mrfeedback said:
and if we use "reality" as a yardstick in measuring what is true, the following statement can NOT be true:
while noise induced by varying magnetic fields is not perfectly differential on twisted wires, most of us in the real world do consider the noise common mode which can be rejected by a balanced circuit.
I think this whole discussion is turning into an ego fight where people refuse to acknowledge their mistakes, however small they are, and say "I was wrong".
I challenge the same people to get Mr. Whitlock to come out and support their view.
Truths And Untruths..........
Steve, Steve, Steve,
.......Fred who subsequently sent a rather irate EMail to Jensen..
I do not believe Fred would do this under any circumstance, and I believe that this is very much NOT the case.
Insults on the forum is one thing, slander is another and irrevocably unforgivable in my view.
Anybody in my employ who slanders a fellow employee would be recieving an instant DCM*.
Steve, I feel that you have had too many chances around here, and it is time for us to let you go.
Eric.
DCM* - Don't Come Monday.
Steve, Steve, Steve,
.......Fred who subsequently sent a rather irate EMail to Jensen..
I do not believe Fred would do this under any circumstance, and I believe that this is very much NOT the case.
Insults on the forum is one thing, slander is another and irrevocably unforgivable in my view.
Anybody in my employ who slanders a fellow employee would be recieving an instant DCM*.
Steve, I feel that you have had too many chances around here, and it is time for us to let you go.
Eric.
DCM* - Don't Come Monday.
We'll Use The French Legal System..........
Steve, if you are going to make public statements like ".......Fred who subsequently sent a rather irate EMail to Jensen.." you need to be able to substantiate this claim in public too.
If you cannot provide proof of your statement, your statement is libelous, and justification for your termination.
Counsel is listening.
Eric.
Steve, if you are going to make public statements like ".......Fred who subsequently sent a rather irate EMail to Jensen.." you need to be able to substantiate this claim in public too.
If you cannot provide proof of your statement, your statement is libelous, and justification for your termination.
Counsel is listening.
Eric.
Steve, Eric, cut it out. If you want to discuss legal, personal, and moral issues, you can do so via email. This area is for technical discussion.
Yeah, I Have Better Things To Do Too.............
Hi Stuart, I posted my above message before seeing your most recent post.
Yes, this forum is intended for technical discussion, and my above post pretty well sums up all that I have to say on the matter.
If this matter is not resolved justfully, you can bet that I will be sending 'please explain' emails to you.
Respectful regards,
Eric.
Hi Stuart, I posted my above message before seeing your most recent post.
Yes, this forum is intended for technical discussion, and my above post pretty well sums up all that I have to say on the matter.
If this matter is not resolved justfully, you can bet that I will be sending 'please explain' emails to you.
Respectful regards,
Eric.
To Clarify Fred's (Infamous) Statement..........
Consider -
Any length of wire acts as an antenna, and any magnetic disturbance will induce (couple) energy into any antenna.
Concerning balanced line systems, magnetic disturbance sources will induce energy into both signal wires in a pair type cable.
This means that voltages (emf) wrt local system earth will be developed at both ends of both wires, and perfectly valid signals will be induced into both wires.
So, that means that both inputs of a balanced input stage are recieving signals due to magnetic disturbances.
The point of closely twisting the two signal wires of a balanced connection is to cause the signals induced into both wires to be as similar as possible (magnitude and relative phase), by means of averaging the positions of both wires in space wrt magnetic disturbances.
If for example the input stage in consideration is a transformer primary winding, and if the two signals are identical, there will be no potential across this primary winding, and therefore no winding induced energy to couple to the transformer secondary winding.
This is a means of common mode rejection.
Any potential difference across the primary constitutes a valid signal that transfers to the secondary.
Solid state techniques can be used instead of a physical transformer.
Both methods have drawbacks and performance compromises.
Eric.
Hmmm, do we have misunderstanding about balanced pair systems ?.millwood said:
Consider -
Any length of wire acts as an antenna, and any magnetic disturbance will induce (couple) energy into any antenna.
Concerning balanced line systems, magnetic disturbance sources will induce energy into both signal wires in a pair type cable.
This means that voltages (emf) wrt local system earth will be developed at both ends of both wires, and perfectly valid signals will be induced into both wires.
So, that means that both inputs of a balanced input stage are recieving signals due to magnetic disturbances.
The point of closely twisting the two signal wires of a balanced connection is to cause the signals induced into both wires to be as similar as possible (magnitude and relative phase), by means of averaging the positions of both wires in space wrt magnetic disturbances.
