Increasing loss of cables to attain dynamic range

Question

RF newbie here. Trying to learn about cables and losses and how they contribute to dynamic range.

I'm creating a shielding effectiveness system for enclosures with a 120dB dynamic range using signal generators, power amplifiers, transmitter and receiver antennas (such as 12" loop, biconical, dipole, log periodic and gain horn antenna for different frequency ranges) pre-amplifier and spectrum analyzer. The frequency range tested is 9kHz to 40GHz.

Thus, is it acceptable to use a long cable to increase loss in order to attain a dynamic range of 120dB, since the gain of the system is relatively higher? For example, the DANL of the spectrum analyzer is -130dB for 9kHz, hence by theoretical calculation and working backwards by taking into account gain of amplifier, antennas and pre amp, I am trying to figure out the cable length between each instrument in order meet the DR.

Also, would it make more sense to use different coax cables for different bands of frequency such as upto 18GHz, 26GHz and 40GHz when designing the system for each frequency band? Or would a 40GHz low loss coax cable meet the needs of the system all the way from 9kHz to 40GHz? However, the attenuation of the 40GHz cable at 9KHz is almost negligible which makes it hard to attain the desired dynamic range.

Any advice on cable lengths and losses would be appreciated as currently I'm working on determining the length of the cable from signal generator to amplifier, amplifier to tx antenna, rx antenna to pre amp and pre amp to spectrum analyzer in order to achieve a DR of 120dB for various frequencies between 9kHz to 40GHz.

Answer 1

Advice on cable lengths - as short as possible consistent with connecting the equipment you need to. But long enough to move equipment for debugging without cable strain or disconnecting, don't stint yourself. Do not increase the lengths of cables to obtain loss.

Cables that are good (have low loss and good match) to 40GHz tend to be expensive, with easily damaged connectors. It is a good idea to have some more robust lower frequency cables as well, for system setup and debug. Once connected though, it is better to leave the good cable in place without frequent reconnection.

Where you need to increase loss in a link, use an attenuator, or several small pads in series. These will have attenuation that varies less with frequency, usually described as 'flatter', than that for cable.

Often you will 'pad' (connect an attenuator to) the output of a sig-gen or the input of a spectrum analyser to improve the port match, which improves the level measurement uncertainty.

Answer 2

In addition to use44635s excellent answer on your cables, in Amateur (HAM) radio we very often use signal attenuators to drop a more powerful incoming signal down to where we get a better reading in our signal strength meters for locating signal sources. The same attenuators could be applied very successfully to your case here.

• First, determine the exact dB attenuation you need for each signal;
• Second, here is an excellent tutorial for making "pi network" attenuators, and here is a nice, quick calculator for determining your needed resistor values;
• Third, shield, shield, and shield again! Every attenuator needs to be independently shielded & the input/output lines from the attenuator need shielded from each other (to ensure that the strong inclming signal doesn't "leak" around the attenuator through EMI/RFI to show up stronger than expected in the output).