Author Topic: Effect of Coax Test Connector on Signal Integrity  (Read 1156 times)

0 Members and 1 Guest are viewing this topic.

Offline yrrapt

  • Contributor
  • Posts: 11
  • Country: scotland
Effect of Coax Test Connector on Signal Integrity
« on: February 07, 2014, 02:57:43 pm »

I'm making up a PCB for a software defined radio and I am wondering about the ramifications of placing a coax connector between the LNA output and mixer input.  I want to be able test the LNA/mixer in isolation in circuit by lifting coupling capacitors/0R resistors to ensure they match the measurements I made before on isolated PCBs. 

Assuming that I place the footprint for an SMA connector with the signal pin in line with the LNA and mixer to trace to avoid any extra trace length or stubs, will the presence of the footprint have any negative effects on the signal integrity?  And what difference would there be between the effect of populating and not populating the connector?

If it's useful: I only really care about 1.575GHz (GPS), the LNA is 50 ohm output and the mixer IC input is 75 ohm so matching is used and the PCB will be standard FR4.

Thanks in advance for any help you can provide,

Offline ConKbot

  • Super Contributor
  • ***
  • Posts: 1220
Re: Effect of Coax Test Connector on Signal Integrity
« Reply #1 on: February 07, 2014, 05:51:48 pm »
Your wavelength is ~9.5 cm in FR4(longer with a coplanar waveguide on the surface), so if your stubs are shorter than 5mm or so you shouldnt be too badly affected.  However the thoughhole/via can start making stuff a bit funky.  I'd put a footprint for a SMT jack (  if you can fit your coplanar waveguide between the 0.125" spacing of the feet.  And put your 0R/coupling cap as close as possible to the jack.

Oh and I'd also pick up at least a short SMA-SMA cable that is really flexible, rather than a standard cable so you are less likely to accidentally break the connector off ;)

Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo