4
\$\begingroup\$

I was reading online about how Ethernet cables use twisted pairs instead of coaxial cables because it's better in some way. Some websites say that twisted pairs cost less, can carry higher frequencies, and filter out noise, cross talk, interference, etc.

This website here says that "The wires are twisted together to reduce noise" and this website here says that twisted pairs provide "high data transmission rates"

If that's the case, why are most oscilloscope probe cables made using coaxial cables? Maybe some advantage I'm missing out on?

I thought oscilloscopes are supposed to capture a higher and wider range of frequencies than ethernet cables, so a better ethernet cable would be made up of coaxial cable.

But after looking at the benefits of ethernet cables, I thought that the better oscilloscope probe cable would be made up of twisted pairs. Hence why I am confused right now.

\$\endgroup\$
5
  • 3
    \$\begingroup\$ Back in the day when Ethernet was invented, it did use coaxial cable as the transmission medium. It would be nice if you would say which sources make these claims so we could say if they are trustworthy or not. \$\endgroup\$
    – Justme
    Dec 6, 2021 at 11:54
  • \$\begingroup\$ @Justme wireandcabletips.com/… \$\endgroup\$
    – loadingnow
    Dec 6, 2021 at 12:01
  • 3
    \$\begingroup\$ That link should be edited into your question, so it is not required to scroll through comments. Also, nowhere in that article is claimed that twisted pairs are better, in fact to me it says coaxial cables are better but for practical installation and cost reasons twisted pairs are used for networking. Please quote the part you think applies to your question. \$\endgroup\$
    – Justme
    Dec 6, 2021 at 12:06
  • 1
    \$\begingroup\$ Some websites say that twisted pairs cost less, can carry higher frequencies, and filter out noise, cross talk, interference, etc. Then these websites are not telling the complete story. For some applications (like ethernet), this is indeed correct. For others (like Gighertz signals) it is not. For very high frequencies like more than a few Gigahertz, twisted pair cables are useless. \$\endgroup\$ Dec 6, 2021 at 12:17
  • \$\begingroup\$ Oscilloscope probes also have specially designed coax (I believe it uses a helical, high-resistivity center conductor?) to make its characteristics as ideal as possible for probing uses. \$\endgroup\$
    – Hearth
    Dec 6, 2021 at 14:05

2 Answers 2

10
\$\begingroup\$

If that's the case, why are most oscilloscope probe cables made using coaxial cables?

Unless you have a specialist differential probe fitted on your o-scope input terminal, the input to the bare terminal is what is known as "unbalanced". This means that the input has a locally grounded reference terminal connected to the o-scope chassis.

  • Coaxial cable is ideal for an unbalanced input but, it can be far from ideal when dealing with a balanced signal.

  • Twisted pair is ideal for a truly balanced input. The scope input is not balanced (unless using a specialist probe) therefore it is far from ideal.

  • An oscilloscope input is high impedance (a feature) and, twisted pair requires termination resistors (circa 50 to 100 Ω) for best performance at high frequencies hence, you also have a clash of requirements.

\$\endgroup\$
5
  • \$\begingroup\$ @loadingnow What he means is that an oscilloscope should not load the signal source so it should have high input impedance, but if you use TP you should terminate the system with a lower impedance. That's the clash of requirements. \$\endgroup\$ Dec 6, 2021 at 11:57
  • \$\begingroup\$ Okay, twisted pair requires low impedance termination, then am i right to say that coaxial cables do not require low impedance termination for high performance? \$\endgroup\$
    – loadingnow
    Dec 6, 2021 at 11:58
  • 3
    \$\begingroup\$ @loadingnow Technically coaxial cables require termination to characteristic impedance too, typical coaxial cables use 50 ohm termination which some scopes have internally available, or in some applications also 75 ohms is also used. But oscilloscope probes use a special lossy coaxial cable which does not require termination. \$\endgroup\$
    – Justme
    Dec 6, 2021 at 12:01
  • \$\begingroup\$ @loadingnow depends on what you mean by "high performance". For example, if the cable is meant to be used for measurement of microwaves then the termination impedance should be 50R. The cable is still a coax. \$\endgroup\$ Dec 6, 2021 at 12:03
  • \$\begingroup\$ @loadingnow basic oscilloscope probes are probably only reliable in terms of reasonable amplitude reproduction up to about 50 MHz (or go for a specialist probe). The golden rule to avoid significant reflections in wires is keep the wavelength of the maximum frequency longer than 10x the length of the probe. So, if the probe is 1.0 metres long (just an example), it has a full \$\lambda\$ at 300 MHz, so 50 MHz won't be too bad. Plus, as frequency rises, the internal probe capacitance does achieve some partial lowering of impedances and reflections are reduced this way too. It's a compromise. \$\endgroup\$
    – Andy aka
    Dec 6, 2021 at 12:31
3
\$\begingroup\$

The perfect oscilloscope would have zero effect on the signals in the circuit under measurement.

Using a twisted pair increases the capacitive coupling between its two conductors compared to side-by-side wiring. It makes the resultant capacitance between conductors and their inductance more consistent.

That is what you don't want to connect across your signal of interest and its ground/Vref. It will interfere with high frequency signals, making the displayed measurement even less true than what the imperfect 'scope already had.

A twisted pair is used into a differential and balanced input. The scope doesn't have that, though it could. But the disadvantages of twisted pair make it an avenue not to go up.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Are you saying that coaxial cables have less capacitance compared to twisted pairs? \$\endgroup\$
    – loadingnow
    Dec 6, 2021 at 12:38
  • \$\begingroup\$ @loadingnow Tony hasn't answered this but I will in his absence. Absolutely coax can have much less capacitance between shield and inner compared to twisted pairs. Twisted pairs seek to reduce the loop inductance at the expense of bigger capacitance. Coax can do the opposite. \$\endgroup\$
    – Andy aka
    Dec 7, 2021 at 14:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.