If for example the input stage in consideration is a transformer primary winding, and if the two signals are identical, there will be no potential across this primary winding, and therefore no winding induced energy to couple to the transformer secondary winding.
This is a means of common mode rejection.
Any potential difference across the primary constitutes a valid signal that transfers to the secondary.
Solid state techniques can be used instead of a physical transformer.
Both methods have drawbacks and performance compromises.
Eric.
Re: To Clarify Fred's (Infamous) Statement..........
While that description's consistent with the Whitlock references and how I've always understood it, it's inconsistent with the capgo reference and one or two others which says that twisting causes noise cancellation to take place in the conductors themselves rather than downstream by the differential input.
In other words, Whitlock seems to be saying that twisting, rather than cancelling any noise in the cable itself, simply averages the distances between the two conductors and the source, effectively converting more of what would otherwise be differential voltages to common-mode voltages which are then rejected at the differential input whereas the capgo reference says that the cancellation's taking place in the cable itself which to me implies that the common-mode rejection capability of the receiver is moot.
se
mrfeedback said:
While that description's consistent with the Whitlock references and how I've always understood it, it's inconsistent with the capgo reference and one or two others which says that twisting causes noise cancellation to take place in the conductors themselves rather than downstream by the differential input.
In other words, Whitlock seems to be saying that twisting, rather than cancelling any noise in the cable itself, simply averages the distances between the two conductors and the source, effectively converting more of what would otherwise be differential voltages to common-mode voltages which are then rejected at the differential input whereas the capgo reference says that the cancellation's taking place in the cable itself which to me implies that the common-mode rejection capability of the receiver is moot.
se
How Much Can A Koala Bear.............
I was being deliberately simplistic in my above explanation.
Further influences are the conductors shielding each other, and the pair twisting cycle length versus wavelength and direction of magnetic inteference sources.
Different coloured insulations affect too.
Fred, Whitlocks and Capgo are all correct.
Eric.
I was being deliberately simplistic in my above explanation.
Further influences are the conductors shielding each other, and the pair twisting cycle length versus wavelength and direction of magnetic inteference sources.
Different coloured insulations affect too.
Fred, Whitlocks and Capgo are all correct.
Eric.
Hi,
Both are correct but are discussing different aspects of noise cancellation:
One addresses the aspects of externally induced noise and the effect it has on a twisted pair, the other refers to induced noise from the source already and the effect it has on the twisted pair.
An enormous advantage of a twisted pair, whether in single ended connection or in balanced connection is that any alternating current carried by it will have a reduced magnetic radiation to the outside world.
In this case the AC current is carried on both wires with opposing polarity but of equal strength.
Another advantage is that it is much less susceptible to magnetic radiation from the outside world, in other words its twisting nulls most of the radiating external field unless more than one external magnetic field of VARYING intensity is operative at close proximity to the wires.
In the latter case the differential field will ride on the wires as a voltage and will show up at the receiving end as a voltage. In case of a balanced connection this phantom voltage will be seen
as a signal at the receiving end we'll have EMI disturbance as a result.
In an unbalanced SE single ended system the sum of both magnetic forces wil be presented, much reduced thanks to cancellation in the twisted pair but at full level in case of parallel wire runs as one will radiate into the other and vice versa.
Enter the drain wires and all kind of degrees of extra shielding...
If anyone wants to experiment with twisted pairs I'd recommend running the heaters of high µ tubes from AC with parallel wires or with a tightly twisted pair, the differences are simply night and day.
Hope I didn't goof here as it's been more than 10 years since I last dabbled with this kind of stuff...
Cheers,
Both are correct but are discussing different aspects of noise cancellation:
One addresses the aspects of externally induced noise and the effect it has on a twisted pair, the other refers to induced noise from the source already and the effect it has on the twisted pair.
An enormous advantage of a twisted pair, whether in single ended connection or in balanced connection is that any alternating current carried by it will have a reduced magnetic radiation to the outside world.
In this case the AC current is carried on both wires with opposing polarity but of equal strength.
Another advantage is that it is much less susceptible to magnetic radiation from the outside world, in other words its twisting nulls most of the radiating external field unless more than one external magnetic field of VARYING intensity is operative at close proximity to the wires.
In the latter case the differential field will ride on the wires as a voltage and will show up at the receiving end as a voltage. In case of a balanced connection this phantom voltage will be seen
as a signal at the receiving end we'll have EMI disturbance as a result.
