1
\$\begingroup\$

I am trying to understand how to choose a coaxial cable (for my current application, I need a 50Ohm coaxial cable that works reasonably well in the DC-10GHZ range). According to the table in section "standards" of Wikipedia, I would choose the cable with the fewer possible "max attenuation", eventually taking into account the size and rigidity of the cable (I don't care about velocity factor). But I don't known if this attenuation measure at 750MHZ does mean something about 10GHZ.

More practical related question:

Does the LMR200 perform reasonably well at 10GHZ?

What about RG402? (I've read somewhere this is the preferred option for 10GHZ radio ham, but I don't understand why).

\$\endgroup\$
5
  • 1
    \$\begingroup\$ " reasonably well " is not very well defined. Make up your mind about that in numbers and read the datasheets of suppliers. There is no single type that will just work, it is a product of a supplier that you have to look at. \$\endgroup\$
    – PlasmaHH
    Dec 20 '17 at 11:33
  • 1
    \$\begingroup\$ You will need to decide what is acceptable to you as the cable will always be a compromise between performance and price. If you want only 0.1 dB loss (just my guess) at 10 GHz for 1 meter of cable expect to pay a lot ($500 perhaps). If you can accept 10 dB loss at 10 GHz over 1 meter cable then your options will be much wider and more affordable. So how much loss is OK? How long does your cable need to be? \$\endgroup\$ Dec 20 '17 at 11:41
  • \$\begingroup\$ I think I can do with 20cm, maybe even 10cm. So, 10 db loss is probably OK. \$\endgroup\$
    – MikeTeX
    Dec 20 '17 at 11:47
  • 1
    \$\begingroup\$ Keep in mind that there are also other metrics in cable performance next to the losses. Examples include phase stability, bend radius, reflections due to bending, dispersion, etc. A cable with extremely low loss but very high dispersion could be better than a higher loss low dispersion cable for say LO distribution, but a horrible choice for a high-speed communication link. \$\endgroup\$
    – Joren Vaes
    Dec 20 '17 at 11:59
  • 1
    \$\begingroup\$ In addition, just how critical is this performance? It could be far cheaper to just get a better, lower noise amplifier to compensate for increased losses in your cable, and just go with a cheaper coaxial cable. \$\endgroup\$
    – Joren Vaes
    Dec 20 '17 at 12:05
3
\$\begingroup\$

10GHz is very high frequency, and good cables for it won't come cheap. You will need to read the datasheets for specific cable products from specific suppliers.

The attenuation at 750MHz will not be a good indicator of performance at 10GHz.

The LMR200 you mention is not suitable. Note that the attenuation graphs and tables stop at 5.8GHz.

Generally thinner cable has higher attenuation loss. However, for any size of cable, there is a maximum frequency above which it will 'mode', that is, allow propagation of other modes in addition to the normal TEM coaxial mode. Above this frequency, all bets on any cable being well-behaved are off. This means that if attenuation is important to you, use the largest diameter cable you can that is still uni-modal at your top frequency.

RG402 is a 'semi-rigid' coax, rated up to 36GHz. This is the best value for money for cables within instruments where you don't need continual flexing. There are two principle sizes of semi-rigid used in instruments, '85thou', and '141', 0.085" and 0.141" outer diameter respectively. The thinner one is easier to bend by hand, but has higher loss than 141.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ RG402 is also the "standard" cable when you go to high-performance metrology connectors at lower frequencies (I see RG402 most often with 3.5 mm connectors). \$\endgroup\$
    – Joren Vaes
    Dec 20 '17 at 12:04

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.