Hi everyone,

I am doing my master thesis on crosstalk. I can calculate crosstalk through simulations and practically but unable to calculate theoritcally, if anyone know the calculation of crosstalk in twisted pair quad cables. It will be helpful to me. i am giving voltages to one pair and want to calculate voltages induced in another pair.so, i can calculate crosstalk. I am attaching a photo to understand it clearly.

Specifications:
V1=10V
V2=0A
V3= -10V
V4=0A
Core diameter: 4.05mm
Screen: 20.7+0.15 mm
Outer sheath:24.5mm
Twist: 180mm*cos(9.7[deg])= 1 laylength

 

Cross talk at DC?

Then it's just about leakage currents?
I think you must be talking about AC signals?

Please clarify the question?

 

Hi everyone,

I am doing my master thesis on crosstalk. I can calculate crosstalk through simulations and practically but unable to calculate theoritcally, if anyone know the calculation of crosstalk in twisted pair quad cables. It will be helpful to me. i am giving voltages to one pair and want to calculate voltages induced in another pair.so, i can calculate crosstalk. I am attaching a photo to understand it clearly.

Specifications:
V1=10V
V2=0A
V3= -10V
V4=0A
Core diameter: 4.05mm
Screen: 20.7+0.15 mm
Outer sheath:24.5mm
Twist: 180mm*cos(9.7[deg])= 1 laylength

View attachment 223148

Cross talk at DC?

Then it's just about leakage currents?
I think you must be talking about AC signals?

Please clarify the question?

Yeah, I am talking about the AC signals and the voltage is induced due to inductive and capacitive coupling on conductor 2 and 4. so , i can calculate crossstalk by difference between the voltages 2 &4 by difference between 1 & 3.

Crosstalk = (V2-V4)/(V1-V2)= (V2-V4)/ (10+10) = (V2-V4)/20

So, This is the reason i want to calculate induced voltages in conductor 2 and 4. Consider the length of the cable as L.

 

This one is beyond my skill level, but i have a feeling the answer is very complex. Theoretically, there should be zero cross-talk between two ideal ballanced twisted pairs? The cross-talk comes from non-ideal behaviors, if I am not mistaken.

I also believe that the answer is frequency dependant, making it all the more tricky.

Maybe some others can chime in with more expertise?

 

hi 199723,
There are a number of videos on that subject, on this link.
https://www.youtube.com/results?search_query=calculating+cross+talk+in+twisted+pairs

E

 

Crosstalk between adjacent conductors is the result of non-identical coupling between either pairs or single conductors, and is a function of the cables being non-ideal. The phone companies have done huge amounts of research on the topic relative to their large cables. So your research should include what you have gained from studying that source, which should answer your questions very well.
Unfortunately my fee for writing another person's master thesis is very high these days and that is why instead I just pointed to the sources that will answer your questions.

 

Hi everyone,

I am doing my master thesis on crosstalk. I can calculate crosstalk through simulations and practically but unable to calculate theoritcally, if anyone know the calculation of crosstalk in twisted pair quad cables. It will be helpful to me. i am giving voltages to one pair and want to calculate voltages induced in another pair.so, i can calculate crosstalk. I am attaching a photo to understand it clearly.

A simple google search on "multiconductor transmission lines with twist" (without the quotes), yields lots of research papers. The leads on IEEE Explore should provide lots of fruit.

I thought that as well as twisting, to further reduce crosstalk in Cat5/6 cables they had different twist pitches on the different pairs, but I could be wrong. Maybe in some cables. Does anyone know?

 

A simple google search on "multiconductor transmission lines with twist" (without the quotes), yields lots of research papers. The leads on IEEE Explore should provide lots of fruit.

I thought that as well as twisting, to further reduce crosstalk in Cat5/6 cables they had different twist pitches on the different pairs, but I could be wrong. Maybe in some cables. Does anyone know?

I knew there would be a whole lot of research on cross-talk because it is a problem for the phone companies. As an experiment years ago I connected an AC coupled balanced input amplifier to our phoneine, with all our phones on-hook, so no dial tone. I could hear lots of talking all at once. That was coupling from other pairs to our pair.(cross-talk) None of it could be heard when we were using the phone, but it was there at some level when not using it. But that was 60 years ago, phone lines might be different today. But cross-coupling is certainly real.

The different twist pitches in cat5 cable are intended to reduce crosstalk between pairs. At least that is what the manufacturers claim it does.Certainly it does keep the cables farther apart in that jacket, and if that is the mechanism of reducing coupling then evidently it works. If they all had the same twist pitch they could get much closer.

 

I knew there would be a whole lot of research on cross-talk because it is a problem for the phone companies. As an experiment years ago I connected an AC coupled balanced input amplifier to our phoneine, with all our phones on-hook, so no dial tone. I could hear lots of talking all at once. That was coupling from other pairs to our pair.(cross-talk) None of it could be heard when we were using the phone, but it was there at some level when not using it. But that was 60 years ago, phone lines might be different today. But cross-coupling is certainly real.

The different twist pitches in cat5 cable are intended to reduce crosstalk between pairs. At least that is what the manufacturers claim it does.Certainly it does keep the cables farther apart in that jacket, and if that is the mechanism of reducing coupling then evidently it works. If they all had the same twist pitch they could get much closer.

I think that there is another benefit apart from spacing, Imagine a twisted pair next to an untwisted pair - in this case the twisting will reduce the coupling as each half-twist reverses the sign of the coupling - so it will tend to average towards zero along the length. Now if both pairs are twisted, then the local coupling tends to retain the sign if both pairs have the same twist pitch - as each time the source pair has a half-twist - changing the sign of the coupling, the receiving pair does the same - undoing the change in sign so it adds in phase to the coupling from the previous half twist. If the twist pitches are different then the coupling phase is more random and you get some cancellation.

I only had peripheral involvement years ago with some ADSL. IIRC coupling between nearby pairs in a bundle tended to make a backward wave coupler leading to near-end crosstalk being a bigger problem than far-end crosstalk. Not sure how universal this is, or whether it is relevant to the OP, but here are some measurements supporting that https://www.hpl.hp.com/hpjournal/95aug/aug95a3a.pdf