1
\$\begingroup\$

I am working between 1-5 MHz, pure sine wave from signal generator.

I have been using transformer to match the impedance of source (50 ohm) to load (2-10 Ohm).

My question is does it matter that I need to match the characteristic impedance of the coaxial cable on primary (connect from source to primary coil) and, specifically, secondary transmission (connect from secondary to load)?

Edit: the cable is about 1 metre or less

\$\endgroup\$
  • 1
    \$\begingroup\$ When you say 1-5 MHz, is that the bandwidth of the signal? Or is the 5 MHz signal something like a square wave or PWM type signal? I suggest you draw a diagram, and show each cable length. Basically, if the cable is short, you don't have to worry much about intrinsic impedance. "Short" means that the roundtrip flight time in the cable is much less than the period of the highest frequency component in your signal. \$\endgroup\$ – mkeith Sep 8 '15 at 20:48
  • 1
    \$\begingroup\$ Whether you need to match may depend on the power levels, or the fragility of your signal source. Matching on the primary side is easy; on the secondary less so. This suggests an electrically short (<< 1/4 wave) connection from sec to load, i.e. place transformer close to load. \$\endgroup\$ – Brian Drummond Sep 8 '15 at 20:49
4
\$\begingroup\$

If the time delay though the cable is small relative to the wavelength of the highest frequency you are transmitting then termination is not important.

Be aware that if you have a pulse rather than a sine wave at 5MHz for example there maybe harmonic components that are much higher than 5Mhz and that may cause problems.

If for example you are sending a digital signal over a coax to be used as the clock into a register at the receiving end then although the fundamental may be low in frequency you may get reflections that could cause double clocking if the cable is not terminated.

Assuming the cable is 1 meter long with a velocity factor of ~0.6 the two way delay will be:

2/(3*10**8 * 0.6) = 11ns. This is equivalent to 89MHz.

If we want this to be 10 times the highest frequency we shouldn't have any frequency components > 8.9MHz. The 5MHz figure you quote meets this.

\$\endgroup\$
  • \$\begingroup\$ Would there be any reflection on the secondary transmission? \$\endgroup\$ – Nhân Lê Sep 8 '15 at 23:04
  • \$\begingroup\$ @NhânLê - Yes the secondary transmission will also be reflected but since every reflection reduces the amplitude (even if terminated properly) it is unlikely that a second reflection would be large enough to cause problems. \$\endgroup\$ – Kevin White Apr 9 '17 at 15:37

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.