0
\$\begingroup\$

I fabbed the gerbers of the eval board for the SKY67153-396LF and used the component values in the datasheet for the 2300 to 2700 MHz band. I tested it yesterday and got some interesting results. There are several different regions of behavior that it has, most of which seem wrong. Here's a video of me sweeping the input power and you can see what's happening: https://www.youtube.com/watch?v=IT8x9wOr9ag Here's a version where you can see the current being used by the amp: https://www.youtube.com/watch?v=eZGxI3nrgM8

It's kind of hard to see the spectrum analyzer, but the marker is at 2.52 GHz, the same as the input freq. The span is from 0 to 6.5 GHz. The amp has very little gain until the input power gets up to a few dBm, and doesn't get close to the correct gain spec until 7 dBm. That's also when all the mixing products (or whatever those spurs are) go away.

Can someone help me understand what's going on in these different regions? What causes those mixing products? Why doesn't amplification begin until a particular input power level?

\$\endgroup\$
  • 1
    \$\begingroup\$ Sounds like it might be oscillating. How carefully have you followed the layout? \$\endgroup\$ – Adil Malik May 16 '17 at 18:56
  • 1
    \$\begingroup\$ Did you get the same components or just the same values? At 2.3GHz, not all capacitors do not behave equally well. \$\endgroup\$ – Joren Vaes May 16 '17 at 18:58
  • 1
    \$\begingroup\$ Are all the inductors rated for the frequencies? Double check all Self Resonant Frequencies of your inductors against the ones in the original BOM \$\endgroup\$ – Adil Malik May 16 '17 at 18:59
  • \$\begingroup\$ And although I think it will make a significant impact at your frequency given the size of the board, but did your PCB manufacturer follow the same board stackup? \$\endgroup\$ – Joren Vaes May 16 '17 at 19:03
  • 1
    \$\begingroup\$ Yes @joren you are right, the Er alone would probably not affect it that much, so the Rogers is probably more for the 6GHz end of the usable range. \$\endgroup\$ – mng May 16 '17 at 19:45
1
\$\begingroup\$

I think I figured out what was going on. My stackup was slightly different than the skyworks eval board, but not by much. I calculated that it should only change my CPW impedance by a few ohms. But I realized that I neglected to order high Q passives. Once I repopulated the board with high Q inductors and capacitors, the oscillation went away: https://www.youtube.com/watch?v=Fq7UKLlVpa4

The difference GJM and GRM series from Murata really really matters! My gain is still not quite what it should be, but I'm going to fab these boards again and put a lot more ground via stitching in (another thing I forgot to do) and see if that improves performance at all.

\$\endgroup\$

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.