I was developer and development team lead of communication systems about 18 years. Also I have university education in automation and communications.

 

1 hour ago, sandyk said:

 Coil have almost zero resistance for DC. By Ohm law is is equivalent that there is nothing too.

51 minutes ago, sandyk said:

That depends entirely on the amount of inductance of the coil,

No.

 

I wrote about DC (direct current) only here.

 

Resistance for DC is

R=0.

 

Resistance for AC is

 

X=2*PI*f*L, where

f- frequency;

L-inductance.

 

If you practically got  resistance R difference for coils with different L, it is wire length and diameter matter only, not inductance.

 

1 hour ago, sandyk said:

Please stop spouting irrelevant Technical stuff to both numlog and myself

I don't like discuss with emotional/non-technical arguments. If you want, I will stop answer to you posts in this topic.

 

Share this post


Link to post
Share on other sites
47 minutes ago, audiventory said:

I don't like discuss with emotional/non-technical arguments. If you want, I will stop answer to you posts in this topic.

Please do.

You are contributing nothing of practical value to this thread .

 

Quote

If you practically got  resistance R difference for coils with different L, it is wire length and diameter matter only, not inductance.

 

DUH !

 

 Do you REALLY think that I don't know those formulas and how to use them ?

 The 10mH choke that I mentioned previously has  a DC resistance of 160mOhms and is rated to 5A. Yes, there is a voltage drop across It when current is drawn through it , and of course the wire length and diameter of the wire

matters, as well as whether it is air wound, or wound on a ferrite material.

 

 BTW,  the formula is XL=2*PI*f*L

 

Share this post


Link to post
Share on other sites
9 minutes ago, sandyk said:

Yes, there is a voltage drop across It when current is drawn through it , and of course the wire length and diameter of the wire

matters, as well as whether it is air wound, or wound on a ferrite material. 

 

DC voltage drop don't depend on ferrite or air there.

 

DC and AC voltage drops are different there.

 

 

13 minutes ago, sandyk said:

You are contributing nothing of practical value to this thread .

 

For this subject I known practice according the theory always.

 

This theory about complex numbers:

 

U=Ua + iUr,

 

I=Ia + iIr,

 

Z=R+iX.

 

Ua, Ia, R - active voltage/current/resistance (wire length + diameter);

iUr, iIr, iX - reactive voltage/current/resistance (depend on inductance and/or capacity).

 

Ua and iUr are ortogonal values. You cannot just sum Ra and iRr, like 2 usual active resistors.

 

When you sum Z1 and Z2,

 

you get Z1+Z2=(R1+R2) + i(X1+X2).

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×