In an unbalanced SE single ended system the sum of both magnetic forces wil be presented, much reduced thanks to cancellation in the twisted pair but at full level in case of parallel wire runs as one will radiate into the other and vice versa.
Enter the drain wires and all kind of degrees of extra shielding...
If anyone wants to experiment with twisted pairs I'd recommend running the heaters of high µ tubes from AC with parallel wires or with a tightly twisted pair, the differences are simply night and day.
Hope I didn't goof here as it's been more than 10 years since I last dabbled with this kind of stuff...
Cheers,
Re: How Much Can A Koala Bear.............
Looking at each on its own leads to two different conclusions. In the Whitlock reference where it's stated as a matter of relative proximity, total noise remains the same and it's only a matter of how much is differential versus how much is common-mode.
Whereas in the capgo reference, it has nothing to do with relative proximity but rather opposing voltages cancelling each other within the cable itself, reducing the total noise.
I don't see how these two square with each other.
se
mrfeedback said:
Looking at each on its own leads to two different conclusions. In the Whitlock reference where it's stated as a matter of relative proximity, total noise remains the same and it's only a matter of how much is differential versus how much is common-mode.
Whereas in the capgo reference, it has nothing to do with relative proximity but rather opposing voltages cancelling each other within the cable itself, reducing the total noise.
I don't see how these two square with each other.
se
fdegrove said:
Did you read the Capgo reference? They're clearly addressing externally induced noise just as the Whitlock reference does. The title is "Magnetically Induced Noise." What are you reading in the Capgo reference that leads you to believe that they're talking about noise that's already in the system?
Under the section on Loop Area Minimization they say:
Reducing the area of the antenna loop is an very effect method method of reducing magnetic pickup and there are many ways in which it may be achieved.
If the noise is already in the system, then why are they talking about the noise being picked up in the cabling?
Yes. In that case you have what amounts to the magnetic field of each conductor cancelling the field of the other. Or to put it another way, the field is concentrated in the area between the conductors. But in this case, you've already got current flowing in the wires whereas what we're interested in here is avoiding any current flowing in the wires as that would mean there's a differential voltage between the wries which will simply be passed on downstream.
Now this is where I think things get a bit sticky.
The only way the conductors can null the external magnetic field impinging on them is if they're carrying current. But we can't have the conductors carrying current because that would mean there is a differential voltage between them which will not be rejected by the differential input.
So it seems the only way to deal with external fields is to make them equal across each conductor, in other words, common-mode, which is the way Whitlock puts it. Twisting the conductors results in each conductor having on average the same distance from the source as the other, so that while you don't reduce the total induced voltage, you make them more common so that ideally no current flow results when used in a balanced interface.
But Capgo is saying that twisting causes the induced voltages to cancel within the cable itself so that instead of simply causing the induced voltage to be common-mode, it's simply not induced at all (at least in the ideal).
And this is where I see the dichotomy between the two.
Again, Whitlock seems to be saying that twisting gives you the same total voltage as a parallel pair, but more of it is common-mode whereas capgo seems to be saying that twisting gives you a lower total voltage.
And if Capgo's explanation is correct, then there would be no difference in how much noise gets amplified at the end of the line whether you're using a balanced interface or an unbalanced interface.
What are you suggesting be measured?
I think the better way to approach this at least in terms of figuring out whether twisting is effective due to keeping each conductor the same distance from the source as the other or whether it's due to cancellation within the cable itself would be to set up a balanced line using both a parallel and twisted pair.
Then measure the voltage across each line. If what Whitlock seems to be saying is the case, the voltage across each line shouldn't change appreciably between parallel and twisted pair, whereas if what Capgo seems to be saying is the case, it should be considerably less.
se
fdegrove said:
You said:
Another advantage is that it is much less susceptible to magnetic radiation from the outside world, in other words its twisting nulls most of the radiating external field...
How do you null a magnetic field? With another magnetic field of the same magnitude but opposite polarity. So if the wires are going to null a magnetic field, they'd have to have current flowing through them.
See above.
Didn't you read the Whitlock reference? By twisting, you cause each conductor to on average be the same distance from the source. If each conductor is equidistant from the source, the magnetic field across each conductor will be the same on average as the other.
Oh, ok, I see what you're saying now.
But what would that do to show HOW the hum was being reduced? In other words, how would that tell you whether the hum was being reduced because differential voltages were being made more common-mode a la Whitlock or due to cancellation within the cable a la Capgo?
We all know that twisted pairs are efficacious. What I've been questioning is WHY, as there are two means described which to me seem at odds with each other.
se
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.